The Critique of PureReason

Immanuel Kant

Wor

k re

prod

uced

with

no

edito

rial

resp

onsi

bilit

y

Notice by Luarna Ediciones

This book is in the public domain becausethe copyrights have expired under Spanish law.

Luarna presents it here as a gift to its cus-tomers, while clarifying the following:

1) Because this edition has not been super-vised by our editorial deparment, wedisclaim responsibility for the fidelity ofits content.

2) Luarna has only adapted the work tomake it easily viewable on common six-inch readers.

3) To all effects, this book must not be con-sidered to have been published byLuarna.

www.luarna.com

INTRODUCTION

I. Of the difference between Pure and EmpiricalKnowledge

That all our knowledge begins with experiencethere can be no doubt. For how is it possiblethat the faculty of cognition should be awak-ened into exercise otherwise than by means ofobjects which affect our senses, and partly ofthemselves produce representations, partlyrouse our powers of understanding into activ-ity, to compare to connect, or to separate these,and so to convert the raw material of our sen-suous impressions into a knowledge of objects,which is called experience? In respect of time,therefore, no knowledge of ours is antecedentto experience, but begins with it.

But, though all our knowledge begins with ex-perience, it by no means follows that all arisesout of experience. For, on the contrary, it is qui-te possible that our empirical knowledge is acompound of that which we receive throughimpressions, and that which the faculty of cog-nition supplies from itself (sensuous impres-sions giving merely the occasion), an additionwhich we cannot distinguish from the originalelement given by sense, till long practice hasmade us attentive to, and skilful in separatingit. It is, therefore, a question which requiresclose investigation, and not to be answered atfirst sight, whether there exists a knowledgealtogether independent of experience, and evenof all sensuous impressions? Knowledge of thiskind is called a priori, in contradistinction toempirical knowledge, which has its sources aposteriori, that is, in experience.

But the expression, "a priori," is not as yet defi-nite enough adequately to indicate the whole

meaning of the question above started. For, inspeaking of knowledge which has its sources inexperience, we are wont to say, that this or thatmay be known a priori, because we do not de-rive this knowledge immediately from experi-ence, but from a general rule, which, however,we have itself borrowed from experience. Thus,if a man undermined his house, we say, "hemight know a priori that it would have fallen;"that is, he needed not to have waited for theexperience that it did actually fall. But still, apriori, he could not know even this much. For,that bodies are heavy, and, consequently, thatthey fall when their supports are taken away,must have been known to him previously, bymeans of experience.

By the term "knowledge a priori," therefore, weshall in the sequel understand, not such as isindependent of this or that kind of experience,but such as is absolutely so of all experience.Opposed to this is empirical knowledge, or that

which is possible only a posteriori, that is,through experience. Knowledge a priori is ei-ther pure or impure. Pure knowledge a priori isthat with which no empirical element is mixedup. For example, the proposition, "Everychange has a cause," is a proposition a priori,but impure, because change is a conceptionwhich can only be derived from experience.

II. The Human Intellect, even in an Unphiloso-phical State, is in Possession of Certain Cogni-tions "a priori".

The question now is as to a criterion, by whichwe may securely distinguish a pure from anempirical cognition. Experience no doubt tea-ches us that this or that object is constituted insuch and such a manner, but not that it couldnot possibly exist otherwise. Now, in the firstplace, if we have a proposition which contains

the idea of necessity in its very conception, it isa if, moreover, it is not derived from any otherproposition, unless from one equally involvingthe idea of necessity, it is absolutely priori. Sec-ondly, an empirical judgement never exhibitsstrict and absolute, but only assumed and com-parative universality (by induction); therefore,the most we can say is—so far as we have hith-erto observed, there is no exception to this orthat rule. If, on the other hand, a judgementcarries with it strict and absolute universality,that is, admits of no possible exception, it is notderived from experience, but is valid absolutelya priori.

Empirical universality is, therefore, only anarbitrary extension of validity, from that whichmay be predicated of a proposition valid inmost cases, to that which is asserted of a propo-sition which holds good in all; as, for example,in the affirmation, "All bodies are heavy."When, on the contrary, strict universality char-

acterizes a judgement, it necessarily indicatesanother peculiar source of knowledge, namely,a faculty of cognition a priori. Necessity andstrict universality, therefore, are infallible testsfor distinguishing pure from empirical knowl-edge, and are inseparably connected with eachother. But as in the use of these criteria the em-pirical limitation is sometimes more easily de-tected than the contingency of the judgement,or the unlimited universality which we attachto a judgement is often a more convincingproof than its necessity, it may be advisable touse the criteria separately, each being by itselfinfallible.

Now, that in the sphere of human cognition wehave judgements which are necessary, and inthe strictest sense universal, consequently purea priori, it will be an easy matter to show. If wedesire an example from the sciences, we needonly take any proposition in mathematics. If wecast our eyes upon the commonest operations

of the understanding, the proposition, "Everychange must have a cause," will amply serveour purpose. In the latter case, indeed, the con-ception of a cause so plainly involves the con-ception of a necessity of connection with aneffect, and of a strict universality of the law,that the very notion of a cause would entirelydisappear, were we to derive it, like Hume,from a frequent association of what happenswith that which precedes; and the habit thenceoriginating of connecting representations—thenecessity inherent in the judgement beingtherefore merely subjective. Besides, withoutseeking for such examples of principles existinga priori in cognition, we might easily show thatsuch principles are the indispensable basis ofthe possibility of experience itself, and conse-quently prove their existence a priori. Forwhence could our experience itself acquire cer-tainty, if all the rules on which it depends werethemselves empirical, and consequently fortui-tous? No one, therefore, can admit the validity

of the use of such rules as first principles. But,for the present, we may content ourselves withhaving established the fact, that we do possessand exercise a faculty of pure a priori cognition;and, secondly, with having pointed out theproper tests of such cognition, namely, univer-sality and necessity.

Not only in judgements, however, but even inconceptions, is an a priori origin manifest. Forexample, if we take away by degrees from ourconceptions of a body all that can be referred tomere sensuous experience—colour, hardness orsoftness, weight, even impenetrability—thebody will then vanish; but the space which itoccupied still remains, and this it is utterly im-possible to annihilate in thought. Again, if wetake away, in like manner, from our empiricalconception of any object, corporeal or incorpo-real, all properties which mere experience hastaught us to connect with it, still we cannotthink away those through which we cogitate it

as substance, or adhering to substance, al-though our conception of substance is moredetermined than that of an object. Compelled,therefore, by that necessity with which the con-ception of substance forces itself upon us, wemust confess that it has its seat in our faculty ofcognition a priori.

III. Philosophy stands in need of a Sciencewhich shall Determine the Possibility, Principles, andExtent of Human Knowledge "a priori"

Of far more importance than all that has beenabove said, is the consideration that certain ofour cognitions rise completely above the sphereof all possible experience, and by means of con-ceptions, to which there exists in the wholeextent of experience no corresponding object,

seem to extend the range of our judgementsbeyond its bounds. And just in this transcen-dental or supersensible sphere, where experi-ence affords us neither instruction nor guid-ance, lie the investigations of reason, which, onaccount of their importance, we consider farpreferable to, and as having a far more elevatedaim than, all that the understanding canachieve within the sphere of sensuous phe-nomena. So high a value do we set upon theseinvestigations, that even at the risk of error, wepersist in following them out, and permit nei-ther doubt nor disregard nor indifference torestrain us from the pursuit. These unavoidableproblems of mere pure reason are God, free-dom (of will), and immortality. The sciencewhich, with all its preliminaries, has for its es-pecial object the solution of these problems isnamed metaphysics—a science which is at thevery outset dogmatical, that is, it confidentlytakes upon itself the execution of this taskwithout any previous investigation of the abil-

ity or inability of reason for such an undertak-ing.

Now the safe ground of experience being thusabandoned, it seems nevertheless natural thatwe should hesitate to erect a building with thecognitions we possess, without knowing when-ce they come, and on the strength of principles,the origin of which is undiscovered. Instead ofthus trying to build without a foundation, it israther to be expected that we should long agohave put the question, how the understandingcan arrive at these a priori cognitions, and whatis the extent, validity, and worth which theymay possess? We say, "This is natural enough,"meaning by the word natural, that which isconsistent with a just and reasonable way ofthinking; but if we understand by the term, thatwhich usually happens, nothing indeed couldbe more natural and more comprehensible thanthat this investigation should be left long unat-tempted. For one part of our pure knowledge,

the science of mathematics, has been longfirmly established, and thus leads us to formflattering expectations with regard to others,though these may be of quite a different nature.Besides, when we get beyond the bounds ofexperience, we are of course safe from opposi-tion in that quarter; and the charm of wideningthe range of our knowledge is so great that,unless we are brought to a standstill by someevident contradiction, we hurry on undoubt-ingly in our course. This, however, may beavoided, if we are sufficiently cautious in theconstruction of our fictions, which are not theless fictions on that account.

Mathematical science affords us a brilliant ex-ample, how far, independently of all experi-ence, we may carry our a priori knowledge. It istrue that the mathematician occupies himselfwith objects and cognitions only in so far asthey can be represented by means of intuition.But this circumstance is easily overlooked, be-

cause the said intuition can itself be given apriori, and therefore is hardly to be distin-guished from a mere pure conception. De-ceived by such a proof of the power of reason,we can perceive no limits to the extension ofour knowledge. The light dove cleaving in freeflight the thin air, whose resistance it feels,might imagine that her movements would befar more free and rapid in airless space. Just inthe same way did Plato, abandoning the worldof sense because of the narrow limits it sets tothe understanding, venture upon the wings ofideas beyond it, into the void space of pureintellect. He did not reflect that he made no realprogress by all his efforts; for he met with noresistance which might serve him for a support,as it were, whereon to rest, and on which hemight apply his powers, in order to let the in-tellect acquire momentum for its progress. It is,indeed, the common fate of human reason inspeculation, to finish the imposing edifice ofthought as rapidly as possible, and then for the

first time to begin to examine whether thefoundation is a solid one or no. Arrived at thispoint, all sorts of excuses are sought after, inorder to console us for its want of stability, orrather, indeed, to enable Us to dispense alto-gether with so late and dangerous an investiga-tion. But what frees us during the process ofbuilding from all apprehension or suspicion,and flatters us into the belief of its solidity, isthis. A great part, perhaps the greatest part, ofthe business of our reason consists in the ana-lysation of the conceptions which we alreadypossess of objects. By this means we gain amultitude of cognitions, which although reallynothing more than elucidations or explanationsof that which (though in a confused manner)was already thought in our conceptions, are, atleast in respect of their form, prized as newintrospections; whilst, so far as regards theirmatter or content, we have really made no ad-dition to our conceptions, but only disinvolvedthem. But as this process does furnish a real

priori knowledge, which has a sure progressand useful results, reason, deceived by this,slips in, without being itself aware of it, asser-tions of a quite different kind; in which, togiven conceptions it adds others, a priori in-deed, but entirely foreign to them, without ourknowing how it arrives at these, and, indeed,without such a question ever suggesting itself. Ishall therefore at once proceed to examine thedifference between these two modes of knowl-edge.

IV. Of the Difference Between Analytical andSynthetical Judgements.

In all judgements wherein the relation of a sub-ject to the predicate is cogitated (I mention af-firmative judgements only here; the applicationto negative will be very easy), this relation ispossible in two different ways. Either the predi-

cate B belongs to the subject A, as somewhatwhich is contained (though covertly) in theconception A; or the predicate B lies completelyout of the conception A, although it stands inconnection with it. In the first instance, I termthe judgement analytical, in the second, syn-thetical. Analytical judgements (affirmative) aretherefore those in which the connection of thepredicate with the subject is cogitated throughidentity; those in which this connection is cogi-tated without identity, are called syntheticaljudgements. The former may be called explica-tive, the latter augmentative judgements; be-cause the former add in the predicate nothingto the conception of the subject, but only ana-lyse it into its constituent conceptions, whichwere thought already in the subject, althoughin a confused manner; the latter add to our con-ceptions of the subject a predicate which wasnot contained in it, and which no analysis couldever have discovered therein. For example,when I say, "All bodies are extended," this is an

analytical judgement. For I need not go beyondthe conception of body in order to find exten-sion connected with it, but merely analyse theconception, that is, become conscious of themanifold properties which I think in that con-ception, in order to discover this predicate in it:it is therefore an analytical judgement. On theother hand, when I say, "All bodies are heavy,"the predicate is something totally differentfrom that which I think in the mere conceptionof a body. By the addition of such a predicate,therefore, it becomes a synthetical judgement.

Judgements of experience, as such, are alwayssynthetical. For it would be absurd to think ofgrounding an analytical judgement on experi-ence, because in forming such a judgement Ineed not go out of the sphere of my concep-tions, and therefore recourse to the testimonyof experience is quite unnecessary. That "bodiesare extended" is not an empirical judgement,but a proposition which stands firm a priori.

For before addressing myself to experience, Ialready have in my conception all the requisiteconditions for the judgement, and I have onlyto extract the predicate from the conception,according to the principle of contradiction, andthereby at the same time become conscious ofthe necessity of the judgement, a necessitywhich I could never learn from experience. Onthe other hand, though at first I do not at allinclude the predicate of weight in my concep-tion of body in general, that conception stillindicates an object of experience, a part of thetotality of experience, to which I can still addother parts; and this I do when I recognize byobservation that bodies are heavy. I can cognizebeforehand by analysis the conception of bodythrough the characteristics of extension, im-penetrability, shape, etc., all which are cogi-tated in this conception. But now I extend myknowledge, and looking back on experiencefrom which I had derived this conception ofbody, I find weight at all times connected with

the above characteristics, and therefore I syn-thetically add to my conceptions this as apredicate, and say, "All bodies are heavy." Thusit is experience upon which rests the possibilityof the synthesis of the predicate of weight withthe conception of body, because both concep-tions, although the one is not contained in theother, still belong to one another (only contin-gently, however), as parts of a whole, namely,of experience, which is itself a synthesis of in-tuitions.

But to synthetical judgements a priori, such aidis entirely wanting. If I go out of and beyondthe conception A, in order to recognize anotherB as connected with it, what foundation have Ito rest on, whereby to render the synthesis pos-sible? I have here no longer the advantage oflooking out in the sphere of experience forwhat I want. Let us take, for example, theproposition, "Everything that happens has acause." In the conception of "something that

happens," I indeed think an existence which acertain time antecedes, and from this I can de-rive analytical judgements. But the conceptionof a cause lies quite out of the above concep-tion, and indicates something entirely differentfrom "that which happens," and is consequentlynot contained in that conception. How then amI able to assert concerning the general concep-tion—"that which happens"—something en-tirely different from that conception, and torecognize the conception of cause although notcontained in it, yet as belonging to it, and evennecessarily? what is here the unknown = X,upon which the understanding rests when itbelieves it has found, out of the conception A aforeign predicate B, which it nevertheless con-siders to be connected with it? It cannot be ex-perience, because the principle adduced an-nexes the two representations, cause and effect,to the representation existence, not only withuniversality, which experience cannot give, butalso with the expression of necessity, therefore

completely a priori and from pure conceptions.Upon such synthetical, that is augmentativepropositions, depends the whole aim of ourspeculative knowledge a priori; for althoughanalytical judgements are indeed highly impor-tant and necessary, they are so, only to arrive atthat clearness of conceptions which is requisitefor a sure and extended synthesis, and thisalone is a real acquisition.

V. In all Theoretical Sciences of Reason, Syn-thetical Judgements "a priori" are contained asPrinciples.

1. Mathematical judgements are always syn-thetical. Hitherto this fact, though incontestablytrue and very important in its consequences,seems to have escaped the analysts of the hu-man mind, nay, to be in complete opposition toall their conjectures. For as it was found that

mathematical conclusions all proceed accord-ing to the principle of contradiction (which thenature of every apodeictic certainty requires),people became persuaded that the fundamentalprinciples of the science also were recognizedand admitted in the same way. But the notion isfallacious; for although a synthetical proposi-tion can certainly be discerned by means of theprinciple of contradiction, this is possible onlywhen another synthetical proposition precedes,from which the latter is deduced, but never ofitself.

Before all, be it observed, that proper mathe-matical propositions are always judgements apriori, and not empirical, because they carryalong with them the conception of necessity,which cannot be given by experience. If this bedemurred to, it matters not; I will then limit myassertion to pure mathematics, the very concep-tion of which implies that it consists of knowl-edge altogether non-empirical and a priori.

We might, indeed at first suppose that the pro-position 7 + 5 = 12 is a merely analytical propo-sition, following (according to the principle ofcontradiction) from the conception of a sum ofseven and five. But if we regard it more nar-rowly, we find that our conception of the sumof seven and five contains nothing more thanthe uniting of both sums into one, whereby itcannot at all be cogitated what this single num-ber is which embraces both. The conception oftwelve is by no means obtained by merely cogi-tating the union of seven and five; and we mayanalyse our conception of such a possible sumas long as we will, still we shall never discoverin it the notion of twelve. We must go beyondthese conceptions, and have recourse to an in-tuition which corresponds to one of the two—our five fingers, for example, or like Segner inhis Arithmetic five points, and so by degrees,add the units contained in the five given in theintuition, to the conception of seven. For I firsttake the number 7, and, for the conception of 5

calling in the aid of the fingers of my hand asobjects of intuition, I add the units, which I be-fore took together to make up the number 5,gradually now by means of the material imagemy hand, to the number 7, and by this process,I at length see the number 12 arise. That 7should be added to 5, I have certainly cogitatedin my conception of a sum = 7 + 5, but not thatthis sum was equal to 12. Arithmetical proposi-tions are therefore always synthetical, of whichwe may become more clearly convinced bytrying large numbers. For it will thus becomequite evident that, turn and twist our concep-tions as we may, it is impossible, without hav-ing recourse to intuition, to arrive at the sumtotal or product by means of the mere analysisof our conceptions. Just as little is any principleof pure geometry analytical. "A straight linebetween two points is the shortest," is a syn-thetical proposition. For my conception ofstraight contains no notion of quantity, but ismerely qualitative. The conception of the short-

est is therefore fore wholly an addition, and byno analysis can it be extracted from our concep-tion of a straight line. Intuition must thereforehere lend its aid, by means of which, and thusonly, our synthesis is possible.

Some few principles preposited by geometri-cians are, indeed, really analytical, and dependon the principle of contradiction. They serve,however, like identical propositions, as links inthe chain of method, not as principles—for ex-ample, a = a, the whole is equal to itself, or(a+b) > a, the whole is greater than its part.And yet even these principles themselves,though they derive their validity from pureconceptions, are only admitted in mathematicsbecause they can be presented in intuition.What causes us here commonly to believe thatthe predicate of such apodeictic judgements isalready contained in our conception, and thatthe judgement is therefore analytical, is merelythe equivocal nature of the expression. We

must join in thought a certain predicate to agiven conception, and this necessity cleavesalready to the conception. But the question is,not what we must join in thought to the givenconception, but what we really think therein,though only obscurely, and then it becomesmanifest that the predicate pertains to theseconceptions, necessarily indeed, yet not asthought in the conception itself, but by virtue ofan intuition, which must be added to the con-ception.

2. The science of natural philosophy (physics)contains in itself synthetical judgements a pri-ori, as principles. I shall adduce two proposi-tions. For instance, the proposition, "In allchanges of the material world, the quantity ofmatter remains unchanged"; or, that, "In allcommunication of motion, action and reactionmust always be equal." In both of these, notonly is the necessity, and therefore their origina priori clear, but also that they are synthetical

propositions. For in the conception of matter, Ido not cogitate its permanency, but merely itspresence in space, which it fills. I thereforereally go out of and beyond the conception ofmatter, in order to think on to it something apriori, which I did not think in it. The proposi-tion is therefore not analytical, but synthetical,and nevertheless conceived a priori; and so it iswith regard to the other propositions of thepure part of natural philosophy.

3. As to metaphysics, even if we look upon itmerely as an attempted science, yet, from thenature of human reason, an indispensable one,we find that it must contain synthetical propo-sitions a priori. It is not merely the duty of me-taphysics to dissect, and thereby analytically toillustrate the conceptions which we form a pri-ori of things; but we seek to widen the range ofour a priori knowledge. For this purpose, wemust avail ourselves of such principles as addsomething to the original conception—

something not identical with, nor contained init, and by means of synthetical judgements apriori, leave far behind us the limits of experi-ence; for example, in the proposition, "theworld must have a beginning," and such like.Thus metaphysics, according to the proper aimof the science, consists merely of syntheticalpropositions a priori.

VI. The Universal Problem of Pure Reason.

It is extremely advantageous to be able to bringa number of investigations under the formulaof a single problem. For in this manner, we notonly facilitate our own labour, inasmuch as wedefine it clearly to ourselves, but also render itmore easy for others to decide whether wehave done justice to our undertaking. The pro-per problem of pure reason, then, is contained

in the question: "How are synthetical judge-ments a priori possible?"

That metaphysical science has hitherto re-mained in so vacillating a state of uncertaintyand contradiction, is only to be attributed to thefact that this great problem, and perhaps eventhe difference between analytical and syntheti-cal judgements, did not sooner suggest itself tophilosophers. Upon the solution of this prob-lem, or upon sufficient proof of the impossibil-ity of synthetical knowledge a priori, dependsthe existence or downfall of the science of me-taphysics. Among philosophers, David Humecame the nearest of all to this problem; yet itnever acquired in his mind sufficient precision,nor did he regard the question in its universal-ity. On the contrary, he stopped short at thesynthetical proposition of the connection of aneffect with its cause (principium causalitatis),insisting that such proposition a priori was im-possible. According to his conclusions, then, all

that we term metaphysical science is a meredelusion, arising from the fancied insight ofreason into that which is in truth borrowedfrom experience, and to which habit has giventhe appearance of necessity. Against this asser-tion, destructive to all pure philosophy, hewould have been guarded, had he had ourproblem before his eyes in its universality. Forhe would then have perceived that, accordingto his own argument, there likewise could notbe any pure mathematical science, which as-suredly cannot exist without synthetical propo-sitions a priori—an absurdity from which hisgood understanding must have saved him.

In the solution of the above problem is at thesame time comprehended the possibility of theuse of pure reason in the foundation and cons-truction of all sciences which contain theoreti-cal knowledge a priori of objects, that is to say,the answer to the following questions:

How is pure mathematical science possible?

How is pure natural science possible?

Respecting these sciences, as they do certainlyexist, it may with propriety be asked, how theyare possible?—for that they must be possible isshown by the fact of their really existing.* Butas to metaphysics, the miserable progress it hashitherto made, and the fact that of no one sys-tem yet brought forward, far as regards its trueaim, can it be said that this science really exists,leaves any one at liberty to doubt with reasonthe very possibility of its existence.

[*Footnote: As to the existence of pure naturalscience, or physics, perhaps many may stillexpress doubts. But we have only to look at thedifferent propositions which are commonlytreated of at the commencement of proper (em-pirical) physical science—those, for example,relating to the permanence of the same quantityof matter, the vis inertiae, the equality of actionand reaction, etc.—to be soon convinced thatthey form a science of pure physics (physica

pura, or rationalis), which well deserves to beseparately exposed as a special science, in itswhole extent, whether that be great or confi-ned.]

Yet, in a certain sense, this kind of knowledgemust unquestionably be looked upon as given;in other words, metaphysics must be consid-ered as really existing, if not as a science, never-theless as a natural disposition of the humanmind (metaphysica naturalis). For human rea-son, without any instigations imputable to themere vanity of great knowledge, unceasinglyprogresses, urged on by its own feeling of need,towards such questions as cannot be answeredby any empirical application of reason, or prin-ciples derived therefrom; and so there has everreally existed in every man some system of me-taphysics. It will always exist, so soon as reasonawakes to the exercise of its power of specula-tion. And now the question arises: "How is me-taphysics, as a natural disposition, possible?" In

other words, how, from the nature of universalhuman reason, do those questions arise whichpure reason proposes to itself, and which it isimpelled by its own feeling of need to answeras well as it can?

But as in all the attempts hitherto made to ans-wer the questions which reason is prompted byits very nature to propose to itself, for example,whether the world had a beginning, or has exis-ted from eternity, it has always met with una-voidable contradictions, we must not rest satis-fied with the mere natural disposition of themind to metaphysics, that is, with the existenceof the faculty of pure reason, whence, indeed,some sort of metaphysical system always ari-ses; but it must be possible to arrive at certaintyin regard to the question whether we know ordo not know the things of which metaphysicstreats. We must be able to arrive at a decisionon the subjects of its questions, or on the abilityor inability of reason to form any judgement

respecting them; and therefore either to extendwith confidence the bounds of our pure reason,or to set strictly defined and safe limits to itsaction. This last question, which arises out ofthe above universal problem, would properlyrun thus: "How is metaphysics possible as ascience?"

Thus, the critique of reason leads at last, natu-rally and necessarily, to science; and, on theother hand, the dogmatical use of reason wit-hout criticism leads to groundless assertions,against which others equally specious can al-ways be set, thus ending unavoidably in scepti-cism.

Besides, this science cannot be of great andformidable prolixity, because it has not to dowith objects of reason, the variety of which isinexhaustible, but merely with Reason herselfand her problems; problems which arise out ofher own bosom, and are not proposed to her bythe nature of outward things, but by her own

nature. And when once Reason has previouslybecome able completely to understand her ownpower in regard to objects which she meetswith in experience, it will be easy to determinesecurely the extent and limits of her attemptedapplication to objects beyond the confines ofexperience.

We may and must, therefore, regard the at-tempts hitherto made to establish metaphysicalscience dogmatically as non-existent. For whatof analysis, that is, mere dissection of concep-tions, is contained in one or other, is not theaim of, but only a preparation for metaphysicsproper, which has for its object the extension,by means of synthesis, of our a priori knowl-edge. And for this purpose, mere analysis is ofcourse useless, because it only shows what iscontained in these conceptions, but not how wearrive, a priori, at them; and this it is her dutyto show, in order to be able afterwards to de-termine their valid use in regard to all objects of

experience, to all knowledge in general. Butlittle self-denial, indeed, is needed to give upthese pretensions, seeing the undeniable, and inthe dogmatic mode of procedure, inevitablecontradictions of Reason with herself, havelong since ruined the reputation of every sys-tem of metaphysics that has appeared up to thistime. It will require more firmness to remainundeterred by difficulty from within, and op-position from without, from endeavouring, bya method quite opposed to all those hithertofollowed, to further the growth and fruitfulnessof a science indispensable to human reason—ascience from which every branch it has bornemay be cut away, but whose roots remain inde-structible.

VII. Idea and Division of a Particular Science,under the Name of a Critique of Pure Reason.

From all that has been said, there results theidea of a particular science, which may be ca-lled the Critique of Pure Reason. For reason isthe faculty which furnishes us with the princi-ples of knowledge a priori. Hence, pure reasonis the faculty which contains the principles ofcognizing anything absolutely a priori. An or-ganon of pure reason would be a compendiumof those principles according to which alone allpure cognitions a priori can be obtained. Thecompletely extended application of such anorganon would afford us a system of pure rea-son. As this, however, is demanding a greatdeal, and it is yet doubtful whether any exten-sion of our knowledge be here possible, or, ifso, in what cases; we can regard a science of themere criticism of pure reason, its sources andlimits, as the propaedeutic to a system of purereason. Such a science must not be called a doc-trine, but only a critique of pure reason; and itsuse, in regard to speculation, would be onlynegative, not to enlarge the bounds of, but to

purify, our reason, and to shield it against er-ror—which alone is no little gain. I apply theterm transcendental to all knowledge which isnot so much occupied with objects as with themode of our cognition of these objects, so far asthis mode of cognition is possible a priori. Asystem of such conceptions would be calledtranscendental philosophy. But this, again, isstill beyond the bounds of our present essay.For as such a science must contain a completeexposition not only of our synthetical a priori,but of our analytical a priori knowledge, it is oftoo wide a range for our present purpose, be-cause we do not require to carry our analysisany farther than is necessary to understand, intheir full extent, the principles of synthesis apriori, with which alone we have to do. Thisinvestigation, which we cannot properly call adoctrine, but only a transcendental critique,because it aims not at the enlargement, but atthe correction and guidance, of our knowledge,and is to serve as a touchstone of the worth or

worthlessness of all knowledge a priori, is thesole object of our present essay. Such a critiqueis consequently, as far as possible, a prepara-tion for an organon; and if this new organonshould be found to fail, at least for a canon ofpure reason, according to which the completesystem of the philosophy of pure reason, whet-her it extend or limit the bounds of that reason,might one day be set forth both analytically andsynthetically. For that this is possible, nay, thatsuch a system is not of so great extent as to pre-clude the hope of its ever being completed, isevident. For we have not here to do with thenature of outward objects, which is infinite, butsolely with the mind, which judges of the na-ture of objects, and, again, with the mind onlyin respect of its cognition a priori. And the ob-ject of our investigations, as it is not to besought without, but, altogether within, our-selves, cannot remain concealed, and in allprobability is limited enough to be completelysurveyed and fairly estimated, according to its

worth or worthlessness. Still less let the readerhere expect a critique of books and systems ofpure reason; our present object is exclusively acritique of the faculty of pure reason itself.Only when we make this critique our founda-tion, do we possess a pure touchstone for esti-mating the philosophical value of ancient andmodern writings on this subject; and withoutthis criterion, the incompetent historian orjudge decides upon and corrects the groundlessassertions of others with his own, which havethemselves just as little foundation.

Transcendental philosophy is the idea of ascience, for which the Critique of Pure Reasonmust sketch the whole plan architectonically,that is, from principles, with a full guaranteefor the validity and stability of all the partswhich enter into the building. It is the system ofall the principles of pure reason. If this Critiqueitself does not assume the title of transcenden-tal philosophy, it is only because, to be a com-

plete system, it ought to contain a full analysisof all human knowledge a priori. Our critiquemust, indeed, lay before us a complete enu-meration of all the radical conceptions whichconstitute the said pure knowledge. But fromthe complete analysis of these conceptionsthemselves, as also from a complete investiga-tion of those derived from them, it abstainswith reason; partly because it would be deviat-ing from the end in view to occupy itself withthis analysis, since this process is not attendedwith the difficulty and insecurity to be found inthe synthesis, to which our critique is entirelydevoted, and partly because it would be incon-sistent with the unity of our plan to burden thisessay with the vindication of the completenessof such an analysis and deduction, with which,after all, we have at present nothing to do. Thiscompleteness of the analysis of these radicalconceptions, as well as of the deduction fromthe conceptions a priori which may be given bythe analysis, we can, however, easily attain,

provided only that we are in possession of allthese radical conceptions, which are to serve asprinciples of the synthesis, and that in respectof this main purpose nothing is wanting.

To the Critique of Pure Reason, therefore, be-longs all that constitutes transcendental philo-sophy; and it is the complete idea of transcen-dental philosophy, but still not the science it-self; because it only proceeds so far with theanalysis as is necessary to the power of judgingcompletely of our synthetical knowledge apriori.

The principal thing we must attend to, in thedivision of the parts of a science like this, is thatno conceptions must enter it which containaught empirical; in other words, that the know-ledge a priori must be completely pure. Hence,although the highest principles and fundamen-tal conceptions of morality are certainly cogni-tions a priori, yet they do not belong to tran-scendental philosophy; because, though they

certainly do not lay the conceptions of pain,pleasure, desires, inclinations, etc. (which areall of empirical origin), at the foundation of itsprecepts, yet still into the conception of duty—as an obstacle to be overcome, or as an incite-ment which should not be made into a mo-tive—these empirical conceptions must neces-sarily enter, in the construction of a system ofpure morality. Transcendental philosophy isconsequently a philosophy of the pure and me-rely speculative reason. For all that is practical,so far as it contains motives, relates to feelings,and these belong to empirical sources of cogni-tion.

If we wish to divide this science from the uni-versal point of view of a science in general, itought to comprehend, first, a Doctrine of theElements, and, secondly, a Doctrine of the Met-hod of pure reason. Each of these main divi-sions will have its subdivisions, the separatereasons for which we cannot here particularize.

Only so much seems necessary, by way of in-troduction of premonition, that there are twosources of human knowledge (which probablyspring from a common, but to us unknownroot), namely, sense and understanding. By theformer, objects are given to us; by the latter,thought. So far as the faculty of sense may con-tain representations a priori, which form theconditions under which objects are given, in sofar it belongs to transcendental philosophy. Thetranscendental doctrine of sense must form thefirst part of our science of elements, because theconditions under which alone the objects ofhuman knowledge are given must precede tho-se under which they are thought.

I. TRANSCENDENTAL DOCTRINE OFELEMENTS.

FIRST PART. TRANSCENDENTAL AEST-HETIC.

SS I. Introductory.

In whatsoever mode, or by whatsoever means,our knowledge may relate to objects, it is atleast quite clear that the only manner in whichit immediately relates to them is by means of anintuition. To this as the indispensable ground-work, all thought points. But an intuition cantake place only in so far as the object is given tous. This, again, is only possible, to man at least,on condition that the object affect the mind in acertain manner. The capacity for receiving re-presentations (receptivity) through the mode inwhich we are affected by objects, objects, iscalled sensibility. By means of sensibility, there-

fore, objects are given to us, and it alone furnis-hes us with intuitions; by the understandingthey are thought, and from it arise conceptions.But an thought must directly, or indirectly, bymeans of certain signs, relate ultimately to in-tuitions; consequently, with us, to sensibility,because in no other way can an object be givento us.

The effect of an object upon the faculty of re-presentation, so far as we are affected by thesaid object, is sensation. That sort of intuitionwhich relates to an object by means of sensationis called an empirical intuition. The undetermi-ned object of an empirical intuition is calledphenomenon. That which in the phenomenoncorresponds to the sensation, I term its matter;but that which effects that the content of thephenomenon can be arranged under certainrelations, I call its form. But that in which oursensations are merely arranged, and by whichthey are susceptible of assuming a certain form,

cannot be itself sensation. It is, then, the matterof all phenomena that is given to us a posterio-ri; the form must lie ready a priori for them inthe mind, and consequently can be regardedseparately from all sensation.

I call all representations pure, in the transcen-dental meaning of the word, wherein nothing ismet with that belongs to sensation. And accord-ingly we find existing in the mind a priori, thepure form of sensuous intuitions in general, inwhich all the manifold content of the phe-nomenal world is arranged and viewed undercertain relations. This pure form of sensibility Ishall call pure intuition. Thus, if I take awayfrom our representation of a body all that theunderstanding thinks as belonging to it, as sub-stance, force, divisibility, etc., and also what-ever belongs to sensation, as impenetrability,hardness, colour, etc.; yet there is still some-thing left us from this empirical intuition,namely, extension and shape. These belong to

pure intuition, which exists a priori in themind, as a mere form of sensibility, and with-out any real object of the senses or any sensa-tion.

The science of all the principles of sensibility apriori, I call transcendental aesthetic.* Theremust, then, be such a science forming the firstpart of the transcendental doctrine of elements,in contradistinction to that part which containsthe principles of pure thought, and which iscalled transcendental logic.

[Footnote: The Germans are the only peoplewho at present use this word to indicate whatothers call the critique of taste. At the founda-tion of this term lies the disappointed hope,which the eminent analyst, Baumgarten, con-ceived, of subjecting the criticism of the beauti-ful to principles of reason, and so of elevatingits rules into a science. But his endeavours werevain. For the said rules or criteria are, in respectto their chief sources, merely empirical, conse-

quently never can serve as determinate laws apriori, by which our judgement in matters oftaste is to be directed. It is rather our judgementwhich forms the proper test as to the correct-ness of the principles. On this account it is ad-visable to give up the use of the term as desig-nating the critique of taste, and to apply it sole-ly to that doctrine, which is true science—thescience of the laws of sensibility—and thus co-me nearer to the language and the sense of theancients in their well-known division of theobjects of cognition into aiotheta kai noeta, orto share it with speculative philosophy, andemploy it partly in a transcendental, partly in apsychological signification.]

In the science of transcendental aesthetic ac-cordingly, we shall first isolate sensibility or thesensuous faculty, by separating from it all thatis annexed to its perceptions by the conceptionsof understanding, so that nothing be left butempirical intuition. In the next place we shall

take away from this intuition all that belongs tosensation, so that nothing may remain but pureintuition, and the mere form of phenomena,which is all that the sensibility can afford a pri-ori. From this investigation it will be found thatthere are two pure forms of sensuous intuition,as principles of knowledge a priori, namely,space and time. To the consideration of thesewe shall now proceed.

SECTION I. Of Space.

SS 2. Metaphysical Exposition of this Concep-tion.

By means of the external sense (a property ofthe mind), we represent to ourselves objects aswithout us, and these all in space. Herein aloneare their shape, dimensions, and relations toeach other determined or determinable. The

internal sense, by means of which the mindcontemplates itself or its internal state, gives,indeed, no intuition of the soul as an object; yetthere is nevertheless a determinate form, underwhich alone the contemplation of our internalstate is possible, so that all which relates to theinward determinations of the mind is repre-sented in relations of time. Of time we cannothave any external intuition, any more than wecan have an internal intuition of space. Whatthen are time and space? Are they real exis-tences? Or, are they merely relations or deter-minations of things, such, however, as wouldequally belong to these things in themselves,though they should never become objects ofintuition; or, are they such as belong only to theform of intuition, and consequently to the sub-jective constitution of the mind, without whichthese predicates of time and space could not beattached to any object? In order to become in-formed on these points, we shall first give anexposition of the conception of space. By expo-

sition, I mean the clear, though not detailed,representation of that which belongs to a con-ception; and an exposition is metaphysicalwhen it contains that which represents the con-ception as given a priori.

1. Space is not a conception which has beenderived from outward experiences. For, in or-der that certain sensations may relate to somet-hing without me (that is, to something whichoccupies a different part of space from that inwhich I am); in like manner, in order that I mayrepresent them not merely as without, of, andnear to each other, but also in separate places,the representation of space must already existas a foundation. Consequently, the representa-tion of space cannot be borrowed from the rela-tions of external phenomena through experi-ence; but, on the contrary, this externalexperience is itself only possible through thesaid antecedent representation.

2. Space then is a necessary representation apriori, which serves for the foundation of allexternal intuitions. We never can imagine ormake a representation to ourselves of the non-existence of space, though we may easilyenough think that no objects are found in it. Itmust, therefore, be considered as the conditionof the possibility of phenomena, and by nomeans as a determination dependent on them,and is a representation a priori, which necessar-ily supplies the basis for external phenomena.

3. Space is no discursive, or as we say, generalconception of the relations of things, but a pureintuition. For, in the first place, we can onlyrepresent to ourselves one space, and, when wetalk of divers spaces, we mean only parts of oneand the same space. Moreover, these parts can-not antecede this one all-embracing space, asthe component parts from which the aggregatecan be made up, but can be cogitated only asexisting in it. Space is essentially one, and mul-

tiplicity in it, consequently the general notionof spaces, of this or that space, depends solelyupon limitations. Hence it follows that an apriori intuition (which is not empirical) lies atthe root of all our conceptions of space. Thus,moreover, the principles of geometry—for ex-ample, that "in a triangle, two sides together aregreater than the third," are never deduced fromgeneral conceptions of line and triangle, butfrom intuition, and this a priori, with apodeicticcertainty.

4. Space is represented as an infinite givenquantity. Now every conception must indeedbe considered as a representation which is con-tained in an infinite multitude of different pos-sible representations, which, therefore, com-prises these under itself; but no conception, assuch, can be so conceived, as if it containedwithin itself an infinite multitude of representa-tions. Nevertheless, space is so conceived of, forall parts of space are equally capable of being

produced to infinity. Consequently, the originalrepresentation of space is an intuition a priori,and not a conception.

SS 3. Transcendental Exposition of the Concep-tion of Space.

By a transcendental exposition, I mean the ex-planation of a conception, as a principle, when-ce can be discerned the possibility of other syn-thetical a priori cognitions. For this purpose, itis requisite, firstly, that such cognitions doreally flow from the given conception; and,secondly, that the said cognitions are only pos-sible under the presupposition of a given modeof explaining this conception.

Geometry is a science which determines theproperties of space synthetically, and yet apriori. What, then, must be our representationof space, in order that such a cognition of it

may be possible? It must be originally intuition,for from a mere conception, no propositionscan be deduced which go out beyond the con-ception, and yet this happens in geometry. (In-trod. V.) But this intuition must be found in themind a priori, that is, before any perception ofobjects, consequently must be pure, not empiri-cal, intuition. For geometrical principles arealways apodeictic, that is, united with the con-sciousness of their necessity, as: "Space has onlythree dimensions." But propositions of this kindcannot be empirical judgements, nor conclu-sions from them. (Introd. II.) Now, how can anexternal intuition anterior to objects them-selves, and in which our conception of objectscan be determined a priori, exist in the humanmind? Obviously not otherwise than in so faras it has its seat in the subject only, as the for-mal capacity of the subject's being affected byobjects, and thereby of obtaining immediaterepresentation, that is, intuition; consequently,

only as the form of the external sense in gen-eral.

Thus it is only by means of our explanation thatthe possibility of geometry, as a syntheticalscience a priori, becomes comprehensible. Eve-ry mode of explanation which does not showus this possibility, although in appearance itmay be similar to ours, can with the utmostcertainty be distinguished from it by thesemarks.

SS 4. Conclusions from the foregoing Concep-tions.

(a) Space does not represent any property ofobjects as things in themselves, nor does it rep-resent them in their relations to each other; inother words, space does not represent to us anydetermination of objects such as attaches to theobjects themselves, and would remain, even

though all subjective conditions of the intuitionwere abstracted. For neither absolute nor rela-tive determinations of objects can be intuitedprior to the existence of the things to whichthey belong, and therefore not a priori.

(b) Space is nothing else than the form of allphenomena of the external sense, that is, thesubjective condition of the sensibility, underwhich alone external intuition is possible. Now,because the receptivity or capacity of the sub-ject to be affected by objects necessarily antece-des all intuitions of these objects, it is easilyunderstood how the form of all phenomena canbe given in the mind previous to all actual per-ceptions, therefore a priori, and how it, as apure intuition, in which all objects must be de-termined, can contain principles of the relationsof these objects prior to all experience.

It is therefore from the human point of viewonly that we can speak of space, extended ob-jects, etc. If we depart from the subjective con-

dition, under which alone we can obtain exter-nal intuition, or, in other words, by means ofwhich we are affected by objects, the represen-tation of space has no meaning whatsoever.This predicate is only applicable to things in sofar as they appear to us, that is, are objects ofsensibility. The constant form of this receptiv-ity, which we call sensibility, is a necessary con-dition of all relations in which objects can beintuited as existing without us, and when ab-straction of these objects is made, is a pure in-tuition, to which we give the name of space. Itis clear that we cannot make the special condi-tions of sensibility into conditions of the possi-bility of things, but only of the possibility oftheir existence as far as they are phenomena.And so we may correctly say that space con-tains all which can appear to us externally, butnot all things considered as things in them-selves, be they intuited or not, or by whatso-ever subject one will. As to the intuitions ofother thinking beings, we cannot judge whether

they are or are not bound by the same condi-tions which limit our own intuition, and whichfor us are universally valid. If we join the limi-tation of a judgement to the conception of thesubject, then the judgement will possess un-conditioned validity. For example, the proposi-tion, "All objects are beside each other inspace," is valid only under the limitation thatthese things are taken as objects of our sensu-ous intuition. But if I join the condition to theconception and say, "All things, as externalphenomena, are beside each other in space,"then the rule is valid universally, and withoutany limitation. Our expositions, consequently,teach the reality (i.e., the objective validity) ofspace in regard of all which can be presented tous externally as object, and at the same timealso the ideality of space in regard to objectswhen they are considered by means of reasonas things in themselves, that is, without refer-ence to the constitution of our sensibility. Wemaintain, therefore, the empirical reality of

space in regard to all possible external experi-ence, although we must admit its transcenden-tal ideality; in other words, that it is nothing, sosoon as we withdraw the condition upon whichthe possibility of all experience depends andlook upon space as something that belongs tothings in themselves.

But, with the exception of space, there is norepresentation, subjective and referring to so-mething external to us, which could be calledobjective a priori. For there are no other subjec-tive representations from which we can deducesynthetical propositions a priori, as we canfrom the intuition of space. (See SS 3.) Therefo-re, to speak accurately, no ideality whateverbelongs to these, although they agree in thisrespect with the representation of space, thatthey belong merely to the subjective nature ofthe mode of sensuous perception; such a mode,for example, as that of sight, of hearing, and offeeling, by means of the sensations of colour,

sound, and heat, but which, because they areonly sensations and not intuitions, do not ofthemselves give us the cognition of any object,least of all, an a priori cognition. My purpose,in the above remark, is merely this: to guardany one against illustrating the asserted idealityof space by examples quite insufficient, forexample, by colour, taste, etc.; for these must becontemplated not as properties of things, butonly as changes in the subject, changes whichmay be different in different men. For, in such acase, that which is originally a mere phenome-non, a rose, for example, is taken by the empiri-cal understanding for a thing in itself, thoughto every different eye, in respect of its colour, itmay appear different. On the contrary, thetranscendental conception of phenomena inspace is a critical admonition, that, in general,nothing which is intuited in space is a thing initself, and that space is not a form which be-longs as a property to things; but that objectsare quite unknown to us in themselves, and

what we call outward objects, are nothing elsebut mere representations of our sensibility,whose form is space, but whose real correlate,the thing in itself, is not known by means ofthese representations, nor ever can be, but re-specting which, in experience, no inquiry isever made.

SECTION II. Of Time.

SS 5. Metaphysical Exposition of this Concep-tion.

1. Time is not an empirical conception. For nei-ther coexistence nor succession would be per-ceived by us, if the representation of time didnot exist as a foundation a priori. Without thispresupposition we could not represent to our-

selves that things exist together at one and thesame time, or at different times, that is, con-temporaneously, or in succession.

2. Time is a necessary representation, lying atthe foundation of all our intuitions. With re-gard to phenomena in general, we cannot thinkaway time from them, and represent them toourselves as out of and unconnected with time,but we can quite well represent to ourselvestime void of phenomena. Time is therefore gi-ven a priori. In it alone is all reality of phenom-ena possible. These may all be annihilated inthought, but time itself, as the universal condi-tion of their possibility, cannot be so annulled.

3. On this necessity a priori is also founded thepossibility of apodeictic principles of the rela-tions of time, or axioms of time in general, suchas: "Time has only one dimension," "Differenttimes are not coexistent but successive" (as dif-ferent spaces are not successive but coexistent).These principles cannot be derived from ex-

perience, for it would give neither strict univer-sality, nor apodeictic certainty. We should onlybe able to say, "so common experience teachesus," but not "it must be so." They are valid asrules, through which, in general, experience ispossible; and they instruct us respecting ex-perience, and not by means of it.

4. Time is not a discursive, or as it is called,general conception, but a pure form of the sen-suous intuition. Different times are merelyparts of one and the same time. But the repre-sentation which can only be given by a singleobject is an intuition. Besides, the propositionthat different times cannot be coexistent couldnot be derived from a general conception. Forthis proposition is synthetical, and thereforecannot spring out of conceptions alone. It istherefore contained immediately in the intui-tion and representation of time.

5. The infinity of time signifies nothing morethan that every determined quantity of time is

possible only through limitations of one timelying at the foundation. Consequently, the ori-ginal representation, time, must be given asunlimited. But as the determinate representa-tion of the parts of time and of every quantityof an object can only be obtained by limitation,the complete representation of time must not befurnished by means of conceptions, for thesecontain only partial representations. Concep-tions, on the contrary, must have immediateintuition for their basis.

SS 6 Transcendental Exposition of the Concep-tion of Time.

I may here refer to what is said above (SS 5, 3),where, for or sake of brevity, I have placed un-der the head of metaphysical exposition, thatwhich is properly transcendental. Here I shalladd that the conception of change, and with it

the conception of motion, as change of place, ispossible only through and in the representationof time; that if this representation were not anintuition (internal) a priori, no conception, ofwhatever kind, could render comprehensiblethe possibility of change, in other words, of aconjunction of contradictorily opposed predi-cates in one and the same object, for example,the presence of a thing in a place and the non-presence of the same thing in the same place. Itis only in time that it is possible to meet withtwo contradictorily opposed determinations inone thing, that is, after each other. Thus ourconception of time explains the possibility of somuch synthetical knowledge a priori, as is ex-hibited in the general doctrine of motion, whichis not a little fruitful.

SS 7. Conclusions from the above Conceptions.

(a) Time is not something which subsists ofitself, or which inheres in things as an objectivedetermination, and therefore remains, whenabstraction is made of the subjective conditionsof the intuition of things. For in the former case,it would be something real, yet without pre-senting to any power of perception any realobject. In the latter case, as an order ordetermination inherent in things themselves, itcould not be antecedent to things, as theircondition, nor discerned or intuited by meansof synthetical propositions a priori. But all thisis quite possible when we regard time asmerely the subjective condition under which allour intuitions take place. For in that case, thisform of the inward intuition can be representedprior to the objects, and consequently a priori.

(b) Time is nothing else than the form of theinternal sense, that is, of the intuitions of selfand of our internal state. For time cannot beany determination of outward phenomena. It

has to do neither with shape nor position; onthe contrary, it determines the relation of repre-sentations in our internal state. And preciselybecause this internal intuition presents to us noshape or form, we endeavour to supply thiswant by analogies, and represent the course oftime by a line progressing to infinity, the con-tent of which constitutes a series which is onlyof one dimension; and we conclude from theproperties of this line as to all the properties oftime, with this single exception, that the partsof the line are coexistent, whilst those of timeare successive. From this it is clear also that therepresentation of time is itself an intuition, be-cause all its relations can be expressed in anexternal intuition.

(c) Time is the formal condition a priori of allphenomena whatsoever. Space, as the pureform of external intuition, is limited as a condi-tion a priori to external phenomena alone. Onthe other hand, because all representations,

whether they have or have not external thingsfor their objects, still in themselves, as determi-nations of the mind, belong to our internal sta-te; and because this internal state is subject tothe formal condition of the internal intuition,that is, to time—time is a condition a priori ofall phenomena whatsoever—the immediatecondition of all internal, and thereby the medi-ate condition of all external phenomena. If I cansay a priori, "All outward phenomena are inspace, and determined a priori according to therelations of space," I can also, from the principleof the internal sense, affirm universally, "Allphenomena in general, that is, all objects of thesenses, are in time and stand necessarily in rela-tions of time."

If we abstract our internal intuition of ourselvesand all external intuitions, possible only byvirtue of this internal intuition and presented tous by our faculty of representation, and conse-quently take objects as they are in themselves,

then time is nothing. It is only of objective vali-dity in regard to phenomena, because these arethings which we regard as objects of our senses.It no longer objective we, make abstraction ofthe sensuousness of our intuition, in otherwords, of that mode of representation which ispeculiar to us, and speak of things in general.Time is therefore merely a subjective conditionof our (human) intuition (which is always sen-suous, that is, so far as we are affected by ob-jects), and in itself, independently of the mindor subject, is nothing. Nevertheless, in respectof all phenomena, consequently of all thingswhich come within the sphere of our experi-ence, it is necessarily objective. We cannot say,"All things are in time," because in this concep-tion of things in general, we abstract and makeno mention of any sort of intuition of things.But this is the proper condition under whichtime belongs to our representation of objects. Ifwe add the condition to the conception, andsay, "All things, as phenomena, that is, objects

of sensuous intuition, are in time," then theproposition has its sound objective validity anduniversality a priori.

What we have now set forth teaches, therefore,the empirical reality of time; that is, its objectivevalidity in reference to all objects which canever be presented to our senses. And as ourintuition is always sensuous, no object ever canbe presented to us in experience, which doesnot come under the conditions of time. On theother hand, we deny to time all claim to abso-lute reality; that is, we deny that it, withouthaving regard to the form of our sensuous in-tuition, absolutely inheres in things as a condi-tion or property. Such properties as belong toobjects as things in themselves never can bepresented to us through the medium of thesenses. Herein consists, therefore, the transcen-dental ideality of time, according to which, ifwe abstract the subjective conditions of sensu-ous intuition, it is nothing, and cannot be reck-

oned as subsisting or inhering in objects asthings in themselves, independently of its rela-tion to our intuition. This ideality, like that ofspace, is not to be proved or illustrated by falla-cious analogies with sensations, for this rea-son—that in such arguments or illustrations,we make the presupposition that the phenome-non, in which such and such predicates inhere,has objective reality, while in this case we canonly find such an objective reality as is itselfempirical, that is, regards the object as a merephenomenon. In reference to this subject, seethe remark in Section I (SS 4)

SS 8. Elucidation.

Against this theory, which grants empiricalreality to time, but denies to it absolute andtranscendental reality, I have heard from inte-lligent men an objection so unanimously urged

that I conclude that it must naturally presentitself to every reader to whom these considera-tions are novel. It runs thus: "Changes are real"(this the continual change in our own represen-tations demonstrates, even though the existenceof all external phenomena, together with theirchanges, is denied). Now, changes are onlypossible in time, and therefore time must besomething real. But there is no difficulty in an-swering this. I grant the whole argument. Time,no doubt, is something real, that is, it is the realform of our internal intuition. It therefore hassubjective reality, in reference to our internalexperience, that is, I have really the representa-tion of time and of my determinations therein.Time, therefore, is not to be regarded as an ob-ject, but as the mode of representation of myselfas an object. But if I could intuite myself, or beintuited by another being, without this condi-tion of sensibility, then those very determina-tions which we now represent to ourselves aschanges, would present to us a knowledge in

which the representation of time, and conse-quently of change, would not appear. The em-pirical reality of time, therefore, remains, as thecondition of all our experience. But absolutereality, according to what has been said above,cannot be granted it. Time is nothing but theform of our internal intuition.* If we take awayfrom it the special condition of our sensibility,the conception of time also vanishes; and it in-heres not in the objects themselves, but solelyin the subject (or mind) which intuites them.

[*Footnote: I can indeed say "my representa-tions follow one another, or are successive"; butthis means only that we are conscious of themas in a succession, that is, according to the formof the internal sense. Time, therefore, is not athing in itself, nor is it any objective determina-tion pertaining to, or inherent in things.]

But the reason why this objection is so unani-mously brought against our doctrine of time,and that too by disputants who cannot start

any intelligible arguments against the doctrineof the ideality of space, is this—they have nohope of demonstrating apodeictically the abso-lute reality of space, because the doctrine ofidealism is against them, according to whichthe reality of external objects is not capable ofany strict proof. On the other hand, the realityof the object of our internal sense (that is, my-self and my internal state) is clear immediatelythrough consciousness. The former—externalobjects in space—might be a mere delusion, butthe latter—the object of my internal percep-tion—is undeniably real. They do not, however,reflect that both, without question of their real-ity as representations, belong only to the genusphenomenon, which has always two aspects,the one, the object considered as a thing in it-self, without regard to the mode of intuiting it,and the nature of which remains for this veryreason problematical, the other, the form of ourintuition of the object, which must be soughtnot in the object as a thing in itself, but in the

subject to which it appears— which form ofintuition nevertheless belongs really and neces-sarily to the phenomenal object.

Time and space are, therefore, two sources ofknowledge, from which, a priori, various synt-hetical cognitions can be drawn. Of this we finda striking example in the cognitions of spaceand its relations, which form the foundation ofpure mathematics. They are the two pure formsof all intuitions, and thereby make syntheticalpropositions a priori possible. But these sourcesof knowledge being merely conditions of oursensibility, do therefore, and as such, strictlydetermine their own range and purpose, in thatthey do not and cannot present objects asthings in themselves, but are applicable to themsolely in so far as they are considered as sensu-ous phenomena. The sphere of phenomena isthe only sphere of their validity, and if we ven-ture out of this, no further objective use can bemade of them. For the rest, this formal reality of

time and space leaves the validity of our empi-rical knowledge unshaken; for our certainty inthat respect is equally firm, whether theseforms necessarily inhere in the things themsel-ves, or only in our intuitions of them. On theother hand, those who maintain the absolutereality of time and space, whether as essentiallysubsisting, or only inhering, as modifications,in things, must find themselves at utter varian-ce with the principles of experience itself. For, ifthey decide for the first view, and make spaceand time into substances, this being the sidetaken by mathematical natural philosophers,they must admit two self-subsisting nonenti-ties, infinite and eternal, which exist (yet wit-hout there being anything real) for the purposeof containing in themselves everything that isreal. If they adopt the second view of inher-ence, which is preferred by some metaphysicalnatural philosophers, and regard space andtime as relations (contiguity in space or succes-sion in time), abstracted from experience,

though represented confusedly in this state ofseparation, they find themselves in that casenecessitated to deny the validity of mathemati-cal doctrines a priori in reference to real things(for example, in space)—at all events their apo-deictic certainty. For such certainty cannot befound in an a posteriori proposition; and theconceptions a priori of space and time are, ac-cording to this opinion, mere creations of theimagination, having their source really in ex-perience, inasmuch as, out of relations ab-stracted from experience, imagination hasmade up something which contains, indeed,general statements of these relations, yet ofwhich no application can be made without therestrictions attached thereto by nature. Theformer of these parties gains this advantage,that they keep the sphere of phenomena freefor mathematical science. On the other hand,these very conditions (space and time) embar-rass them greatly, when the understandingendeavours to pass the limits of that sphere.

The latter has, indeed, this advantage, that therepresentations of space and time do not comein their way when they wish to judge of objects,not as phenomena, but merely in their relationto the understanding. Devoid, however, of atrue and objectively valid a priori intuition,they can neither furnish any basis for the possi-bility of mathematical cognitions a priori, norbring the propositions of experience into neces-sary accordance with those of mathematics. Inour theory of the true nature of these two origi-nal forms of the sensibility, both difficulties aresurmounted.

In conclusion, that transcendental aestheticcannot contain any more than these two ele-ments—space and time, is sufficiently obviousfrom the fact that all other conceptions apper-taining to sensibility, even that of motion,which unites in itself both elements, presuppo-se something empirical. Motion, for example,presupposes the perception of something mov-

able. But space considered in itself containsnothing movable, consequently motion must besomething which is found in space onlythrough experience— in other words, an em-pirical datum. In like manner, transcendentalaesthetic cannot number the conception ofchange among its data a priori; for time itselfdoes not change, but only something which isin time. To acquire the conception of change,therefore, the perception of some existing objectand of the succession of its determinations, inone word, experience, is necessary.

SS 9. General Remarks on Transcendental Aes-thetic.

I. In order to prevent any misunderstanding, itwill be requisite, in the first place, to recapitula-te, as clearly as possible, what our opinion iswith respect to the fundamental nature of our

sensuous cognition in general. We have in-tended, then, to say that all our intuition isnothing but the representation of phenomena;that the things which we intuite, are not inthemselves the same as our representations ofthem in intuition, nor are their relations inthemselves so constituted as they appear to us;and that if we take away the subject, or evenonly the subjective constitution of our senses ingeneral, then not only the nature and relationsof objects in space and time, but even space andtime themselves disappear; and that these, asphenomena, cannot exist in themselves, butonly in us. What may be the nature of objectsconsidered as things in themselves and withoutreference to the receptivity of our sensibility isquite unknown to us. We know nothing morethan our mode of perceiving them, which ispeculiar to us, and which, though not of neces-sity pertaining to every animated being, is so tothe whole human race. With this alone we haveto do. Space and time are the pure forms

thereof; sensation the matter. The former alonecan we cognize a priori, that is, antecedent toall actual perception; and for this reason suchcognition is called pure intuition. The latter isthat in our cognition which is called cognition aposteriori, that is, empirical intuition. The for-mer appertain absolutely and necessarily to oursensibility, of whatsoever kind our sensationsmay be; the latter may be of very diversifiedcharacter. Supposing that we should carry ourempirical intuition even to the very highestdegree of clearness, we should not thereby ad-vance one step nearer to a knowledge of theconstitution of objects as things in themselves.For we could only, at best, arrive at a completecognition of our own mode of intuition, that isof our sensibility, and this always under theconditions originally attaching to the subject,namely, the conditions of space and time; whilethe question: "What are objects considered asthings in themselves?" remains unanswerable

even after the most thorough examination ofthe phenomenal world.

To say, then, that all our sensibility is nothingbut the confused representation of things con-taining exclusively that which belongs to themas things in themselves, and this under an ac-cumulation of characteristic marks and partialrepresentations which we cannot distinguish inconsciousness, is a falsification of the concep-tion of sensibility and phenomenization, whichrenders our whole doctrine thereof empty anduseless. The difference between a confused anda clear representation is merely logical and hasnothing to do with content. No doubt the con-ception of right, as employed by a sound un-derstanding, contains all that the most subtleinvestigation could unfold from it, although, inthe ordinary practical use of the word, we arenot conscious of the manifold representationscomprised in the conception. But we cannot forthis reason assert that the ordinary conception

is a sensuous one, containing a mere phenome-non, for right cannot appear as a phenomenon;but the conception of it lies in the understand-ing, and represents a property (the moral prop-erty) of actions, which belongs to them inthemselves. On the other hand, the representa-tion in intuition of a body contains nothingwhich could belong to an object considered as athing in itself, but merely the phenomenon orappearance of something, and the mode inwhich we are affected by that appearance; andthis receptivity of our faculty of cognition iscalled sensibility, and remains toto caelo differ-ent from the cognition of an object in itself,even though we should examine the content ofthe phenomenon to the very bottom.

It must be admitted that the Leibnitz-Wolfianphilosophy has assigned an entirely erroneouspoint of view to all investigations into the na-ture and origin of our cognitions, inasmuch asit regards the distinction between the sensuous

and the intellectual as merely logical, whereasit is plainly transcendental, and concerns notmerely the clearness or obscurity, but the con-tent and origin of both. For the faculty of sensi-bility not only does not present us with an in-distinct and confused cognition of objects asthings in themselves, but, in fact, gives us noknowledge of these at all. On the contrary, sosoon as we abstract in thought our own subjec-tive nature, the object represented, with theproperties ascribed to it by sensuous intuition,entirely disappears, because it was only thissubjective nature that determined the form ofthe object as a phenomenon.

In phenomena, we commonly, indeed, distin-guish that which essentially belongs to the in-tuition of them, and is valid for the sensuousfaculty of every human being, from that whichbelongs to the same intuition accidentally, asvalid not for the sensuous faculty in general,but for a particular state or organization of this

or that sense. Accordingly, we are accustomedto say that the former is a cognition which re-presents the object itself, whilst the latter pre-sents only a particular appearance or phe-nomenon thereof. This distinction, however, isonly empirical. If we stop here (as is usual), anddo not regard the empirical intuition as itself amere phenomenon (as we ought to do), inwhich nothing that can appertain to a thing initself is to be found, our transcendental distinc-tion is lost, and we believe that we cognize ob-jects as things in themselves, although in thewhole range of the sensuous world, investigatethe nature of its objects as profoundly as wemay, we have to do with nothing but phenom-ena. Thus, we call the rainbow a mere appear-ance of phenomenon in a sunny shower, andthe rain, the reality or thing in itself; and this isright enough, if we understand the latter con-ception in a merely physical sense, that is, asthat which in universal experience, and underwhatever conditions of sensuous perception, is

known in intuition to be so and so determined,and not otherwise. But if we consider this em-pirical datum generally, and inquire, withoutreference to its accordance with all our senses,whether there can be discovered in it aughtwhich represents an object as a thing in itself(the raindrops of course are not such, for theyare, as phenomena, empirical objects), the ques-tion of the relation of the representation to theobject is transcendental; and not only are theraindrops mere phenomena, but even their cir-cular form, nay, the space itself through whichthey fall, is nothing in itself, but both are meremodifications or fundamental dispositions ofour sensuous intuition, whilst the transcenden-tal object remains for us utterly unknown.

The second important concern of our aestheticis that it does not obtain favour merely as aplausible hypothesis, but possess as undoubteda character of certainty as can be demanded ofany theory which is to serve for an organon. In

order fully to convince the reader of this cer-tainty, we shall select a case which will serve tomake its validity apparent, and also to illustratewhat has been said in SS 3.

Suppose, then, that space and time are in them-selves objective, and conditions of the—possibility of objects as things in themselves. Inthe first place, it is evident that both present us,with very many apodeictic and synthetic pro-positions a priori, but especially space—and forthis reason we shall prefer it for investigation atpresent. As the propositions of geometry arecognized synthetically a priori, and with apo-deictic certainty, I inquire: Whence do you ob-tain propositions of this kind, and on what ba-sis does the understanding rest, in order to ar-rive at such absolutely necessary and univer-sally valid truths?

There is no other way than through intuitionsor conceptions, as such; and these are giveneither a priori or a posteriori. The latter, name-

ly, empirical conceptions, together with theempirical intuition on which they are founded,cannot afford any synthetical proposition, ex-cept such as is itself also empirical, that is, aproposition of experience. But an empiricalproposition cannot possess the qualities of ne-cessity and absolute universality, which, never-theless, are the characteristics of all geometricalpropositions. As to the first and only means toarrive at such cognitions, namely, throughmere conceptions or intuitions a priori, it isquite clear that from mere conceptions no syn-thetical cognitions, but only analytical ones, canbe obtained. Take, for example, the proposition:"Two straight lines cannot enclose a space, andwith these alone no figure is possible," and tryto deduce it from the conception of a straightline and the number two; or take the proposi-tion: "It is possible to construct a figure withthree straight lines," and endeavour, in likemanner, to deduce it from the mere conceptionof a straight line and the number three. All

your endeavours are in vain, and you findyourself forced to have recourse to intuition, as,in fact, geometry always does. You thereforegive yourself an object in intuition. But of whatkind is this intuition? Is it a pure a priori, or is itan empirical intuition? If the latter, then neitheran universally valid, much less an apodeicticproposition can arise from it, for experiencenever can give us any such proposition. Youmust, therefore, give yourself an object a prioriin intuition, and upon that ground your syn-thetical proposition. Now if there did not existwithin you a faculty of intuition a priori; if thissubjective condition were not in respect to itsform also the universal condition a priori underwhich alone the object of this external intuitionis itself possible; if the object (that is, the trian-gle) were something in itself, without relationto you the subject; how could you affirm thatthat which lies necessarily in your subjectiveconditions in order to construct a triangle, mustalso necessarily belong to the triangle in itself?

For to your conceptions of three lines, youcould not add anything new (that is, the fig-ure); which, therefore, must necessarily befound in the object, because the object is givenbefore your cognition, and not by means of it.If, therefore, space (and time also) were not amere form of your intuition, which containsconditions a priori, under which alone thingscan become external objects for you, and with-out which subjective conditions the objects arein themselves nothing, you could not constructany synthetical proposition whatsoever regard-ing external objects. It is therefore not merelypossible or probable, but indubitably certain,that space and time, as the necessary conditionsof all our external and internal experience, aremerely subjective conditions of all our intui-tions, in relation to which all objects are there-fore mere phenomena, and not things in them-selves, presented to us in this particular man-ner. And for this reason, in respect to the formof phenomena, much may be said a priori,

whilst of the thing in itself, which may lie at thefoundation of these phenomena, it is impossi-ble to say anything.

II. In confirmation of this theory of the idealityof the external as well as internal sense, conse-quently of all objects of sense, as mere phe-nomena, we may especially remark that all inour cognition that belongs to intuition containsnothing more than mere relations. (The feelingsof pain and pleasure, and the will, which arenot cognitions, are excepted.) The relations, towit, of place in an intuition (extension), changeof place (motion), and laws according to whichthis change is determined (moving forces).That, however, which is present in this or thatplace, or any operation going on, or result tak-ing place in the things themselves, with theexception of change of place, is not given to usby intuition. Now by means of mere relations, athing cannot be known in itself; and it may the-refore be fairly concluded, that, as through the

external sense nothing but mere representa-tions of relations are given us, the said externalsense in its representation can contain only therelation of the object to the subject, but not theessential nature of the object as a thing in itself.

The same is the case with the internal intuition,not only because, in the internal intuition, therepresentation of the external senses constitutesthe material with which the mind is occupied;but because time, in which we place, and whichitself antecedes the consciousness of, these rep-resentations in experience, and which, as theformal condition of the mode according towhich objects are placed in the mind, lies at thefoundation of them, contains relations of thesuccessive, the coexistent, and of that whichalways must be coexistent with succession, thepermanent. Now that which, as representation,can antecede every exercise of thought (of anobject), is intuition; and when it contains noth-ing but relations, it is the form of the intuition,

which, as it presents us with no representation,except in so far as something is placed in themind, can be nothing else than the mode inwhich the mind is affected by its own activity,to wit—its presenting to itself representations,consequently the mode in which the mind isaffected by itself; that is, it can be nothing butan internal sense in respect to its form. Every-thing that is represented through the mediumof sense is so far phenomenal; consequently, wemust either refuse altogether to admit an inter-nal sense, or the subject, which is the object ofthat sense, could only be represented by it asphenomenon, and not as it would judge of it-self, if its intuition were pure spontaneous ac-tivity, that is, were intellectual. The difficultyhere lies wholly in the question: How can thesubject have an internal intuition of itself? Butthis difficulty is common to every theory. Theconsciousness of self (apperception) is the sim-ple representation of the "ego"; and if by meansof that representation alone, all the manifold

representations in the subject were spontane-ously given, then our internal intuition wouldbe intellectual. This consciousness in man re-quires an internal perception of the manifoldrepresentations which are previously given inthe subject; and the manner in which these rep-resentations are given in the mind withoutspontaneity, must, on account of this difference(the want of spontaneity), be called sensibility.If the faculty of self-consciousness is to appre-hend what lies in the mind, it must all act thatand can in this way alone produce an intuitionof self. But the form of this intuition, which liesin the original constitution of the mind, deter-mines, in the representation of time, the man-ner in which the manifold representations areto combine themselves in the mind; since thesubject intuites itself, not as it would representitself immediately and spontaneously, but ac-cording to the manner in which the mind isinternally affected, consequently, as it appears,and not as it is.

III. When we say that the intuition of externalobjects, and also the self-intuition of the subject,represent both, objects and subject, in spaceand time, as they affect our senses, that is, asthey appear—this is by no means equivalent toasserting that these objects are mere illusoryappearances. For when we speak of things asphenomena, the objects, nay, even the proper-ties which we ascribe to them, are looked uponas really given; only that, in so far as this or thatproperty depends upon the mode of intuitionof the subject, in the relation of the given objectto the subject, the object as phenomenon is tobe distinguished from the object as a thing initself. Thus I do not say that bodies seem orappear to be external to me, or that my soulseems merely to be given in my self-consciousness, although I maintain that theproperties of space and time, in conformity towhich I set both, as the condition of their exis-tence, abide in my mode of intuition, and not inthe objects in themselves. It would be my own

fault, if out of that which I should reckon asphenomenon, I made mere illusory appear-ance.* But this will not happen, because of ourprinciple of the ideality of all sensuous intui-tions. On the contrary, if we ascribe objectivereality to these forms of representation, it be-comes impossible to avoid changing everythinginto mere appearance. For if we regard spaceand time as properties, which must be found inobjects as things in themselves, as sine quibusnon of the possibility of their existence, andreflect on the absurdities in which we then findourselves involved, inasmuch as we are com-pelled to admit the existence of two infinitethings, which are nevertheless not substances,nor anything really inhering in substances, nay,to admit that they are the necessary conditionsof the existence of all things, and moreover,that they must continue to exist, although allexisting things were annihilated— we cannotblame the good Berkeley for degrading bodiesto mere illusory appearances. Nay, even our

own existence, which would in this case de-pend upon the self-existent reality of such amere nonentity as time, would necessarily bechanged with it into mere appearance—an ab-surdity which no one has as yet been guilty of.

[*Footnote: The predicates of the phenomenoncan be affixed to the object itself in relation toour sensuous faculty; for example, the red col-our or the perfume to the rose. But (illusory)appearance never can be attributed as a predi-cate to an object, for this very reason, that itattributes to this object in itself that which be-longs to it only in relation to our sensuous fac-ulty, or to the subject in general, e.g., the twohandles which were formerly ascribed to Sat-urn. That which is never to be found in the ob-ject itself, but always in the relation of the ob-ject to the subject, and which moreover is in-separable from our representation of the object,we denominate phenomenon. Thus the predi-cates of space and time are rightly attributed to

objects of the senses as such, and in this there isno illusion. On the contrary, if I ascribe rednessof the rose as a thing in itself, or to Saturn hishandles, or extension to all external objects,considered as things in themselves, withoutregarding the determinate relation of these ob-jects to the subject, and without limiting myjudgement to that relation—then, and thenonly, arises illusion.]

IV. In natural theology, where we think of anobject—God—which never can be an object ofintuition to us, and even to himself can neverbe an object of sensuous intuition, we carefullyavoid attributing to his intuition the conditionsof space and time—and intuition all his cogni-tion must be, and not thought, which alwaysincludes limitation. But with what right can wedo this if we make them forms of objects asthings in themselves, and such, moreover, aswould continue to exist as a priori conditions ofthe existence of things, even though the things

themselves were annihilated? For as conditionsof all existence in general, space and time mustbe conditions of the existence of the SupremeBeing also. But if we do not thus make themobjective forms of all things, there is no otherway left than to make them subjective forms ofour mode of intuition—external and internal;which is called sensuous, because it is not pri-mitive, that is, is not such as gives in itself theexistence of the object of the intuition (a modeof intuition which, so far as we can judge, canbelong only to the Creator), but is dependenton the existence of the object, is possible, there-fore, only on condition that the representativefaculty of the subject is affected by the object.

It is, moreover, not necessary that we shouldlimit the mode of intuition in space and time tothe sensuous faculty of man. It may well be thatall finite thinking beings must necessarily inthis respect agree with man (though as to thiswe cannot decide), but sensibility does not on

account of this universality cease to be sensibil-ity, for this very reason, that it is a deduced(intuitus derivativus), and not an original (in-tuitus originarius), consequently not an intel-lectual intuition, and this intuition, as such, forreasons above mentioned, seems to belong sole-ly to the Supreme Being, but never to a beingdependent, quoad its existence, as well as itsintuition (which its existence determines andlimits relatively to given objects). This latterremark, however, must be taken only as anillustration, and not as any proof of the truth ofour aesthetical theory.

SS 10. Conclusion of the Transcendental Aes-thetic.

We have now completely before us one part ofthe solution of the grand general problem oftranscendental philosophy, namely, the ques-

tion: "How are synthetical propositions a prioripossible?" That is to say, we have shown thatwe are in possession of pure a priori intuitions,namely, space and time, in which we find,when in a judgement a priori we pass out be-yond the given conception, something which isnot discoverable in that conception, but is cer-tainly found a priori in the intuition which cor-responds to the conception, and can be unitedsynthetically with it. But the judgements whichthese pure intuitions enable us to make, neverreach farther than to objects of the senses, andare valid only for objects of possible experience.

SECOND PART. TRANSCENDENTALLOGIC.

INTRODUCTION. Idea of a TranscendentalLogic.

I. Of Logic in General.

Our knowledge springs from two main sourcesin the mind, first of which is the faculty or po-wer of receiving representations (receptivity forimpressions); the second is the power of cog-nizing by means of these representations (spon-taneity in the production of conceptions).Through the first an object is given to us;through the second, it is, in relation to the rep-resentation (which is a mere determination ofthe mind), thought. Intuition and conceptionsconstitute, therefore, the elements of all ourknowledge, so that neither conceptions withoutan intuition in some way corresponding tothem, nor intuition without conceptions, can

afford us a cognition. Both are either pure orempirical. They are empirical, when sensation(which presupposes the actual presence of theobject) is contained in them; and pure, when nosensation is mixed with the representation.Sensations we may call the matter of sensuouscognition. Pure intuition consequently containsmerely the form under which something is in-tuited, and pure conception only the form ofthe thought of an object. Only pure intuitionsand pure conceptions are possible a priori; theempirical only a posteriori.

We apply the term sensibility to the receptivityof the mind for impressions, in so far as it is insome way affected; and, on the other hand, wecall the faculty of spontaneously producingrepresentations, or the spontaneity of cognition,understanding. Our nature is so constitutedthat intuition with us never can be other thansensuous, that is, it contains only the mode inwhich we are affected by objects. On the other

hand, the faculty of thinking the object of sen-suous intuition is the understanding. Neither ofthese faculties has a preference over the other.Without the sensuous faculty no object wouldbe given to us, and without the understandingno object would be thought. Thoughts withoutcontent are void; intuitions without concep-tions, blind. Hence it is as necessary for themind to make its conceptions sensuous (that is,to join to them the object in intuition), as to ma-ke its intuitions intelligible (that is, to bringthem under conceptions). Neither of these fac-ulties can exchange its proper function. Under-standing cannot intuite, and the sensuous fac-ulty cannot think. In no other way than fromthe united operation of both, can knowledgearise. But no one ought, on this account, tooverlook the difference of the elements contrib-uted by each; we have rather great reason care-fully to separate and distinguish them. Wetherefore distinguish the science of the laws of

sensibility, that is, aesthetic, from the science ofthe laws of the understanding, that is, logic.

Now, logic in its turn may be considered astwofold—namely, as logic of the general, or ofthe particular use of the understanding. Thefirst contains the absolutely necessary laws ofthought, without which no use whatsoever ofthe understanding is possible, and gives lawstherefore to the understanding, without regardto the difference of objects on which it may beemployed. The logic of the particular use of theunderstanding contains the laws of correctthinking upon a particular class of objects. Theformer may be called elemental logic—the lat-ter, the organon of this or that particular sci-ence. The latter is for the most part employed inthe schools, as a propaedeutic to the sciences,although, indeed, according to the course ofhuman reason, it is the last thing we arrive at,when the science has been already matured,and needs only the finishing touches towards

its correction and completion; for our knowl-edge of the objects of our attempted sciencemust be tolerably extensive and complete be-fore we can indicate the laws by which a sci-ence of these objects can be established.

General logic is again either pure or applied. Inthe former, we abstract all the empirical condi-tions under which the understanding is exer-cised; for example, the influence of the senses,the play of the fantasy or imagination, the lawsof the memory, the force of habit, of inclination,etc., consequently also, the sources of preju-dice—in a word, we abstract all causes fromwhich particular cognitions arise, because thesecauses regard the understanding under certaincircumstances of its application, and, to theknowledge of them experience is required. Pu-re general logic has to do, therefore, merelywith pure a priori principles, and is a canon ofunderstanding and reason, but only in respectof the formal part of their use, be the content

what it may, empirical or transcendental. Gen-eral logic is called applied, when it is directedto the laws of the use of the understanding,under the subjective empirical conditionswhich psychology teaches us. It has thereforeempirical principles, although, at the sametime, it is in so far general, that it applies to theexercise of the understanding, without regardto the difference of objects. On this account,moreover, it is neither a canon of the under-standing in general, nor an organon of a par-ticular science, but merely a cathartic of thehuman understanding.

In general logic, therefore, that part which con-stitutes pure logic must be carefully distin-guished from that which constitutes applied(though still general) logic. The former alone isproperly science, although short and dry, as themethodical exposition of an elemental doctrineof the understanding ought to be. In this, there-

fore, logicians must always bear in mind tworules:

1. As general logic, it makes abstraction of allcontent of the cognition of the understanding,and of the difference of objects, and has to dowith nothing but the mere form of thought.

2. As pure logic, it has no empirical principles,and consequently draws nothing (contrary tothe common persuasion) from psychology,which therefore has no influence on the canonof the understanding. It is a demonstrated doc-trine, and everything in it must be certain com-pletely a priori.

What I called applied logic (contrary to thecommon acceptation of this term, according towhich it should contain certain exercises for thescholar, for which pure logic gives the rules), isa representation of the understanding, and ofthe rules of its necessary employment in con-creto, that is to say, under the accidental condi-

tions of the subject, which may either hinder orpromote this employment, and which are allgiven only empirically. Thus applied logictreats of attention, its impediments and conse-quences, of the origin of error, of the state ofdoubt, hesitation, conviction, etc., and to it isrelated pure general logic in the same way thatpure morality, which contains only the neces-sary moral laws of a free will, is related to prac-tical ethics, which considers these laws underall the impediments of feelings, inclinations,and passions to which men are more or lesssubjected, and which never can furnish us witha true and demonstrated science, because it, aswell as applied logic, requires empirical andpsychological principles.

II. Of Transcendental Logic.

General logic, as we have seen, makes abstrac-tion of all content of cognition, that is, of allrelation of cognition to its object, and regardsonly the logical form in the relation of cogni-tions to each other, that is, the form of thoughtin general. But as we have both pure and em-pirical intuitions (as transcendental aestheticproves), in like manner a distinction might bedrawn between pure and empirical thought (ofobjects). In this case, there would exist a kind oflogic, in which we should not make abstractionof all content of cognition; for or logic whichshould comprise merely the laws of purethought (of an object), would of course excludeall those cognitions which were of empiricalcontent. This kind of logic would also examinethe origin of our cognitions of objects, so far asthat origin cannot be ascribed to the objectsthemselves; while, on the contrary, general lo-gic has nothing to do with the origin of ourcognitions, but contemplates our representa-tions, be they given primitively a priori in our-

selves, or be they only of empirical origin, sole-ly according to the laws which the understand-ing observes in employing them in the processof thought, in relation to each other. Conse-quently, general logic treats of the form of theunderstanding only, which can be applied torepresentations, from whatever source theymay have arisen.

And here I shall make a remark, which the rea-der must bear well in mind in the course of thefollowing considerations, to wit, that not everycognition a priori, but only those throughwhich we cognize that and how certain repre-sentations (intuitions or conceptions) are ap-plied or are possible only a priori; that is to say,the a priori possibility of cognition and the apriori use of it are transcendental. Thereforeneither is space, nor any a priori geometricaldetermination of space, a transcendental Repre-sentation, but only the knowledge that such arepresentation is not of empirical origin, and

the possibility of its relating to objects of ex-perience, although itself a priori, can be calledtranscendental. So also, the application of spaceto objects in general would be transcendental;but if it be limited to objects of sense it is em-pirical. Thus, the distinction of the transcenden-tal and empirical belongs only to the critique ofcognitions, and does not concern the relation ofthese to their object.

Accordingly, in the expectation that there mayperhaps be conceptions which relate a priori toobjects, not as pure or sensuous intuitions, butmerely as acts of pure thought (which are the-refore conceptions, but neither of empirical noraesthetical origin)—in this expectation, I say,we form to ourselves, by anticipation, the ideaof a science of pure understanding and rationalcognition, by means of which we may cogitateobjects entirely a priori. A science of this kind,which should determine the origin, the extent,and the objective validity of such cognitions,

must be called transcendental logic, because ithas not, like general logic, to do with the lawsof understanding and reason in relation to em-pirical as well as pure rational cognitions with-out distinction, but concerns itself with theseonly in an a priori relation to objects.

III. Of the Division of General Logic into Ana-lytic and Dialectic.

The old question with which people sought topush logicians into a corner, so that they musteither have recourse to pitiful sophisms or con-fess their ignorance, and consequently the van-ity of their whole art, is this: "What is truth?"The definition of the word truth, to wit, "theaccordance of the cognition with its object," ispresupposed in the question; but we desire tobe told, in the answer to it, what is the univer-

sal and secure criterion of the truth of everycognition.

To know what questions we may reasonablypropose is in itself a strong evidence of sagacityand intelligence. For if a question be in itselfabsurd and unsusceptible of a rational answer,it is attended with the danger—not to mentionthe shame that falls upon the person who pro-poses it—of seducing the unguarded listenerinto making absurd answers, and we are pre-sented with the ridiculous spectacle of one (asthe ancients said) "milking the he-goat, and theother holding a sieve."

If truth consists in the accordance of a cognitionwith its object, this object must be, ipso facto,distinguished from all others; for a cognition isfalse if it does not accord with the object towhich it relates, although it contains somethingwhich may be affirmed of other objects. Nowan universal criterion of truth would be thatwhich is valid for all cognitions, without dis-

tinction of their objects. But it is evident thatsince, in the case of such a criterion, we makeabstraction of all the content of a cognition (thatis, of all relation to its object), and truth relatesprecisely to this content, it must be utterly ab-surd to ask for a mark of the truth of this con-tent of cognition; and that, accordingly, a suffi-cient, and at the same time universal, test oftruth cannot possibly be found. As we havealready termed the content of a cognition itsmatter, we shall say: "Of the truth of our cogni-tions in respect of their matter, no universal testcan be demanded, because such a demand isself-contradictory."

On the other hand, with regard to our cognitionin respect of its mere form (excluding all con-tent), it is equally manifest that logic, in so faras it exhibits the universal and necessary lawsof the understanding, must in these very lawspresent us with criteria of truth. Whatever con-tradicts these rules is false, because thereby the

understanding is made to contradict its ownuniversal laws of thought; that is, to contradictitself. These criteria, however, apply solely tothe form of truth, that is, of thought in general,and in so far they are perfectly accurate, yet notsufficient. For although a cognition may be per-fectly accurate as to logical form, that is, notself-contradictory, it is notwithstanding quitepossible that it may not stand in agreementwith its object. Consequently, the merely logicalcriterion of truth, namely, the accordance of acognition with the universal and formal laws ofunderstanding and reason, is nothing morethan the conditio sine qua non, or negativecondition of all truth. Farther than this logiccannot go, and the error which depends not onthe form, but on the content of the cognition, ithas no test to discover.

General logic, then, resolves the whole formalbusiness of understanding and reason into itselements, and exhibits them as principles of all

logical judging of our cognitions. This part oflogic may, therefore, be called analytic, and isat least the negative test of truth, because allcognitions must first of an be estimated andtried according to these laws before we proceedto investigate them in respect of their content,in order to discover whether they contain posi-tive truth in regard to their object. Because,however, the mere form of a cognition, accu-rately as it may accord with logical laws, is in-sufficient to supply us with material (objective)truth, no one, by means of logic alone, can ven-ture to predicate anything of or decide concern-ing objects, unless he has obtained, independ-ently of logic, well-grounded information aboutthem, in order afterwards to examine, accord-ing to logical laws, into the use and connection,in a cohering whole, of that information, or,what is still better, merely to test it by them.Notwithstanding, there lies so seductive acharm in the possession of a specious art likethis—an art which gives to all our cognitions

the form of the understanding, although withrespect to the content thereof we may be sadlydeficient—that general logic, which is merely acanon of judgement, has been employed as anorganon for the actual production, or rather forthe semblance of production, of objective asser-tions, and has thus been grossly misapplied.Now general logic, in its assumed character oforganon, is called dialectic.

Different as are the significations in which theancients used this term for a science or an art,we may safely infer, from their actual employ-ment of it, that with them it was nothing elsethan a logic of illusion—a sophistical art forgiving ignorance, nay, even intentional sophis-tries, the colouring of truth, in which the thor-oughness of procedure which logic requireswas imitated, and their topic employed to cloakthe empty pretensions. Now it may be taken asa safe and useful warning, that general logic,considered as an organon, must always be a

logic of illusion, that is, be dialectical, for, as itteaches us nothing whatever respecting thecontent of our cognitions, but merely the for-mal conditions of their accordance with theunderstanding, which do not relate to and arequite indifferent in respect of objects, any at-tempt to employ it as an instrument (organon)in order to extend and enlarge the range of ourknowledge must end in mere prating; any onebeing able to maintain or oppose, with someappearance of truth, any single assertion what-ever.

Such instruction is quite unbecoming the dig-nity of philosophy. For these reasons we havechosen to denominate this part of logic dialec-tic, in the sense of a critique of dialectical illu-sion, and we wish the term to be so understoodin this place.

IV. Of the Division of Transcendental Logicinto Transcendental Analytic and Dialectic.

In transcendental logic we isolate the under-standing (as in transcendental aesthetic thesensibility) and select from our cognitionmerely that part of thought which has its originin the understanding alone. The exercise of thispure cognition, however, depends upon this asits condition, that objects to which it may beapplied be given to us in intuition, for withoutintuition the whole of our cognition is withoutobjects, and is therefore quite void. That part oftranscendental logic, then, which treats of theelements of pure cognition of the understand-ing, and of the principles without which noobject at all can be thought, is transcendentalanalytic, and at the same time a logic of truth.For no cognition can contradict it, without los-ing at the same time all content, that is, losingall reference to an object, and therefore all

truth. But because we are very easily seducedinto employing these pure cognitions and prin-ciples of the understanding by themselves, andthat even beyond the boundaries of experience,which yet is the only source whence we canobtain matter (objects) on which those pureconceptions may be employed—understandingruns the risk of making, by means of emptysophisms, a material and objective use of themere formal principles of the pure understand-ing, and of passing judgements on objectswithout distinction—objects which are not gi-ven to us, nay, perhaps cannot be given to us inany way. Now, as it ought properly to be only acanon for judging of the empirical use of theunderstanding, this kind of logic is misusedwhen we seek to employ it as an organon of theuniversal and unlimited exercise of the under-standing, and attempt with the pure under-standing alone to judge synthetically, affirm,and determine respecting objects in general. Inthis case the exercise of the pure understanding

becomes dialectical. The second part of ourtranscendental logic must therefore be a cri-tique of dialectical illusion, and this critique weshall term transcendental dialectic— not mean-ing it as an art of producing dogmatically suchillusion (an art which is unfortunately too cur-rent among the practitioners of metaphysicaljuggling), but as a critique of understandingand reason in regard to their hyperphysical use.This critique will expose the groundless natureof the pretensions of these two faculties, andinvalidate their claims to the discovery andenlargement of our cognitions merely by meansof transcendental principles, and show that theproper employment of these faculties is to testthe judgements made by the pure understand-ing, and to guard it from sophistical delusion.

TRANSCENDENTAL LOGIC. FIRST DI-VISION.

TRANSCENDENTAL ANALYTIC.

SS I.

Transcendental analytic is the dissection of thewhole of our a priori knowledge into the ele-ments of the pure cognition of the understand-ing. In order to effect our purpose, it is neces-sary: (1) That the conceptions be pure and notempirical; (2) That they belong not to intuitionand sensibility, but to thought and understand-ing; (3) That they be elementary conceptions,and as such, quite different from deduced orcompound conceptions; (4) That our table ofthese elementary conceptions be complete, andfill up the whole sphere of the pure under-standing. Now this completeness of a sciencecannot be accepted with confidence on the

guarantee of a mere estimate of its existence inan aggregate formed only by means of repeatedexperiments and attempts. The completenesswhich we require is possible only by means ofan idea of the totality of the a priori cognitionof the understanding, and through the therebydetermined division of the conceptions whichform the said whole; consequently, only bymeans of their connection in a system. Pureunderstanding distinguishes itself not merelyfrom everything empirical, but also completelyfrom all sensibility. It is a unity self-subsistent,self-sufficient, and not to be enlarged by anyadditions from without. Hence the sum of itscognition constitutes a system to be determinedby and comprised under an idea; and the com-pleteness and articulation of this system can atthe same time serve as a test of the correctnessand genuineness of all the parts of cognitionthat belong to it. The whole of this part of tran-scendental logic consists of two books, of which

the one contains the conceptions, and the otherthe principles of pure understanding.

BOOK I.

SS 2. Analytic of Conceptions.

By the term Analytic of Conceptions, I do notunderstand the analysis of these, or the usualprocess in philosophical investigations of dis-secting the conceptions which present them-selves, according to their content, and so mak-ing them clear; but I mean the hitherto littleattempted dissection of the faculty of under-standing itself, in order to investigate the pos-sibility of conceptions a priori, by looking forthem in the understanding alone, as theirbirthplace, and analysing the pure use of thisfaculty. For this is the proper duty of a tran-scendental philosophy; what remains is the

logical treatment of the conceptions in philoso-phy in general. We shall therefore follow up thepure conceptions even to their germs and be-ginnings in the human understanding, in whichthey lie, until they are developed on occasionspresented by experience, and, freed by thesame understanding from the empirical condi-tions attaching to them, are set forth in theirunalloyed purity.

CHAPTER I. Of the Transcendental Clue tothe Discovery of all Pure Conceptions of theUnderstanding.

SS 3. Introductory.

When we call into play a faculty of cognition,different conceptions manifest themselves ac-cording to the different circumstances, and ma-ke known this faculty, and assemble them-

selves into a more or less extensive collection,according to the time or penetration that hasbeen applied to the consideration of them.Where this process, conducted as it is mechani-cally, so to speak, will end, cannot be deter-mined with certainty. Besides, the conceptionswhich we discover in this haphazard mannerpresent themselves by no means in order andsystematic unity, but are at last coupled to-gether only according to resemblances to eachother, and arranged in series, according to thequantity of their content, from the simpler tothe more complex—series which are anythingbut systematic, though not altogether without acertain kind of method in their construction.

Transcendental philosophy has the advantage,and moreover the duty, of searching for its con-ceptions according to a principle; because theseconceptions spring pure and unmixed out ofthe understanding as an absolute unity, andtherefore must be connected with each other

according to one conception or idea. A connec-tion of this kind, however, furnishes us with aready prepared rule, by which its proper placemay be assigned to every pure conception ofthe understanding, and the completeness of thesystem of all be determined a priori—bothwhich would otherwise have been dependenton mere choice or chance.

SS 4. SECTION 1. Of defined above Use of un-derstanding in General.

The understanding was defined above onlynegatively, as a non-sensuous faculty of cogni-tion. Now, independently of sensibility, wecannot possibly have any intuition; conse-quently, the understanding is no faculty of in-tuition. But besides intuition there is no othermode of cognition, except through conceptions;consequently, the cognition of every, at least of

every human, understanding is a cognitionthrough conceptions—not intuitive, but discur-sive. All intuitions, as sensuous, depend onaffections; conceptions, therefore, upon func-tions. By the word function I understand theunity of the act of arranging diverse representa-tions under one common representation. Con-ceptions, then, are based on the spontaneity ofthought, as sensuous intuitions are on the re-ceptivity of impressions. Now, the understand-ing cannot make any other use of these concep-tions than to judge by means of them. As norepresentation, except an intuition, relates im-mediately to its object, a conception never re-lates immediately to an object, but only to someother representation thereof, be that an intui-tion or itself a conception. A judgement, there-fore, is the mediate cognition of an object, con-sequently the representation of a representationof it. In every judgement there is a conceptionwhich applies to, and is valid for many otherconceptions, and which among these compre-

hends also a given representation, this last be-ing immediately connected with an object. Forexample, in the judgement— "All bodies aredivisible," our conception of divisible applies tovarious other conceptions; among these, how-ever, it is here particularly applied to the con-ception of body, and this conception of bodyrelates to certain phenomena which occur to us.These objects, therefore, are mediately repre-sented by the conception of divisibility. Alljudgements, accordingly, are functions of unityin our representations, inasmuch as, instead ofan immediate, a higher representation, whichcomprises this and various others, is used forour cognition of the object, and thereby manypossible cognitions are collected into one. Butwe can reduce all acts of the understanding tojudgements, so that understanding may be rep-resented as the faculty of judging. For it is, ac-cording to what has been said above, a facultyof thought. Now thought is cognition by meansof conceptions. But conceptions, as predicates

of possible judgements, relate to some repre-sentation of a yet undetermined object. Thusthe conception of body indicates something—for example, metal—which can be cognized bymeans of that conception. It is therefore a con-ception, for the reason alone that other repre-sentations are contained under it, by means ofwhich it can relate to objects. It is therefore thepredicate to a possible judgement; for example:"Every metal is a body." All the functions of theunderstanding therefore can be discovered,when we can completely exhibit the functionsof unity in judgements. And that this may beeffected very easily, the following section willshow.

SS 5. SECTION II. Of the Logical Function ofthe Understanding in Judgements.

If we abstract all the content of a judgement,and consider only the intellectual form thereof,we find that the function of thought in a jud-gement can be brought under four heads, ofwhich each contains three momenta. These maybe conveniently represented in the followingtable:

1 Quantity of judgements Universal Particular Singular

2 3 Quality Relation Affirmative Categorical Negative Hypothetical Infinite Disjunctive

4 Modality Problematical

Assertorical Apodeictical

As this division appears to differ in some,though not essential points, from the usualtechnique of logicians, the following observa-tions, for the prevention of otherwise possiblemisunderstanding, will not be without theiruse.

1. Logicians say, with justice, that in the use ofjudgements in syllogisms, singular judgementsmay be treated like universal ones. For, pre-cisely because a singular judgement has no ex-tent at all, its predicate cannot refer to a part ofthat which is contained in the conception of thesubject and be excluded from the rest. Thepredicate is valid for the whole conception justas if it were a general conception, and had ex-tent, to the whole of which the predicate ap-plied. On the other hand, let us compare a sin-

gular with a general judgement, merely as acognition, in regard to quantity. The singularjudgement relates to the general one, as unityto infinity, and is therefore in itself essentiallydifferent. Thus, if we estimate a singularjudgement (judicium singulare) not merelyaccording to its intrinsic validity as a judge-ment, but also as a cognition generally, accord-ing to its quantity in comparison with that ofother cognitions, it is then entirely differentfrom a general judgement (judicium com-mune), and in a complete table of the momentaof thought deserves a separate place—though,indeed, this would not be necessary in a logiclimited merely to the consideration of the use ofjudgements in reference to each other.

2. In like manner, in transcendental logic, infi-nite must be distinguished from affirmativejudgements, although in general logic they arerightly enough classed under affirmative. Gen-eral logic abstracts all content of the predicate

(though it be negative), and only considerswhether the said predicate be affirmed or de-nied of the subject. But transcendental logicconsiders also the worth or content of this logi-cal affirmation—an affirmation by means of amerely negative predicate, and inquires howmuch the sum total of our cognition gains bythis affirmation. For example, if I say of thesoul, "It is not mortal"—by this negative jud-gement I should at least ward off error. Now,by the proposition, "The soul is not mortal," Ihave, in respect of the logical form, really af-firmed, inasmuch as I thereby place the soul inthe unlimited sphere of immortal beings. Now,because of the whole sphere of possible exis-tences, the mortal occupies one part, and theimmortal the other, neither more nor less isaffirmed by the proposition than that the soul isone among the infinite multitude of thingswhich remain over, when I take away the who-le mortal part. But by this proceeding we ac-complish only this much, that the infinite sphe-

re of all possible existences is in so far limitedthat the mortal is excluded from it, and the soulis placed in the remaining part of the extent ofthis sphere. But this part remains, notwith-standing this exception, infinite, and more andmore parts may be taken away from the wholesphere, without in the slightest degree therebyaugmenting or affirmatively determining ourconception of the soul. These judgements, the-refore, infinite in respect of their logical extent,are, in respect of the content of their cognition,merely limitative; and are consequently entitledto a place in our transcendental table of all themomenta of thought in judgements, becausethe function of the understanding exercised bythem may perhaps be of importance in the fieldof its pure a priori cognition.

3. All relations of thought in judgements arethose (a) of the predicate to the subject; (b) ofthe principle to its consequence; (c) of the di-vided cognition and all the members of the di-

vision to each other. In the first of these threeclasses, we consider only two conceptions; inthe second, two judgements; in the third, sev-eral judgements in relation to each other. Thehypothetical proposition, "If perfect justice ex-ists, the obstinately wicked are punished," con-tains properly the relation to each other of twopropositions, namely, "Perfect justice exists,"and "The obstinately wicked are punished."Whether these propositions are in themselvestrue is a question not here decided. Nothing iscogitated by means of this judgement except acertain consequence. Finally, the disjunctivejudgement contains a relation of two or morepropositions to each other—a relation not ofconsequence, but of logical opposition, in so faras the sphere of the one proposition excludesthat of the other. But it contains at the sametime a relation of community, in so far as all thepropositions taken together fill up the sphere ofthe cognition. The disjunctive judgement con-tains, therefore, the relation of the parts of the

whole sphere of a cognition, since the sphere ofeach part is a complemental part of the sphereof the other, each contributing to form the sumtotal of the divided cognition. Take, for exam-ple, the proposition, "The world exists eitherthrough blind chance, or through internal ne-cessity, or through an external cause." Each ofthese propositions embraces a part of the sphe-re of our possible cognition as to the existenceof a world; all of them taken together, the who-le sphere. To take the cognition out of one ofthese spheres, is equivalent to placing it in oneof the others; and, on the other hand, to place itin one sphere is equivalent to taking it out ofthe rest. There is, therefore, in a disjunctivejudgement a certain community of cognitions,which consists in this, that they mutually ex-clude each other, yet thereby determine, as awhole, the true cognition, inasmuch as, takentogether, they make up the complete content ofa particular given cognition. And this is all that

I find necessary, for the sake of what follows, toremark in this place.

4. The modality of judgements is a quite pecu-liar function, with this distinguishing character-istic, that it contributes nothing to the contentof a judgement (for besides quantity, quality,and relation, there is nothing more that consti-tutes the content of a judgement), but concernsitself only with the value of the copula in rela-tion to thought in general. Problematical jud-gements are those in which the affirmation ornegation is accepted as merely possible (adlibitum). In the assertorical, we regard the pro-position as real (true); in the apodeictical, welook on it as necessary.* Thus the two judge-ments (antecedens et consequens), the relationof which constitutes a hypothetical judgement,likewise those (the members of the division) inwhose reciprocity the disjunctive consists, areonly problematical. In the example above giventhe proposition, "There exists perfect justice," is

not stated assertorically, but as an ad libitumjudgement, which someone may choose toadopt, and the consequence alone is assertori-cal. Hence such judgements may be obviouslyfalse, and yet, taken problematically, be condi-tions of our cognition of the truth. Thus theproposition, "The world exists only by blindchance," is in the disjunctive judgement of pro-blematical import only: that is to say, one mayaccept it for the moment, and it helps us (likethe indication of the wrong road among all theroads that one can take) to find out the trueproposition. The problematical proposition is,therefore, that which expresses only logicalpossibility (which is not objective); that is, itexpresses a free choice to admit the validity ofsuch a proposition—a merely arbitrary recep-tion of it into the understanding. The assertori-cal speaks of logical reality or truth; as, for ex-ample, in a hypothetical syllogism, the antece-dens presents itself in a problematical form inthe major, in an assertorical form in the minor,

and it shows that the proposition is in harmonywith the laws of the understanding. The apo-deictical proposition cogitates the assertoricalas determined by these very laws of the under-standing, consequently as affirming a priori,and in this manner it expresses logical neces-sity. Now because all is here gradually incorpo-rated with the understanding—inasmuch as inthe first place we judge problematically; thenaccept assertorically our judgement as true;lastly, affirm it as inseparably united with theunderstanding, that is, as necessary and apo-deictical— we may safely reckon these threefunctions of modality as so many momenta ofthought.

[*Footnote: Just as if thought were in the firstinstance a function of the understanding; in thesecond, of judgement; in the third, of reason. Aremark which will be explained in the sequel.]

SS 6. SECTION III. Of the Pure Conceptions ofthe Understanding, or Categories.

General logic, as has been repeatedly said, ma-kes abstraction of all content of cognition, andexpects to receive representations from someother quarter, in order, by means of analysis, toconvert them into conceptions. On the contrary,transcendental logic has lying before it the ma-nifold content of a priori sensibility, whichtranscendental aesthetic presents to it in orderto give matter to the pure conceptions of theunderstanding, without which transcendentallogic would have no content, and be thereforeutterly void. Now space and time contain aninfinite diversity of determinations of pure apriori intuition, but are nevertheless the condi-tion of the mind's receptivity, under which alo-ne it can obtain representations of objects, andwhich, consequently, must always affect theconception of these objects. But the spontaneity

of thought requires that this diversity be exam-ined after a certain manner, received into themind, and connected, in order afterwards toform a cognition out of it. This Process I callsynthesis.

By the word synthesis, in its most general signi-fication, I understand the process of joiningdifferent representations to each other and ofcomprehending their diversity in one cognition.This synthesis is pure when the diversity is notgiven empirically but a priori (as that in spaceand time). Our representations must be givenpreviously to any analysis of them; and no con-ceptions can arise, quoad their content, analyti-cally. But the synthesis of a diversity (be it gi-ven a priori or empirically) is the first requisitefor the production of a cognition, which in itsbeginning, indeed, may be crude and confused,and therefore in need of analysis—still, synthe-sis is that by which alone the elements of ourcognitions are collected and united into a cer-

tain content, consequently it is the first thing onwhich we must fix our attention, if we wish toinvestigate the origin of our knowledge.

Synthesis, generally speaking, is, as we shallafterwards see, the mere operation of the ima-gination—a blind but indispensable function ofthe soul, without which we should have nocognition whatever, but of the working ofwhich we are seldom even conscious. But toreduce this synthesis to conceptions is a func-tion of the understanding, by means of whichwe attain to cognition, in the proper meaning ofthe term.

Pure synthesis, represented generally, gives usthe pure conception of the understanding. Butby this pure synthesis, I mean that which restsupon a basis of a priori synthetical unity. Thus,our numeration (and this is more observable inlarge numbers) is a synthesis according to con-ceptions, because it takes place according to acommon basis of unity (for example, the dec-

ade). By means of this conception, therefore,the unity in the synthesis of the manifold be-comes necessary.

By means of analysis different representationsare brought under one conception—an opera-tion of which general logic treats. On the otherhand, the duty of transcendental logic is to re-duce to conceptions, not representations, butthe pure synthesis of representations. The firstthing which must be given to us for the sake ofthe a priori cognition of all objects, is the diver-sity of the pure intuition; the synthesis of thisdiversity by means of the imagination is thesecond; but this gives, as yet, no cognition. Theconceptions which give unity to this pure syn-thesis, and which consist solely in the represen-tation of this necessary synthetical unity, fur-nish the third requisite for the cognition of anobject, and these conceptions are given by theunderstanding.

The same function which gives unity to thedifferent representation in a judgement, givesalso unity to the mere synthesis of differentrepresentations in an intuition; and this unitywe call the pure conception of the understand-ing. Thus, the same understanding, and by thesame operations, whereby in conceptions, bymeans of analytical unity, it produced the logi-cal form of a judgement, introduces, by meansof the synthetical unity of the manifold in intui-tion, a transcendental content into its represen-tations, on which account they are called pureconceptions of the understanding, and theyapply a priori to objects, a result not within thepower of general logic.

In this manner, there arise exactly so many pu-re conceptions of the understanding, applying apriori to objects of intuition in general, as thereare logical functions in all possible judgements.For there is no other function or faculty existingin the understanding besides those enumerated

in that table. These conceptions we shall, withAristotle, call categories, our purpose beingoriginally identical with his, notwithstandingthe great difference in the execution.

TABLE OF THE CATEGORIES

1 2

Of Quantity Of Quality Unity Reality Plurality Negation Totality Limitation

3 Of Relation Of Inherence and Subsistence (substantia etaccidens) Of Causality and Dependence (cause and ef-fect) Of Community (reciprocity between the agentand patient)

4 Of Modality Possibility—Impossibility Existence—Non-existence Necessity—Contingence

This, then, is a catalogue of all the originallypure conceptions of the synthesis which theunderstanding contains a priori, and these con-ceptions alone entitle it to be called a pure un-derstanding; inasmuch as only by them it canrender the manifold of intuition conceivable, inother words, think an object of intuition. Thisdivision is made systematically from a commonprinciple, namely the faculty of judgement(which is just the same as the power ofthought), and has not arisen rhapsodically froma search at haphazard after pure conceptions,respecting the full number of which we nevercould be certain, inasmuch as we employ in-duction alone in our search, without consider-

ing that in this way we can never understandwherefore precisely these conceptions, and no-ne others, abide in the pure understanding. Itwas a design worthy of an acute thinker likeAristotle, to search for these fundamental con-ceptions. Destitute, however, of any guidingprinciple, he picked them up just as they oc-curred to him, and at first hunted out ten,which he called categories (predicaments). Af-terwards be believed that he had discoveredfive others, which were added under the nameof post predicaments. But his catalogue stillremained defective. Besides, there are to befound among them some of the modes of puresensibility (quando, ubi, situs, also prius,simul), and likewise an empirical conception(motus)—which can by no means belong to thisgenealogical register of the pure understand-ing. Moreover, there are deduced conceptions(actio, passio) enumerated among the originalconceptions, and, of the latter, some are entirelywanting.

With regard to these, it is to be remarked, thatthe categories, as the true primitive conceptionsof the pure understanding, have also their purededuced conceptions, which, in a completesystem of transcendental philosophy, must byno means be passed over; though in a merelycritical essay we must be contented with thesimple mention of the fact.

Let it be allowed me to call these pure, but de-duced conceptions of the understanding, thepredicables of the pure understanding, in con-tradistinction to predicaments. If we are in pos-session of the original and primitive, the de-duced and subsidiary conceptions can easily beadded, and the genealogical tree of the under-standing completely delineated. As my presentaim is not to set forth a complete system, butmerely the principles of one, I reserve this taskfor another time. It may be easily executed byany one who will refer to the ontological ma-nuals, and subordinate to the category of cau-

sality, for example, the predicables of force,action, passion; to that of community, those ofpresence and resistance; to the categories ofmodality, those of origination, extinction, chan-ge; and so with the rest. The categories com-bined with the modes of pure sensibility, orwith one another, afford a great number of de-duced a priori conceptions; a complete enu-meration of which would be a useful and notunpleasant, but in this place a perfectly dispen-sable, occupation.

I purposely omit the definitions of the catego-ries in this treatise. I shall analyse these concep-tions only so far as is necessary for the doctrineof method, which is to form a part of this cri-tique. In a system of pure reason, definitions ofthem would be with justice demanded of me,but to give them here would only bide fromour view the main aim of our investigation, atthe same time raising doubts and objections,the consideration of which, without injustice to

our main purpose, may be very well postponedtill another opportunity. Meanwhile, it ought tobe sufficiently clear, from the little we havealready said on this subject, that the formationof a complete vocabulary of pure conceptions,accompanied by all the requisite explanations,is not only a possible, but an easy undertaking.The compartments already exist; it is only nec-essary to fill them up; and a systematic topiclike the present, indicates with perfect precisionthe proper place to which each conception be-longs, while it readily points out any that havenot yet been filled up.

SS 7.

Our table of the categories suggests considera-tions of some importance, which may perhapshave significant results in regard to the scien-tific form of all rational cognitions. For, that this

table is useful in the theoretical part of philoso-phy, nay, indispensable for the sketching of thecomplete plan of a science, so far as that sciencerests upon conceptions a priori, and for divid-ing it mathematically, according to fixed prin-ciples, is most manifest from the fact that it con-tains all the elementary conceptions of the un-derstanding, nay, even the form of a system ofthese in the understanding itself, and conse-quently indicates all the momenta, and also theinternal arrangement of a projected speculativescience, as I have elsewhere shown. [Footnote:In the Metaphysical Principles of Natural Sci-ence.] Here follow some of these observations.

I. This table, which contains four classes of con-ceptions of the understanding, may, in the firstinstance, be divided into two classes, the first ofwhich relates to objects of intuition—pure aswell as empirical; the second, to the existence ofthese objects, either in relation to one another,or to the understanding.

The former of these classes of categories Iwould entitle the mathematical, and the latterthe dynamical categories. The former, as wesee, has no correlates; these are only to befound in the second class. This difference musthave a ground in the nature of the human un-derstanding.

II. The number of the categories in each class isalways the same, namely, three—a fact whichalso demands some consideration, because inall other cases division a priori through concep-tions is necessarily dichotomy. It is to be added,that the third category in each triad always ari-ses from the combination of the second withthe first.

Thus totality is nothing else but plurality con-templated as unity; limitation is merely realityconjoined with negation; community is the cau-sality of a substance, reciprocally determining,and determined by other substances; and fi-nally, necessity is nothing but existence, which

is given through the possibility itself. Let it notbe supposed, however, that the third categoryis merely a deduced, and not a primitive con-ception of the pure understanding. For the con-junction of the first and second, in order toproduce the third conception, requires a par-ticular function of the understanding, which isby no means identical with those which areexercised in the first and second. Thus, the con-ception of a number (which belongs to the ca-tegory of totality) is not always possible, wherethe conceptions of multitude and unity exist(for example, in the representation of the infi-nite). Or, if I conjoin the conception of a causewith that of a substance, it does not follow thatthe conception of influence, that is, how onesubstance can be the cause of something in an-other substance, will be understood from that.Thus it is evident that a particular act of theunderstanding is here necessary; and so in theother instances.

III. With respect to one category, namely, thatof community, which is found in the third class,it is not so easy as with the others to detect itsaccordance with the form of the disjunctivejudgement which corresponds to it in the tableof the logical functions.

In order to assure ourselves of this accordance,we must observe that in every disjunctive jud-gement, the sphere of the judgement (that is,the complex of all that is contained in it) is rep-resented as a whole divided into parts; and,since one part cannot be contained in the other,they are cogitated as co-ordinated with, notsubordinated to each other, so that they do notdetermine each other unilaterally, as in a linearseries, but reciprocally, as in an aggregate—(ifone member of the division is posited, all therest are excluded; and conversely).

Now a like connection is cogitated in a whole ofthings; for one thing is not subordinated, aseffect, to another as cause of its existence, but,

on the contrary, is co-ordinated contemporane-ously and reciprocally, as a cause in relation tothe determination of the others (for example, ina body—the parts of which mutually attractand repel each other). And this is an entirelydifferent kind of connection from that whichwe find in the mere relation of the cause to theeffect (the principle to the consequence), for insuch a connection the consequence does not inits turn determine the principle, and thereforedoes not constitute, with the latter, a whole—just as the Creator does not with the world ma-ke up a whole. The process of understandingby which it represents to itself the sphere of adivided conception, is employed also when wethink of a thing as divisible; and in the samemanner as the members of the division in theformer exclude one another, and yet are con-nected in one sphere, so the understandingrepresents to itself the parts of the latter, ashaving—each of them—an existence (as sub-

stances), independently of the others, and yetas united in one whole.

SS 8.

In the transcendental philosophy of the an-cients there exists one more leading division,which contains pure conceptions of the under-standing, and which, although not numberedamong the categories, ought, according tothem, as conceptions a priori, to be valid ofobjects. But in this case they would augmentthe number of the categories; which cannot be.These are set forth in the proposition, so re-nowned among the schoolmen—"Quodlibetens est UNUM, VERUM, BONUM." Now,though the inferences from this principle weremere tautological propositions, and though it isallowed only by courtesy to retain a place inmodern metaphysics, yet a thought which

maintained itself for such a length of time,however empty it seems to be, deserves an in-vestigation of its origin, and justifies the conjec-ture that it must be grounded in some law ofthe understanding, which, as is often the case,has only been erroneously interpreted. Thesepretended transcendental predicates are, infact, nothing but logical requisites and criteriaof all cognition of objects, and they employ, asthe basis for this cognition, the categories ofquantity, namely, unity, plurality, and totality.But these, which must be taken as materialconditions, that is, as belonging to the possibil-ity of things themselves, they employed merelyin a formal signification, as belonging to thelogical requisites of all cognition, and yet mostunguardedly changed these criteria of thoughtinto properties of objects, as things in them-selves. Now, in every cognition of an object,there is unity of conception, which may becalled qualitative unity, so far as by this termwe understand only the unity in our connection

of the manifold; for example, unity of thetheme in a play, an oration, or a story. Sec-ondly, there is truth in respect of the deduc-tions from it. The more true deductions wehave from a given conception, the more criteriaof its objective reality. This we might call thequalitative plurality of characteristic marks,which belong to a conception as to a commonfoundation, but are not cogitated as a quantityin it. Thirdly, there is perfection—which con-sists in this, that the plurality falls back uponthe unity of the conception, and accords com-pletely with that conception and with no other.This we may denominate qualitative complete-ness. Hence it is evident that these logical crite-ria of the possibility of cognition are merely thethree categories of quantity modified and trans-formed to suit an unauthorized manner of ap-plying them. That is to say, the three categories,in which the unity in the production of thequantum must be homogeneous throughout,are transformed solely with a view to the con-

nection of heterogeneous parts of cognition inone act of consciousness, by means of the qual-ity of the cognition, which is the principle ofthat connection. Thus the criterion of the possi-bility of a conception (not of its object) is thedefinition of it, in which the unity of the con-ception, the truth of all that may be immedi-ately deduced from it, and finally, the com-pleteness of what has been thus deduced, con-stitute the requisites for the reproduction of thewhole conception. Thus also, the criterion ortest of an hypothesis is the intelligibility of thereceived principle of explanation, or its unity(without help from any subsidiary hypothe-sis)—the truth of our deductions from it (con-sistency with each other and with experi-ence)—and lastly, the completeness of the prin-ciple of the explanation of these deductions,which refer to neither more nor less than whatwas admitted in the hypothesis, restoring ana-lytically and a posteriori, what was cogitatedsynthetically and a priori. By the conceptions,

therefore, of unity, truth, and perfection, wehave made no addition to the transcendentaltable of the categories, which is complete with-out them. We have, on the contrary, merelyemployed the three categories of quantity, set-ting aside their application to objects of experi-ence, as general logical laws of the consistencyof cognition with itself.

CHAPTER II Of the Deduction of the PureConceptions of the Understanding.

SS 9. SECTION I Of the Principles of a Tran-scendental Deduction in general.

Teachers of jurisprudence, when speaking ofrights and claims, distinguish in a cause thequestion of right (quid juris) from the questionof fact (quid facti), and while they demandproof of both, they give to the proof of the for-

mer, which goes to establish right or claim inlaw, the name of deduction. Now we make useof a great number of empirical conceptions,without opposition from any one; and considerourselves, even without any attempt at deduc-tion, justified in attaching to them a sense, anda supposititious signification, because we havealways experience at hand to demonstrate theirobjective reality. There exist also, however,usurped conceptions, such as fortune, fate,which circulate with almost universal indul-gence, and yet are occasionally challenged bythe question, "quid juris?" In such cases, wehave great difficulty in discovering any deduc-tion for these terms, inasmuch as we cannotproduce any manifest ground of right, eitherfrom experience or from reason, on which theclaim to employ them can be founded.

Among the many conceptions, which make upthe very variegated web of human cognition,some are destined for pure use a priori, inde-

pendent of all experience; and their title to beso employed always requires a deduction, in-asmuch as, to justify such use of them, proofsfrom experience are not sufficient; but it is nec-essary to know how these conceptions can ap-ply to objects without being derived from ex-perience. I term, therefore, an examination ofthe manner in which conceptions can apply apriori to objects, the transcendental deductionof conceptions, and I distinguish it from theempirical deduction, which indicates the modein which conception is obtained through ex-perience and reflection thereon; consequently,does not concern itself with the right, but onlywith the fact of our obtaining conceptions insuch and such a manner. We have already seenthat we are in possession of two perfectly dif-ferent kinds of conceptions, which neverthelessagree with each other in this, that they bothapply to objects completely a priori. These arethe conceptions of space and time as forms ofsensibility, and the categories as pure concep-

tions of the understanding. To attempt an em-pirical deduction of either of these classeswould be labour in vain, because the distin-guishing characteristic of their nature consistsin this, that they apply to their objects, withouthaving borrowed anything from experiencetowards the representation of them. Conse-quently, if a deduction of these conceptions isnecessary, it must always be transcendental.

Meanwhile, with respect to these conceptions,as with respect to all our cognition, we cer-tainly may discover in experience, if not theprinciple of their possibility, yet the occasion-ing causes of their production. It will be foundthat the impressions of sense give the first occa-sion for bringing into action the whole facultyof cognition, and for the production of experi-ence, which contains two very dissimilar ele-ments, namely, a matter for cognition, given bythe senses, and a certain form for the arrange-ment of this matter, arising out of the inner

fountain of pure intuition and thought; andthese, on occasion given by sensuous impres-sions, are called into exercise and produce con-ceptions. Such an investigation into the firstefforts of our faculty of cognition to mountfrom particular perceptions to general concep-tions is undoubtedly of great utility; and wehave to thank the celebrated Locke for havingfirst opened the way for this inquiry. But a de-duction of the pure a priori conceptions ofcourse never can be made in this way, seeingthat, in regard to their future employment,which must be entirely independent of experi-ence, they must have a far different certificateof birth to show from that of a descent fromexperience. This attempted physiological deri-vation, which cannot properly be called deduc-tion, because it relates merely to a quaestiofacti, I shall entitle an explanation of the pos-session of a pure cognition. It is therefore mani-fest that there can only be a transcendental de-duction of these conceptions and by no means

an empirical one; also, that all attempts at anempirical deduction, in regard to pure a prioriconceptions, are vain, and can only be made byone who does not understand the altogetherpeculiar nature of these cognitions.

But although it is admitted that the only possi-ble deduction of pure a priori cognition is atranscendental deduction, it is not, for that rea-son, perfectly manifest that such a deduction isabsolutely necessary. We have already traced totheir sources the conceptions of space and time,by means of a transcendental deduction, andwe have explained and determined their objec-tive validity a priori. Geometry, nevertheless,advances steadily and securely in the provinceof pure a priori cognitions, without needing toask from philosophy any certificate as to thepure and legitimate origin of its fundamentalconception of space. But the use of the concep-tion in this science extends only to the externalworld of sense, the pure form of the intuition of

which is space; and in this world, therefore, allgeometrical cognition, because it is foundedupon a priori intuition, possesses immediateevidence, and the objects of this cognition aregiven a priori (as regards their form) in intui-tion by and through the cognition itself. Withthe pure conceptions of understanding, on thecontrary, commences the absolute necessity ofseeking a transcendental deduction, not only ofthese conceptions themselves, but likewise ofspace, because, inasmuch as they make affirma-tions concerning objects not by means of thepredicates of intuition and sensibility, but ofpure thought a priori, they apply to objectswithout any of the conditions of sensibility.Besides, not being founded on experience, theyare not presented with any object in a prioriintuition upon which, antecedently to experi-ence, they might base their synthesis. Henceresults, not only doubt as to the objective valid-ity and proper limits of their use, but that evenour conception of space is rendered equivocal;

inasmuch as we are very ready with the aid ofthe categories, to carry the use of this concep-tion beyond the conditions of sensuous intui-tion—and, for this reason, we have alreadyfound a transcendental deduction of it needful.The reader, then, must be quite convinced ofthe absolute necessity of a transcendental de-duction, before taking a single step in the fieldof pure reason; because otherwise he goes towork blindly, and after he has wondered aboutin all directions, returns to the state of utterignorance from which he started. He ought,moreover, clearly to recognize beforehand theunavoidable difficulties in his undertaking, sothat he may not afterwards complain of theobscurity in which the subject itself is deeplyinvolved, or become too soon impatient of theobstacles in his path; because we have a choiceof only two things—either at once to give up allpretensions to knowledge beyond the limits ofpossible experience, or to bring this critical in-vestigation to completion.

We have been able, with very little trouble, tomake it comprehensible how the conceptions ofspace and time, although a priori cognitions,must necessarily apply to external objects, andrender a synthetical cognition of these possible,independently of all experience. For inasmuchas only by means of such pure form of sensibil-ity an object can appear to us, that is, be an ob-ject of empirical intuition, space and time arepure intuitions, which contain a priori the con-dition of the possibility of objects as phenom-ena, and an a priori synthesis in these intuitionspossesses objective validity.

On the other hand, the categories of the under-standing do not represent the conditions underwhich objects are given to us in intuition; ob-jects can consequently appear to us withoutnecessarily connecting themselves with these,and consequently without any necessity bind-ing on the understanding to contain a priori theconditions of these objects. Thus we find our-

selves involved in a difficulty which did notpresent itself in the sphere of sensibility, that isto say, we cannot discover how the subjectiveconditions of thought can have objective valid-ity, in other words, can become conditions ofthe possibility of all cognition of objects; forphenomena may certainly be given to us inintuition without any help from the functionsof the understanding. Let us take, for example,the conception of cause, which indicates a pe-culiar kind of synthesis, namely, that with so-mething, A, something entirely different, B, isconnected according to a law. It is not a priorimanifest why phenomena should contain any-thing of this kind (we are of course debarredfrom appealing for proof to experience, for theobjective validity of this conception must bedemonstrated a priori), and it hence remainsdoubtful a priori, whether such a conception benot quite void and without any correspondingobject among phenomena. For that objects ofsensuous intuition must correspond to the for-

mal conditions of sensibility existing a priori inthe mind is quite evident, from the fact thatwithout these they could not be objects for us;but that they must also correspond to the con-ditions which understanding requires for thesynthetical unity of thought is an assertion, thegrounds for which are not so easily to be dis-covered. For phenomena might be so consti-tuted as not to correspond to the conditions ofthe unity of thought; and all things might lie insuch confusion that, for example, nothing couldbe met with in the sphere of phenomena tosuggest a law of synthesis, and so correspondto the conception of cause and effect; so thatthis conception would be quite void, null, andwithout significance. Phenomena would never-theless continue to present objects to our intui-tion; for mere intuition does not in any respectstand in need of the functions of thought.

If we thought to free ourselves from the labourof these investigations by saying: "Experience is

constantly offering us examples of the relationof cause and effect in phenomena, and presentsus with abundant opportunity of abstractingthe conception of cause, and so at the same ti-me of corroborating the objective validity ofthis conception"; we should in this case be over-looking the fact, that the conception of causecannot arise in this way at all; that, on the con-trary, it must either have an a priori basis in theunderstanding, or be rejected as a mere chi-mera. For this conception demands that some-thing, A, should be of such a nature that some-thing else, B, should follow from it necessarily,and according to an absolutely universal law.We may certainly collect from phenomena alaw, according to which this or that usuallyhappens, but the element of necessity is not tobe found in it. Hence it is evident that to thesynthesis of cause and effect belongs a dignity,which is utterly wanting in any empirical syn-thesis; for it is no mere mechanical synthesis,by means of addition, but a dynamical one; that

is to say, the effect is not to be cogitated asmerely annexed to the cause, but as posited byand through the cause, and resulting from it.The strict universality of this law never can be acharacteristic of empirical laws, which obtainthrough induction only a comparative univer-sality, that is, an extended range of practicalapplication. But the pure conceptions of theunderstanding would entirely lose all their pe-culiar character, if we treated them merely asthe productions of experience.

SS 10. Transition to the Transcendental Deduc-tion of the Categories.

There are only two possible ways in which syn-thetical representation and its objects can coin-cide with and relate necessarily to each other,and, as it were, meet together. Either the object

alone makes the representation possible, or therepresentation alone makes the object possible.In the former case, the relation between them isonly empirical, and an a priori representation isimpossible. And this is the case with phenom-ena, as regards that in them which is referableto mere sensation. In the latter case—althoughrepresentation alone (for of its causality, bymeans of the will, we do not here speak) doesnot produce the object as to its existence, itmust nevertheless be a priori determinative inregard to the object, if it is only by means of therepresentation that we can cognize anything asan object. Now there are only two conditions ofthe possibility of a cognition of objects; firstly,intuition, by means of which the object, thoughonly as phenomenon, is given; secondly, con-ception, by means of which the object whichcorresponds to this intuition is thought. But it isevident from what has been said on aestheticthat the first condition, under which alone ob-jects can be intuited, must in fact exist, as a

formal basis for them, a priori in the mind.With this formal condition of sensibility, there-fore, all phenomena necessarily correspond,because it is only through it that they can bephenomena at all; that is, can be empiricallyintuited and given. Now the question is whet-her there do not exist, a priori in the mind, con-ceptions of understanding also, as conditionsunder which alone something, if not intuited, isyet thought as object. If this question be an-swered in the affirmative, it follows that allempirical cognition of objects is necessarilyconformable to such conceptions, since, if theyare not presupposed, it is impossible that any-thing can be an object of experience. Now allexperience contains, besides the intuition of thesenses through which an object is given, a con-ception also of an object that is given in intui-tion. Accordingly, conceptions of objects ingeneral must lie as a priori conditions at thefoundation of all empirical cognition; and con-sequently, the objective validity of the catego-

ries, as a priori conceptions, will rest upon this,that experience (as far as regards the form ofthought) is possible only by their means. For inthat case they apply necessarily and a priori toobjects of experience, because only throughthem can an object of experience be thought.

The whole aim of the transcendental deductionof all a priori conceptions is to show that theseconceptions are a priori conditions of the possi-bility of all experience. Conceptions which af-ford us the objective foundation of the possibil-ity of experience are for that very reason neces-sary. But the analysis of the experiences inwhich they are met with is not deduction, butonly an illustration of them, because from ex-perience they could never derive the attributeof necessity. Without their original applicabilityand relation to all possible experience, in whichall objects of cognition present themselves, therelation of the categories to objects, of whatevernature, would be quite incomprehensible.

The celebrated Locke, for want of due reflectionon these points, and because he met with pureconceptions of the understanding in experi-ence, sought also to deduce them from experi-ence, and yet proceeded so inconsequently asto attempt, with their aid, to arrive it cognitionswhich lie far beyond the limits of all experi-ence. David Hume perceived that, to renderthis possible, it was necessary that the concep-tions should have an a priori origin. But as hecould not explain how it was possible that con-ceptions which are not connected with eachother in the understanding must neverthelessbe thought as necessarily connected in the ob-ject—and it never occurred to him that the un-derstanding itself might, perhaps, by means ofthese conceptions, be the author of the experi-ence in which its objects were presented to it—he was forced to drive these conceptions fromexperience, that is, from a subjective necessityarising from repeated association of experi-ences erroneously considered to be objective—

in one word, from habit. But he proceeded withperfect consequence and declared it to be im-possible, with such conceptions and the princi-ples arising from them, to overstep the limits ofexperience. The empirical derivation, however,which both of these philosophers attributed tothese conceptions, cannot possibly be recon-ciled with the fact that we do possess scientifica priori cognitions, namely, those of pure mat-hematics and general physics.

The former of these two celebrated men openeda wide door to extravagance—(for if reason hasonce undoubted right on its side, it will notallow itself to be confined to set limits, by va-gue recommendations of moderation); the lattergave himself up entirely to scepticism—a natu-ral consequence, after having discovered, as hethought, that the faculty of cognition was nottrustworthy. We now intend to make a trialwhether it be not possible safely to conductreason between these two rocks, to assign her

determinate limits, and yet leave open for herthe entire sphere of her legitimate activity.

I shall merely premise an explanation of whatthe categories are. They are conceptions of anobject in general, by means of which its intui-tion is contemplated as determined in relationto one of the logical functions of judgement.The following will make this plain. The func-tion of the categorical judgement is that of therelation of subject to predicate; for example, inthe proposition: "All bodies are divisible." Butin regard to the merely logical use of the under-standing, it still remains undetermined towhich Of these two conceptions belongs thefunction Of subject and to which that of predi-cate. For we could also say: "Some divisible is abody." But the category of substance, when theconception of a body is brought under it, de-termines that; and its empirical intuition in ex-perience must be contemplated always as sub-

ject and never as mere predicate. And so withall the other categories.

SS 11. SECTION II Transcendental Deductionof the pure Conceptions of the Understanding.

Of the Possibility of a Conjunction of the mani-fold representations given by Sense.

The manifold content in our representationscan be given in an intuition which is merelysensuous—in other words, is nothing but sus-ceptibility; and the form of this intuition canexist a priori in our faculty of representation,without being anything else but the mode inwhich the subject is affected. But the conjunc-tion (conjunctio) of a manifold in intuition ne-ver can be given us by the senses; it cannot the-refore be contained in the pure form of sensu-ous intuition, for it is a spontaneous act of the

faculty of representation. And as we must, todistinguish it from sensibility, entitle this fac-ulty understanding; so all conjunction whetherconscious or unconscious, be it of the manifoldin intuition, sensuous or non-sensuous, or ofseveral conceptions—is an act of the under-standing. To this act we shall give the generalappellation of synthesis, thereby to indicate, atthe same time, that we cannot represent any-thing as conjoined in the object without havingpreviously conjoined it ourselves. Of all mentalnotions, that of conjunction is the only onewhich cannot be given through objects, but canbe originated only by the subject itself, becauseit is an act of its purely spontaneous activity.The reader will easily enough perceive that thepossibility of conjunction must be grounded inthe very nature of this act, and that it must beequally valid for all conjunction, and that ana-lysis, which appears to be its contrary, must,nevertheless, always presuppose it; for wherethe understanding has not previously con-

joined, it cannot dissect or analyse, becauseonly as conjoined by it, must that which is to beanalysed have been given to our faculty of rep-resentation.

But the conception of conjunction includes,besides the conception of the manifold and ofthe synthesis of it, that of the unity of it also.Conjunction is the representation of the syn-thetical unity of the manifold.* This idea of uni-ty, therefore, cannot arise out of that of con-junction; much rather does that idea, by com-bining itself with the representation of themanifold, render the conception of conjunctionpossible. This unity, which a priori precedes allconceptions of conjunction, is not the categoryof unity (SS 6); for all the categories are basedupon logical functions of judgement, and inthese functions we already have conjunction,and consequently unity of given conceptions. Itis therefore evident that the category of unitypresupposes conjunction. We must therefore

look still higher for this unity (as qualitative, SS8), in that, namely, which contains the groundof the unity of diverse conceptions in judge-ments, the ground, consequently, of the possi-bility of the existence of the understanding,even in regard to its logical use.

[*Footnote: Whether the representations are inthemselves identical, and consequently whet-her one can be thought analytically by means ofand through the other, is a question which weneed not at present consider. Our Conscious-ness of the one, when we speak of the manifold,is always distinguishable from our conscious-ness of the other; and it is only respecting thesynthesis of this (possible) consciousness thatwe here treat.]

SS 12. Of the Originally Synthetical Unity ofApperception.

The "I think" must accompany all my represen-tations, for otherwise something would be rep-resented in me which could not be thought; inother words, the representation would either beimpossible, or at least be, in relation to me, not-hing. That representation which can be givenpreviously to all thought is called intuition. Allthe diversity or manifold content of intuition,has, therefore, a necessary relation to the "Ithink," in the subject in which this diversity isfound. But this representation, "I think," is anact of spontaneity; that is to say, it cannot beregarded as belonging to mere sensibility. I callit pure apperception, in order to distinguish itfrom empirical; or primitive apperception, be-cause it is self-consciousness which, whilst itgives birth to the representation "I think," mustnecessarily be capable of accompanying all ourrepresentations. It is in all acts of consciousnessone and the same, and unaccompanied by it, norepresentation can exist for me. The unity ofthis apperception I call the transcendental unity

of self-consciousness, in order to indicate thepossibility of a priori cognition arising from it.For the manifold representations which aregiven in an intuition would not all of them bemy representations, if they did not all belong toone self-consciousness, that is, as my represen-tations (even although I am not conscious ofthem as such), they must conform to the condi-tion under which alone they can exist togetherin a common self-consciousness, because oth-erwise they would not all without exceptionbelong to me. From this primitive conjunctionfollow many important results.

For example, this universal identity of the ap-perception of the manifold given in intuitioncontains a synthesis of representations and ispossible only by means of the consciousness ofthis synthesis. For the empirical consciousnesswhich accompanies different representations isin itself fragmentary and disunited, and with-out relation to the identity of the subject. This

relation, then, does not exist because I accom-pany every representation with consciousness,but because I join one representation to an-other, and am conscious of the synthesis ofthem. Consequently, only because I can connecta variety of given representations in one con-sciousness, is it possible that I can represent tomyself the identity of consciousness in theserepresentations; in other words, the analyticalunity of apperception is possible only underthe presupposition of a synthetical unity.* Thethought, "These representations given in intui-tion belong all of them to me," is accordinglyjust the same as, "I unite them in one self-consciousness, or can at least so unite them";and although this thought is not itself the con-sciousness of the synthesis of representations, itpresupposes the possibility of it; that is to say,for the reason alone that I can comprehend thevariety of my representations in one conscious-ness, do I call them my representations, for oth-erwise I must have as many-coloured and vari-

ous a self as are the representations of which Iam conscious. Synthetical unity of the manifoldin intuitions, as given a priori, is therefore thefoundation of the identity of apperception it-self, which antecedes a priori all determinatethought. But the conjunction of representationsinto a conception is not to be found in objectsthemselves, nor can it be, as it were, borrowedfrom them and taken up into the understand-ing by perception, but it is on the contrary anoperation of the understanding itself, which isnothing more than the faculty of conjoining apriori and of bringing the variety of given rep-resentations under the unity of apperception.This principle is the highest in all human cogni-tion.

[*Footnote: All general conceptions—as such—depend, for their existence, on the analyticalunity of consciousness. For example, when Ithink of red in general, I thereby think to my-self a property which (as a characteristic mark)

can be discovered somewhere, or can be unitedwith other representations; consequently, it isonly by means of a forethought possible syn-thetical unity that I can think to myself the ana-lytical. A representation which is cogitated ascommon to different representations, is re-garded as belonging to such as, besides thiscommon representation, contain somethingdifferent; consequently it must be previouslythought in synthetical unity with other al-though only possible representations, before Ican think in it the analytical unity of conscious-ness which makes it a conceptas communis.And thus the synthetical unity of apperceptionis the highest point with which we must con-nect every operation of the understanding,even the whole of logic, and after it our tran-scendental philosophy; indeed, this faculty isthe understanding itself.]

This fundamental principle of the necessaryunity of apperception is indeed an identical,

and therefore analytical, proposition; but it ne-vertheless explains the necessity for a synthesisof the manifold given in an intuition, withoutwhich the identity of self-consciousness wouldbe incogitable. For the ego, as a simple repre-sentation, presents us with no manifold con-tent; only in intuition, which is quite differentfrom the representation ego, can it be given us,and by means of conjunction it is cogitated inone self-consciousness. An understanding, inwhich all the manifold should be given bymeans of consciousness itself, would be intui-tive; our understanding can only think andmust look for its intuition to sense. I am, there-fore, conscious of my identical self, in relationto all the variety of representations given to mein an intuition, because I call all of them myrepresentations. In other words, I am consciousmyself of a necessary a priori synthesis of myrepresentations, which is called the originalsynthetical unity of apperception, under which

rank all the representations presented to me,but that only by means of a synthesis.

SS 13. The Principle of the Synthetical Unity ofApperception is the highest Principle of all ex-ercise of the Understanding.

The supreme principle of the possibility of allintuition in relation to sensibility was, accord-ing to our transcendental aesthetic, that all themanifold in intuition be subject to the formalconditions of space and time. The supremeprinciple of the possibility of it in relation to theunderstanding is that all the manifold in it besubject to conditions of the originally syntheti-cal unity or apperception.* To the former ofthese two principles are subject all the variousrepresentations of intuition, in so far as they aregiven to us; to the latter, in so far as they mustbe capable of conjunction in one consciousness;

for without this nothing can be thought or cog-nized, because the given representations wouldnot have in common the act Of the appercep-tion "I think" and therefore could not be con-nected in one self-consciousness.

[*Footnote: Space and time, and all portionsthereof, are intuitions; consequently are, with amanifold for their content, single representa-tions. (See the Transcendental Aesthetic.) Con-sequently, they are not pure conceptions, bymeans of which the same consciousness isfound in a great number of representations;but, on the contrary, they are many representa-tions contained in one, the consciousness ofwhich is, so to speak, compounded. The unityof consciousness is nevertheless syntheticaland, therefore, primitive. From this peculiarcharacter of consciousness follow many impor-tant consequences. (See SS 21.)]

Understanding is, to speak generally, the fac-ulty Of cognitions. These consist in the deter-

mined relation of given representation to anobject. But an object is that, in the conception ofwhich the manifold in a given intuition is uni-ted. Now all union of representations requiresunity of consciousness in the synthesis of them.Consequently, it is the unity of consciousnessalone that constitutes the possibility of repre-sentations relating to an object, and therefore oftheir objective validity, and of their becomingcognitions, and consequently, the possibility ofthe existence of the understanding itself.

The first pure cognition of understanding, then,upon which is founded all its other exercise,and which is at the same time perfectly inde-pendent of all conditions of mere sensuous in-tuition, is the principle of the original syntheti-cal unity of apperception. Thus the mere formof external sensuous intuition, namely, space,affords us, per se, no cognition; it merely con-tributes the manifold in a priori intuition to apossible cognition. But, in order to cognize so-

mething in space (for example, a line), I mustdraw it, and thus produce synthetically a de-termined conjunction of the given manifold, sothat the unity of this act is at the same time theunity of consciousness (in the conception of aline), and by this means alone is an object (adeterminate space) cognized. The syntheticalunity of consciousness is, therefore, an objectivecondition of all cognition, which I do not mere-ly require in order to cognize an object, but towhich every intuition must necessarily be sub-ject, in order to become an object for me; be-cause in any other way, and without this syn-thesis, the manifold in intuition could not beunited in one consciousness.

This proposition is, as already said, itself ana-lytical, although it constitutes the syntheticalunity, the condition of all thought; for it statesnothing more than that all my representationsin any given intuition must be subject to thecondition which alone enables me to connect

them, as my representation with the identicalself, and so to unite them synthetically in oneapperception, by means of the general expres-sion, "I think."

But this principle is not to be regarded as aprinciple for every possible understanding, butonly for the understanding by means of whosepure apperception in the thought I am, no ma-nifold content is given. The understanding ormind which contained the manifold in intui-tion, in and through the act itself of its ownself-consciousness, in other words, an under-standing by and in the representation of whichthe objects of the representation should at thesame time exist, would not require a special actof synthesis of the manifold as the condition ofthe unity of its consciousness, an act of whichthe human understanding, which thinks onlyand cannot intuite, has absolute need. But thisprinciple is the first principle of all the opera-tions of our understanding, so that we cannot

form the least conception of any other possibleunderstanding, either of one such as should beitself intuition, or possess a sensuous intuition,but with forms different from those of spaceand time.

SS 14. What Objective Unity of Self-consciousness is.

It is by means of the transcendental unity ofapperception that all the manifold, given in anintuition is united into a conception of the ob-ject. On this account it is called objective, andmust be distinguished from the subjective unityof consciousness, which is a determination ofthe internal sense, by means of which the saidmanifold in intuition is given empirically to beso united. Whether I can be empirically con-scious of the manifold as coexistent or as suc-cessive, depends upon circumstances, or em-

pirical conditions. Hence the empirical unity ofconsciousness by means of association of repre-sentations, itself relates to a phenomenal worldand is wholly contingent. On the contrary, thepure form of intuition in time, merely as anintuition, which contains a given manifold, issubject to the original unity of consciousness,and that solely by means of the necessary rela-tion of the manifold in intuition to the "I think,"consequently by means of the pure synthesis ofthe understanding, which lies a priori at thefoundation of all empirical synthesis. The tran-scendental unity of apperception is alone objec-tively valid; the empirical which we do notconsider in this essay, and which is merely aunity deduced from the former under givenconditions in concreto, possesses only subjec-tive validity. One person connects the notionconveyed in a word with one thing, anotherwith another thing; and the unity of conscious-ness in that which is empirical, is, in relation to

that which is given by experience, not necessar-ily and universally valid.

SS 15. The Logical Form of all Judgements con-sists in the Objective Unity of Apperception of the Conceptionscontained therein.

I could never satisfy myself with the definitionwhich logicians give of a judgement. It is, ac-cording to them, the representation of a relationbetween two conceptions. I shall not dwell hereon the faultiness of this definition, in that itsuits only for categorical and not for hypotheti-cal or disjunctive judgements, these latter con-taining a relation not of conceptions but of jud-gements themselves— a blunder from whichmany evil results have followed.* It is moreimportant for our present purpose to observe,

that this definition does not determine in whatthe said relation consists.

[*Footnote: The tedious doctrine of the foursyllogistic figures concerns only categoricalsyllogisms; and although it is nothing morethan an artifice by surreptitiously introducingimmediate conclusions (consequentiae imme-diatae) among the premises of a pure syllo-gism, to give ism' give rise to an appearance ofmore modes of drawing a conclusion than thatin the first figure, the artifice would not havehad much success, had not its authors suc-ceeded in bringing categorical judgements intoexclusive respect, as those to which all othersmust be referred—a doctrine, however, which,according to SS 5, is utterly false.]

But if I investigate more closely the relation ofgiven cognitions in every judgement, and dis-tinguish it, as belonging to the understanding,from the relation which is produced accordingto laws of the reproductive imagination (which

has only subjective validity), I find that judge-ment is nothing but the mode of bringing givencognitions under the objective unit of appercep-tion. This is plain from our use of the term ofrelation is in judgements, in order to distin-guish the objective unity of given representa-tions from the subjective unity. For this termindicates the relation of these representations tothe original apperception, and also their neces-sary unity, even although the judgement is em-pirical, therefore contingent, as in the judge-ment: "All bodies are heavy." I do not mean bythis, that these representations do necessarilybelong to each other in empirical intuition, butthat by means of the necessary unity of appre-ciation they belong to each other in the synthe-sis of intuitions, that is to say, they belong toeach other according to principles of the objec-tive determination of all our representations, inso far as cognition can arise from them, theseprinciples being all deduced from the mainprinciple of the transcendental unity of apper-

ception. In this way alone can there arise fromthis relation a judgement, that is, a relationwhich has objective validity, and is perfectlydistinct from that relation of the very same rep-resentations which has only subjective valid-ity—a relation, to wit, which is produced ac-cording to laws of association. According tothese laws, I could only say: "When I hold inmy hand or carry a body, I feel an impressionof weight"; but I could not say: "It, the body, isheavy"; for this is tantamount to saying boththese representations are conjoined in the ob-ject, that is, without distinction as to the condi-tion of the subject, and do not merely standtogether in my perception, however frequentlythe perceptive act may be repeated.

SS 16. All Sensuous Intuitions are subject to theCategories, as Conditions under which alone

the manifold Content of them can be united inone Consciousness.

The manifold content given in a sensuous intui-tion comes necessarily under the original syn-thetical unity of apperception, because therebyalone is the unity of intuition possible (SS 13).But that act of the understanding, by which themanifold content of given representations(whether intuitions or conceptions) is broughtunder one apperception, is the logical functionof judgements (SS 15). All the manifold, there-fore, in so far as it is given in one empirical in-tuition, is determined in relation to one of thelogical functions of judgement, by means ofwhich it is brought into union in one con-sciousness. Now the categories are nothing elsethan these functions of judgement so far as themanifold in a given intuition is determined inrelation to them (SS 9). Consequently, the mani-fold in a given intuition is necessarily subject tothe categories of the understanding.

SS 17. Observation.

The manifold in an intuition, which I call mine,is represented by means of the synthesis of theunderstanding, as belonging to the necessaryunity of self-consciousness, and this takes placeby means of the category.* The category indi-cates accordingly that the empirical conscious-ness of a given manifold in an intuition is sub-ject to a pure self-consciousness a priori, in thesame manner as an empirical intuition is sub-ject to a pure sensuous intuition, which is also apriori. In the above proposition, then, lies thebeginning of a deduction of the pure concep-tions of the understanding. Now, as the catego-ries have their origin in the understanding alo-ne, independently of sensibility, I must in mydeduction make abstraction of the mode inwhich the manifold of an empirical intuition isgiven, in order to fix my attention exclusivelyon the unity which is brought by the under-standing into the intuition by means of the cate-

gory. In what follows (SS 22), it will be shown,from the mode in which the empirical intuitionis given in the faculty of sensibility, that theunity which belongs to it is no other than thatwhich the category (according to SS 16) im-poses on the manifold in a given intuition, andthus, its a priori validity in regard to all objectsof sense being established, the purpose of ourdeduction will be fully attained.

[*Footnote: The proof of this rests on the repre-sented unity of intuition, by means of which anobject is given, and which always includes initself a synthesis of the manifold to be intuited,and also the relation of this latter to unity ofapperception.]

But there is one thing in the above demonstra-tion of which I could not make abstraction, na-mely, that the manifold to be intuited must begiven previously to the synthesis of the under-standing, and independently of it. How thistakes place remains here undetermined. For if I

cogitate an understanding which was itself in-tuitive (as, for example, a divine understandingwhich should not represent given objects, butby whose representation the objects themselvesshould be given or produced), the categorieswould possess no significance in relation tosuch a faculty of cognition. They are merelyrules for an understanding, whose whole po-wer consists in thought, that is, in the act ofsubmitting the synthesis of the manifold whichis presented to it in intuition from a very differ-ent quarter, to the unity of apperception; a fac-ulty, therefore, which cognizes nothing per se,but only connects and arranges the material ofcognition, the intuition, namely, which must bepresented to it by means of the object. But toshow reasons for this peculiar character of ourunderstandings, that it produces unity of ap-perception a priori only by means of categories,and a certain kind and number thereof, is asimpossible as to explain why we are endowedwith precisely so many functions of judgement

and no more, or why time and space are theonly forms of our intuition.

SS 18. In Cognition, its Application to Objectsof Experience is the only legitimate use of theCategory.

To think an object and to cognize an object areby no means the same thing. In cognition thereare two elements: firstly, the conception, whe-reby an object is cogitated (the category); and,secondly, the intuition, whereby the object isgiven. For supposing that to the conception acorresponding intuition could not be given, itwould still be a thought as regards its form, butwithout any object, and no cognition of any-thing would be possible by means of it, inas-much as, so far as I knew, there existed andcould exist nothing to which my thought couldbe applied. Now all intuition possible to us is

sensuous; consequently, our thought of an ob-ject by means of a pure conception of the un-derstanding, can become cognition for us onlyin so far as this conception is applied to objectsof the senses. Sensuous intuition is either pureintuition (space and time) or empirical intui-tion—of that which is immediately representedin space and time by means of sensation as real.Through the determination of pure intuition weobtain a priori cognitions of objects, as in mat-hematics, but only as regards their form as phe-nomena; whether there can exist things whichmust be intuited in this form is not thereby es-tablished. All mathematical conceptions, there-fore, are not per se cognition, except in so far aswe presuppose that there exist things whichcan only be represented conformably to theform of our pure sensuous intuition. But thingsin space and time are given only in so far asthey are perceptions (representations accompa-nied with sensation), therefore only by empiri-cal representation. Consequently the pure con-

ceptions of the understanding, even when theyare applied to intuitions a priori (as in mathe-matics), produce cognition only in so far asthese (and therefore the conceptions of the un-derstanding by means of them) can be appliedto empirical intuitions. Consequently the cate-gories do not, even by means of pure intuitionafford us any cognition of things; they can onlydo so in so far as they can be applied to empiri-cal intuition. That is to say, the categories serveonly to render empirical cognition possible. Butthis is what we call experience. Consequently,in cognition, their application to objects of ex-perience is the only legitimate use of the cate-gories.

SS 19.

The foregoing proposition is of the utmost im-portance, for it determines the limits of the ex-

ercise of the pure conceptions of the under-standing in regard to objects, just as transcen-dental aesthetic determined the limits of theexercise of the pure form of our sensuous intui-tion. Space and time, as conditions of the possi-bility of the presentation of objects to us, arevalid no further than for objects of sense, con-sequently, only for experience. Beyond theselimits they represent to us nothing, for theybelong only to sense, and have no reality apartfrom it. The pure conceptions of the under-standing are free from this limitation, and ex-tend to objects of intuition in general, be theintuition like or unlike to ours, provided only itbe sensuous, and not intellectual. But this ex-tension of conceptions beyond the range of ourintuition is of no advantage; for they are thenmere empty conceptions of objects, as to thepossibility or impossibility of the existence ofwhich they furnish us with no means of discov-ery. They are mere forms of thought, withoutobjective reality, because we have no intuition

to which the synthetical unity of apperception,which alone the categories contain, could beapplied, for the purpose of determining an ob-ject. Our sensuous and empirical intuition canalone give them significance and meaning.

If, then, we suppose an object of a non-sensuous intuition to be given we can in thatcase represent it by all those predicates whichare implied in the presupposition that nothingappertaining to sensuous intuition belongs to it;for example, that it is not extended, or in space;that its duration is not time; that in it no change(the effect of the determinations in time) is to bemet with, and so on. But it is no proper knowl-edge if I merely indicate what the intuition ofthe object is not, without being able to say whatis contained in it, for I have not shown the pos-sibility of an object to which my pure concep-tion of understanding could be applicable, be-cause I have not been able to furnish any intui-tion corresponding to it, but am only able to

say that our intuition is not valid for it. But themost important point is this, that to a some-thing of this kind not one category can befound applicable. Take, for example, the con-ception of substance, that is, something that canexist as subject, but never as mere predicate; inregard to this conception I am quite ignorantwhether there can really be anything to corre-spond to such a determination of thought, ifempirical intuition did not afford me the occa-sion for its application. But of this more in thesequel.

SS 20. Of the Application of the Categories toObjects of the Senses in general.

The pure conceptions of the understandingapply to objects of intuition in general, throughthe understanding alone, whether the intuition

be our own or some other, provided only it besensuous, but are, for this very reason, mereforms of thought, by means of which alone nodetermined object can be cognized. The synthe-sis or conjunction of the manifold in these con-ceptions relates, we have said, only to the unityof apperception, and is for this reason theground of the possibility of a priori cognition,in so far as this cognition is dependent on theunderstanding. This synthesis is, therefore, notmerely transcendental, but also purely intellec-tual. But because a certain form of sensuousintuition exists in the mind a priori which restson the receptivity of the representative faculty(sensibility), the understanding, as a spontane-ity, is able to determine the internal sense bymeans of the diversity of given representations,conformably to the synthetical unity of apper-ception, and thus to cogitate the syntheticalunity of the apperception of the manifold ofsensuous intuition a priori, as the condition towhich must necessarily be submitted all objects

of human intuition. And in this manner thecategories as mere forms of thought receiveobjective reality, that is, application to objectswhich are given to us in intuition, but that onlyas phenomena, for it is only of phenomena thatwe are capable of a priori intuition.

This synthesis of the manifold of sensuous in-tuition, which is possible and necessary a pri-ori, may be called figurative (synthesisspeciosa), in contradistinction to that which iscogitated in the mere category in regard to themanifold of an intuition in general, and iscalled connection or conjunction of theunderstanding (synthesis intellectualis). Bothare transcendental, not merely because theythemselves precede a priori all experience, butalso because they form the basis for thepossibility of other cognition a priori.

But the figurative synthesis, when it has rela-tion only to the originally synthetical unity ofapperception, that is to the transcendental unity

cogitated in the categories, must, to be distin-guished from the purely intellectual conjunc-tion, be entitled the transcendental synthesis ofimagination. Imagination is the faculty of rep-resenting an object even without its presence inintuition. Now, as all our intuition is sensuous,imagination, by reason of the subjective condi-tion under which alone it can give a corre-sponding intuition to the conceptions of theunderstanding, belongs to sensibility. But in sofar as the synthesis of the imagination is an actof spontaneity, which is determinative, and not,like sense, merely determinable, and which isconsequently able to determine sense a priori,according to its form, conformably to the unityof apperception, in so far is the imagination afaculty of determining sensibility a priori, andits synthesis of intuitions according to the cate-gories must be the transcendental synthesis ofthe imagination. It is an operation of the under-standing on sensibility, and the first applicationof the understanding to objects of possible in-

tuition, and at the same time the basis for theexercise of the other functions of that faculty.As figurative, it is distinguished from the mere-ly intellectual synthesis, which is produced bythe understanding alone, without the aid ofimagination. Now, in so far as imagination isspontaneity, I sometimes call it also the produc-tive imagination, and distinguish it from thereproductive, the synthesis of which is subjectentirely to empirical laws, those of association,namely, and which, therefore, contributes noth-ing to the explanation of the possibility of apriori cognition, and for this reason belongs notto transcendental philosophy, but to psychol-ogy.

We have now arrived at the proper place forexplaining the paradox which must have struckevery one in our exposition of the internal sen-se (SS 6), namely—how this sense represents usto our own consciousness, only as we appear toourselves, not as we are in ourselves, because,

to wit, we intuite ourselves only as we are in-wardly affected. Now this appears to be con-tradictory, inasmuch as we thus stand in a pas-sive relation to ourselves; and therefore in thesystems of psychology, the internal sense iscommonly held to be one with the faculty ofapperception, while we, on the contrary, care-fully distinguish them.

That which determines the internal sense is theunderstanding, and its original power of con-joining the manifold of intuition, that is, ofbringing this under an apperception (uponwhich rests the possibility of the understandingitself). Now, as the human understanding is notin itself a faculty of intuition, and is unable toexercise such a power, in order to conjoin, as itwere, the manifold of its own intuition, the syn-thesis of understanding is, considered per se,nothing but the unity of action, of which, assuch, it is self-conscious, even apart from sensi-bility, by which, moreover, it is able to deter-

mine our internal sense in respect of the mani-fold which may be presented to it according tothe form of sensuous intuition. Thus, under thename of a transcendental synthesis of imagina-tion, the understanding exercises an activityupon the passive subject, whose faculty it is;and so we are right in saying that the internalsense is affected thereby. Apperception and itssynthetical unity are by no means one and thesame with the internal sense. The former, as thesource of all our synthetical conjunction, ap-plies, under the name of the categories, to themanifold of intuition in general, prior to allsensuous intuition of objects. The internal sen-se, on the contrary, contains merely the form ofintuition, but without any synthetical conjunc-tion of the manifold therein, and consequentlydoes not contain any determined intuition,which is possible only through consciousnessof the determination of the manifold by thetranscendental act of the imagination (syntheti-cal influence of the understanding on the inter-

nal sense), which I have named figurative syn-thesis.

This we can indeed always perceive in our-selves. We cannot cogitate a geometrical linewithout drawing it in thought, nor a circlewithout describing it, nor represent the threedimensions of space without drawing threelines from the same point perpendicular to oneanother. We cannot even cogitate time, unless,in drawing a straight line (which is to serve asthe external figurative representation of time),we fix our attention on the act of the synthesisof the manifold, whereby we determine succes-sively the internal sense, and thus attend also tothe succession of this determination. Motion asan act of the subject (not as a determination ofan object),* consequently the synthesis of themanifold in space, if we make abstraction ofspace and attend merely to the act by which wedetermine the internal sense according to itsform, is that which produces the conception of

succession. The understanding, therefore, doesby no means find in the internal sense any suchsynthesis of the manifold, but produces it, inthat it affects this sense. At the same time, how"I who think" is distinct from the "I" which intu-ites itself (other modes of intuition being cogi-table as at least possible), and yet one and thesame with this latter as the same subject; how,therefore, I am able to say: "I, as an intelligenceand thinking subject, cognize myself as an ob-ject thought, so far as I am, moreover, given tomyself in intuition—only, like other phenom-ena, not as I am in myself, and as considered bythe understanding, but merely as I appear"—isa question that has in it neither more nor lessdifficulty than the question—"How can I be anobject to myself?" or this—"How I can be anobject of my own intuition and internal percep-tions?" But that such must be the fact, if weadmit that space is merely a pure form of thephenomena of external sense, can be clearlyproved by the consideration that we cannot

represent time, which is not an object of exter-nal intuition, in any other way than under theimage of a line, which we draw in thought, amode of representation without which wecould not cognize the unity of its dimension,and also that we are necessitated to take ourdetermination of periods of time, or of points oftime, for all our internal perceptions from thechanges which we perceive in outward things.It follows that we must arrange the determina-tions of the internal sense, as phenomena intime, exactly in the same manner as we arrangethose of the external senses in space. And con-sequently, if we grant, respecting this latter,that by means of them we know objects only inso far as we are affected externally, we mustalso confess, with regard to the internal sense,that by means of it we intuite ourselves only aswe are internally affected by ourselves; in otherwords, as regards internal intuition, we cognizeour own subject only as phenomenon, and notas it is in itself.*[2]

[*Footnote: Motion of an object in space doesnot belong to a pure science, consequently notto geometry; because, that a thing is movablecannot be known a priori, but only from ex-perience. But motion, considered as thedescription of a space, is a pure act of thesuccessive synthesis of the manifold in externalintuition by means of productive imagination,and belongs not only to geometry, but even totranscendental philosophy.]

[*[2]Footnote: I do not see why so much diffi-culty should be found in admitting that ourinternal sense is affected by ourselves. Everyact of attention exemplifies it. In such an act theunderstanding determines the internal sense bythe synthetical conjunction which it cogitates,conformably to the internal intuition whichcorresponds to the manifold in the synthesis ofthe understanding. How much the mind isusually affected thereby every one will be ableto perceive in himself.]

SS 21.

On the other hand, in the transcendental syn-thesis of the manifold content of representa-tions, consequently in the synthetical unity ofapperception, I am conscious of myself, not as Iappear to myself, nor as I am in myself, butonly that "I am." This representation is athought, not an intuition. Now, as in order tocognize ourselves, in addition to the act ofthinking, which subjects the manifold of everypossible intuition to the unity of apperception,there is necessary a determinate mode of intui-tion, whereby this manifold is given; althoughmy own existence is certainly not mere phe-nomenon (much less mere illusion), the deter-mination of my existence* Can only take placeconformably to the form of the internal sense,according to the particular mode in which themanifold which I conjoin is given in internalintuition, and I have therefore no knowledge ofmyself as I am, but merely as I appear to my-

self. The consciousness of self is thus very farfrom a knowledge of self, in which I do not usethe categories, whereby I cogitate an object, bymeans of the conjunction of the manifold in oneapperception. In the same way as I require, forthe sake of the cognition of an object distinctfrom myself, not only the thought of an objectin general (in the category), but also an intui-tion by which to determine that general concep-tion, in the same way do I require, in order tothe cognition of myself, not only the conscious-ness of myself or the thought that I think my-self, but in addition an intuition of the manifoldin myself, by which to determine this thought.It is true that I exist as an intelligence which isconscious only of its faculty of conjunction orsynthesis, but subjected in relation to the mani-fold which this intelligence has to conjoin to alimitative conjunction called the internal sense.My intelligence (that is, I) can render that con-junction or synthesis perceptible only accordingto the relations of time, which are quite beyond

the proper sphere of the conceptions of the un-derstanding and consequently cognize itself inrespect to an intuition (which cannot possiblybe intellectual, nor given by the understand-ing), only as it appears to itself, and not as itwould cognize itself, if its intuition were intel-lectual.

[*Footnote: The "I think" expresses the act ofdetermining my own existence. My existence isthus already given by the act of consciousness;but the mode in which I must determine myexistence, that is, the mode in which I mustplace the manifold belonging to my existence,is not thereby given. For this purpose intuitionof self is required, and this intuition possesses aform given a priori, namely, time, which is sen-suous, and belongs to our receptivity of thedeterminable. Now, as I do not possess anotherintuition of self which gives the determining inme (of the spontaneity of which I am con-

scious), prior to the act of determination, in thesame manner as time gives the determinable, itis clear that I am unable to determine my ownexistence as that of a spontaneous being, but Iam only able to represent to myself the sponta-neity of my thought, that is, of my determina-tion, and my existence remains ever determin-able in a purely sensuous manner, that is to say,like the existence of a phenomenon. But it isbecause of this spontaneity that I call myself anintelligence.]

SS 22. Transcendental Deduction of the univer-sally possible employment in experience of thePure Conceptions of the Understanding.

In the metaphysical deduction, the a priori ori-gin of categories was proved by their completeaccordance with the general logical of thought;in the transcendental deduction was exhibited

the possibility of the categories as a priori cog-nitions of objects of an intuition in general (SS16 and 17).At present we are about to explainthe possibility of cognizing, a priori, by meansof the categories, all objects which can possiblybe presented to our senses, not, indeed, accord-ing to the form of their intuition, but accordingto the laws of their conjunction or synthesis,and thus, as it were, of prescribing laws to na-ture and even of rendering nature possible. Forif the categories were inadequate to this task, itwould not be evident to us why everything thatis presented to our senses must be subject tothose laws which have an a priori origin in theunderstanding itself.

I premise that by the term synthesis of appre-hension I understand the combination of themanifold in an empirical intuition, wherebyperception, that is, empirical consciousness ofthe intuition (as phenomenon), is possible.

We have a priori forms of the external and in-ternal sensuous intuition in the representationsof space and time, and to these must the syn-thesis of apprehension of the manifold in a phe-nomenon be always comformable, because thesynthesis itself can only take place according tothese forms. But space and time are not merelyforms of sensuous intuition, but intuitionsthemselves (which contain a manifold), andtherefore contain a priori the determination ofthe unity of this manifold.* (See the Transcen-dent Aesthetic.) Therefore is unity of the syn-thesis of the manifold without or within us,consequently also a conjunction to which allthat is to be represented as determined in spaceor time must correspond, given a priori alongwith (not in) these intuitions, as the conditionof the synthesis of all apprehension of them.But this synthetical unity can be no other thanthat of the conjunction of the manifold of a gi-ven intuition in general, in a primitive act ofconsciousness, according to the categories, but

applied to our sensuous intuition. Conse-quently all synthesis, whereby alone is evenperception possible, is subject to the categories.And, as experience is cognition by means ofconjoined perceptions, the categories are condi-tions of the possibility of experience and aretherefore valid a priori for all objects of experi-ence.

[*Footnote: Space represented as an object (asgeometry really requires it to be) contains morethan the mere form of the intuition; namely, acombination of the manifold given according tothe form of sensibility into a representation thatcan be intuited; so that the form of the intuitiongives us merely the manifold, but the formalintuition gives unity of representation. In theaesthetic, I regarded this unity as belongingentirely to sensibility, for the purpose of indi-cating that it antecedes all conceptions, al-though it presupposes a synthesis which doesnot belong to sense, through which alone, how-

ever, all our conceptions of space and time arepossible. For as by means of this unity alone(the understanding determining the sensibility)space and time are given as intuitions, it fol-lows that the unity of this intuition a priori be-longs to space and time, and not to the concep-tion of the understanding (SS 20).]

When, then, for example, I make the empiricalintuition of a house by apprehension of themanifold contained therein into a perception,the necessary unity of space and of my externalsensuous intuition lies at the foundation of thisact, and I, as it were, draw the form of the hou-se conformably to this synthetical unity of themanifold in space. But this very syntheticalunity remains, even when I abstract the form ofspace, and has its seat in the understanding,and is in fact the category of the synthesis ofthe homogeneous in an intuition; that is to say,the category of quantity, to which the aforesaid

synthesis of apprehension, that is, the percep-tion, must be completely conformable.*

[*Footnote: In this manner it is proved, that thesynthesis of apprehension, which is empirical,must necessarily be conformable to the synthe-sis of apperception, which is intellectual, andcontained a priori in the category. It is one andthe same spontaneity which at one time, underthe name of imagination, at another under thatof understanding, produces conjunction in themanifold of intuition.]

To take another example, when I perceive thefreezing of water, I apprehend two states (fluid-ity and solidity), which, as such, stand towardeach other mutually in a relation of time. But inthe time, which I place as an internal intuition,at the foundation of this phenomenon, I repre-sent to myself synthetical unity of the manifold,without which the aforesaid relation could notbe given in an intuition as determined (in re-gard to the succession of time). Now this syn-

thetical unity, as the a priori condition underwhich I conjoin the manifold of an intuition, is,if I make abstraction of the permanent form ofmy internal intuition (that is to say, of time), thecategory of cause, by means of which, whenapplied to my sensibility, I determine every-thing that occurs according to relations of time.Consequently apprehension in such an event,and the event itself, as far as regards the possi-bility of its perception, stands under the con-ception of the relation of cause and effect: andso in all other cases.

Categories are conceptions which prescribelaws a priori to phenomena, consequently tonature as the complex of all phenomena (naturamaterialiter spectata). And now the questionarises— inasmuch as these categories are notderived from nature, and do not regulate them-selves according to her as their model (for inthat case they would be empirical)—how it isconceivable that nature must regulate herself

according to them, in other words, how thecategories can determine a priori the synthesisof the manifold of nature, and yet not derivetheir origin from her. The following is the solu-tion of this enigma.

It is not in the least more difficult to conceivehow the laws of the phenomena of nature mustharmonize with the understanding and with itsa priori form—that is, its faculty of conjoiningthe manifold—than it is to understand how thephenomena themselves must correspond withthe a priori form of our sensuous intuition. Forlaws do not exist in the phenomena any morethan the phenomena exist as things in them-selves. Laws do not exist except by relation tothe subject in which the phenomena inhere, inso far as it possesses understanding, just asphenomena have no existence except by rela-tion to the same existing subject in so far as ithas senses. To things as things in themselves,conformability to law must necessarily belong

independently of an understanding to cognizethem. But phenomena are only representationsof things which are utterly unknown in respectto what they are in themselves. But as mererepresentations, they stand under no law ofconjunction except that which the conjoiningfaculty prescribes. Now that which conjoins themanifold of sensuous intuition is imagination, amental act to which understanding contributesunity of intellectual synthesis, and sensibility,manifoldness of apprehension. Now as all pos-sible perception depends on the synthesis ofapprehension, and this empirical synthesis it-self on the transcendental, consequently on thecategories, it is evident that all possible percep-tions, and therefore everything that can attainto empirical consciousness, that is, all phenom-ena of nature, must, as regards their conjunc-tion, be subject to the categories. And nature(considered merely as nature in general) is de-pendent on them, as the original ground of hernecessary conformability to law (as natura for-

maliter spectata). But the pure faculty (of theunderstanding) of prescribing laws a priori tophenomena by means of mere categories, is notcompetent to enounce other or more laws thanthose on which a nature in general, as a con-formability to law of phenomena of space andtime, depends. Particular laws, inasmuch asthey concern empirically determined phenom-ena, cannot be entirely deduced from purelaws, although they all stand under them. Ex-perience must be superadded in order to knowthese particular laws; but in regard to experi-ence in general, and everything that can be cog-nized as an object thereof, these a priori lawsare our only rule and guide.

SS 23. Result of this Deduction of the Concep-tions of the Understanding.

We cannot think any object except by means ofthe categories; we cannot cognize any thoughtexcept by means of intuitions corresponding tothese conceptions. Now all our intuitions aresensuous, and our cognition, in so far as theobject of it is given, is empirical. But empiricalcognition is experience; consequently no a pri-ori cognition is possible for us, except of objectsof possible experience.*

[Footnote: Lest my readers should stumble atthis assertion, and the conclusions that may betoo rashly drawn from it, I must remind themthat the categories in the act of thought are byno means limited by the conditions of our sen-suous intuition, but have an unbounded sphereof action. It is only the cognition of the object ofthought, the determining of the object, whichrequires intuition. In the absence of intuition,our thought of an object may still have true anduseful consequences in regard to the exercise ofreason by the subject. But as this exercise of

reason is not always directed on the determina-tion of the object, in other words, on cognitionthereof, but also on the determination of thesubject and its volition, I do not intend to treatof it in this place.]

But this cognition, which is limited to objects ofexperience, is not for that reason derived en-tirely, from, experience, but—and this is as-serted of the pure intuitions and the pure con-ceptions of the understanding—there are, un-questionably, elements of cognition, which ex-ist in the mind a priori. Now there are only twoways in which a necessary harmony of experi-ence with the conceptions of its objects can becogitated. Either experience makes these con-ceptions possible, or the conceptions make ex-perience possible. The former of these state-ments will not bold good with respect to thecategories (nor in regard to pure sensuous in-tuition), for they are a priori conceptions, andtherefore independent of experience. The asser-

tion of an empirical origin would attribute tothem a sort of generatio aequivoca. Conse-quently, nothing remains but to adopt the sec-ond alternative (which presents us with a sys-tem, as it were, of the epigenesis of pure rea-son), namely, that on the part of the under-standing the categories do contain the groundsof the possibility of all experience. But withrespect to the questions how they make experi-ence possible, and what are the principles ofthe possibility thereof with which they presentus in their application to phenomena, the fol-lowing section on the transcendental exercise ofthe faculty of judgement will inform the reader.

It is quite possible that someone may propose aspecies of preformation-system of pure rea-son—a middle way between the two—to wit,that the categories are neither innate and first apriori principles of cognition, nor derived fromexperience, but are merely subjective aptitudesfor thought implanted in us contemporane-

ously with our existence, which were so or-dered and disposed by our Creator, that theirexercise perfectly harmonizes with the laws ofnature which regulate experience. Now, not tomention that with such an hypothesis it is im-possible to say at what point we must stop inthe employment of predetermined aptitudes,the fact that the categories would in this caseentirely lose that character of necessity which isessentially involved in the very conception ofthem, is a conclusive objection to it. The con-ception of cause, for example, which expressesthe necessity of an effect under a presupposedcondition, would be false, if it rested only uponsuch an arbitrary subjective necessity of unitingcertain empirical representations according tosuch a rule of relation. I could not then say—"The effect is connected with its cause in theobject (that is, necessarily)," but only, "I am soconstituted that I can think this representationas so connected, and not otherwise." Now thisis just what the sceptic wants. For in this case,

all our knowledge, depending on the supposedobjective validity of our judgement, is nothingbut mere illusion; nor would there be wantingpeople who would deny any such subjectivenecessity in respect to themselves, though theymust feel it. At all events, we could not disputewith any one on that which merely depends onthe manner in which his subject is organized.

Short view of the above Deduction.

The foregoing deduction is an exposition of thepure conceptions of the understanding (andwith them of all theoretical a priori cognition),as principles of the possibility of experience,but of experience as the determination of allphenomena in space and time in general—ofexperience, finally, from the principle of theoriginal synthetical unity of apperception, as

the form of the understanding in relation totime and space as original forms of sensibility.

I consider the division by paragraphs to be nec-essary only up to this point, because we had totreat of the elementary conceptions. As we nowproceed to the exposition of the employment ofthese, I shall not designate the chapters in thismanner any further.

BOOK II.

Analytic of Principles.

General logic is constructed upon a plan whichcoincides exactly with the division of the higherfaculties of cognition. These are, understand-ing, judgement, and reason. This science, ac-cordingly, treats in its analytic of conceptions,

judgements, and conclusions in exact corre-spondence with the functions and order ofthose mental powers which we include gener-ally under the generic denomination of under-standing.

As this merely formal logic makes abstractionof all content of cognition, whether pure orempirical, and occupies itself with the mereform of thought (discursive cognition), it mustcontain in its analytic a canon for reason. Forthe form of reason has its law, which, withouttaking into consideration the particular natureof the cognition about which it is employed,can be discovered a priori, by the simple analy-sis of the action of reason into its momenta.

Transcendental logic, limited as it is to a deter-minate content, that of pure a priori cognitions,to wit, cannot imitate general logic in this divi-sion. For it is evident that the transcendentalemployment of reason is not objectively valid,and therefore does not belong to the logic of

truth (that is, to analytic), but as a logic of illu-sion, occupies a particular department in thescholastic system under the name of transcen-dental dialectic.

Understanding and judgement accordinglypossess in transcendental logic a canon of ob-jectively valid, and therefore true exercise, andare comprehended in the analytical departmentof that logic. But reason, in her endeavours toarrive by a priori means at some true statementconcerning objects and to extend cognition be-yond the bounds of possible experience, is alto-gether dialectic, and her illusory assertions can-not be constructed into a canon such as an ana-lytic ought to contain.

Accordingly, the analytic of principles will bemerely a canon for the faculty of judgement, forthe instruction of this faculty in its applicationto phenomena of the pure conceptions of theunderstanding, which contain the necessarycondition for the establishment of a priori laws.

On this account, although the subject of thefollowing chapters is the especial principles ofunderstanding, I shall make use of the termDoctrine of the faculty of judgement, in orderto define more particularly my present purpo-se.

INTRODUCTION. Of the Transcendental Fa-culty of judgement in General.

If understanding in general be defined as thefaculty of laws or rules, the faculty of judge-ment may be termed the faculty of subsump-tion under these rules; that is, of distinguishingwhether this or that does or does not stand un-der a given rule (casus datae legis). Generallogic contains no directions or precepts for thefaculty of judgement, nor can it contain anysuch. For as it makes abstraction of all contentof cognition, no duty is left for it, except that of

exposing analytically the mere form of cogni-tion in conceptions, judgements, and conclu-sions, and of thereby establishing formal rulesfor all exercise of the understanding. Now ifthis logic wished to give some general directionhow we should subsume under these rules,that is, how we should distinguish whether thisor that did or did not stand under them, thisagain could not be done otherwise than bymeans of a rule. But this rule, precisely becauseit is a rule, requires for itself direction from thefaculty of judgement. Thus, it is evident thatthe understanding is capable of being instruc-ted by rules, but that the judgement is a pecu-liar talent, which does not, and cannot requiretuition, but only exercise. This faculty is there-fore the specific quality of the so-called motherwit, the want of which no scholastic disciplinecan compensate.

For although education may furnish, and, as itwere, engraft upon a limited understanding

rules borrowed from other minds, yet the po-wer of employing these rules correctly mustbelong to the pupil himself; and no rule whichwe can prescribe to him with this purpose is, inthe absence or deficiency of this gift of nature,secure from misuse.* A physician therefore, ajudge or a statesman, may have in his headmany admirable pathological, juridical, or poli-tical rules, in a degree that may enable him tobe a profound teacher in his particular science,and yet in the application of these rules he mayvery possibly blunder—either because he iswanting in natural judgement (though not inunderstanding) and, whilst he can comprehendthe general in abstracto, cannot distinguishwhether a particular case in concreto ought torank under the former; or because his faculty ofjudgement has not been sufficiently exercisedby examples and real practice. Indeed, thegrand and only use of examples, is to sharpenthe judgement. For as regards the correctnessand precision of the insight of the understan-

ding, examples are commonly injurious ratherthan otherwise, because, as casus in terministhey seldom adequately fulfil the conditions ofthe rule. Besides, they often weaken the powerof our understanding to apprehend rules orlaws in their universality, independently ofparticular circumstances of experience; andhence, accustom us to employ them more asformulae than as principles. Examples are thusthe go-cart of the judgement, which he who isnaturally deficient in that faculty cannot affordto dispense with.

[*Footnote: Deficiency in judgement is properlythat which is called stupidity; and for such afailing we know no remedy. A dull or narrow-minded person, to whom nothing is wantingbut a proper degree of understanding, may beimproved by tuition, even so far as to deservethe epithet of learned. But as such persons fre-quently labour under a deficiency in the facultyof judgement, it is not uncommon to find men

extremely learned who in the application oftheir science betray a lamentable degree thisirremediable want.]

But although general logic cannot give direc-tions to the faculty of judgement, the case isvery different as regards transcendental logic,insomuch that it appears to be the especial dutyof the latter to secure and direct, by means ofdeterminate rules, the faculty of judgement inthe employment of the pure understanding.For, as a doctrine, that is, as an endeavour toenlarge the sphere of the understanding in re-gard to pure a priori cognitions, philosophy isworse than useless, since from all the attemptshitherto made, little or no ground has been gai-ned. But, as a critique, in order to guard againstthe mistakes of the faculty of judgement (lapsusjudicii) in the employment of the few pure con-ceptions of the understanding which we pos-sess, although its use is in this case purely ne-

gative, philosophy is called upon to apply allits acuteness and penetration.

But transcendental philosophy has this peculia-rity, that besides indicating the rule, or ratherthe general condition for rules, which is givenin the pure conception of the understanding, itcan, at the same time, indicate a priori the caseto which the rule must be applied. The cause ofthe superiority which, in this respect, transcen-dental philosophy possesses above all othersciences except mathematics, lies in this: ittreats of conceptions which must relate a priorito their objects, whose objective validity conse-quently cannot be demonstrated a posteriori,and is, at the same time, under the obligation ofpresenting in general but sufficient tests, theconditions under which objects can be given inharmony with those conceptions; otherwisethey would be mere logical forms, without con-tent, and not pure conceptions of the unders-tanding.

Our transcendental doctrine of the faculty ofjudgement will contain two chapters. The firstwill treat of the sensuous condition underwhich alone pure conceptions of the unders-tanding can be employed— that is, of theschematism of the pure understanding. Thesecond will treat of those synthetical judge-ments which are derived a priori from pureconceptions of the understanding under thoseconditions, and which lie a priori at the founda-tion of all other cognitions, that is to say, it willtreat of the principles of the pure understan-ding.

TRANSCENDENTAL DOCTRINE OFTHE FACULTY OF JUDGEMENT OR, ANA-LYTIC OF PRINCIPLES.

CHAPTER I. Of the Schematism at of thePure Conceptions of the Understanding.

In all subsumptions of an object under a con-ception, the representation of the object must behomogeneous with the conception; in otherwords, the conception must contain that whichis represented in the object to be subsumedunder it. For this is the meaning of the expres-sion: "An object is contained under a concep-tion." Thus the empirical conception of a plateis homogeneous with the pure geometrical con-ception of a circle, inasmuch as the roundnesswhich is cogitated in the former is intuited inthe latter.

But pure conceptions of the understanding,when compared with empirical intuitions, or

even with sensuous intuitions in general, arequite heterogeneous, and never can be discove-red in any intuition. How then is the subsump-tion of the latter under the former, and conse-quently the application of the categories to phe-nomena, possible?—For it is impossible to say,for example: "Causality can be intuited throughthe senses and is contained in the phenome-non."—This natural and important questionforms the real cause of the necessity of a trans-cendental doctrine of the faculty of judgement,with the purpose, to wit, of showing how pureconceptions of the understanding can be ap-plied to phenomena. In all other sciences, whe-re the conceptions by which the object isthought in the general are not so different andheterogeneous from those which represent theobject in concreto—as it is given, it is quite un-necessary to institute any special inquiries con-cerning the application of the former to the lat-ter.

Now it is quite clear that there must be somethird thing, which on the one side is homoge-neous with the category, and with the pheno-menon on the other, and so makes the applica-tion of the former to the latter possible. Thismediating representation must be pure (wit-hout any empirical content), and yet must onthe one side be intellectual, on the other sen-suous. Such a representation is the transcen-dental schema.

The conception of the understanding containspure synthetical unity of the manifold in gene-ral. Time, as the formal condition of the mani-fold of the internal sense, consequently of theconjunction of all representations, contains apriori a manifold in the pure intuition. Now atranscendental determination of time is so farhomogeneous with the category, which consti-tutes the unity thereof, that it is universal andrests upon a rule a priori. On the other hand, itis so far homogeneous with the phenomenon,

inasmuch as time is contained in every empiri-cal representation of the manifold. Thus an ap-plication of the category to phenomena beco-mes possible, by means of the transcendentaldetermination of time, which, as the schema ofthe conceptions of the understanding, mediatesthe subsumption of the latter under the former.

After what has been proved in our deduction ofthe categories, no one, it is to be hoped, canhesitate as to the proper decision of the ques-tion, whether the employment of these pureconceptions of the understanding ought to bemerely empirical or also transcendental; in ot-her words, whether the categories, as condi-tions of a possible experience, relate a priorisolely to phenomena, or whether, as conditionsof the possibility of things in general, their ap-plication can be extended to objects as things inthemselves. For we have there seen that con-ceptions are quite impossible, and utterly wit-hout signification, unless either to them, or at

least to the elements of which they consist, anobject be given; and that, consequently, theycannot possibly apply to objects as things inthemselves without regard to the questionwhether and how these may be given to us;and, further, that the only manner in whichobjects can be given to us is by means of themodification of our sensibility; and, finally, thatpure a priori conceptions, in addition to thefunction of the understanding in the category,must contain a priori formal conditions of sen-sibility (of the internal sense, namely), whichagain contain the general condition underwhich alone the category can be applied to anyobject. This formal and pure condition of sensi-bility, to which the conception of the unders-tanding is restricted in its employment, weshall name the schema of the conception of theunderstanding, and the procedure of the un-derstanding with these schemata we shall callthe schematism of the pure understanding.

The schema is, in itself, always a mere productof the imagination. But, as the synthesis of ima-gination has for its aim no single intuition, butmerely unity in the determination of sensibility,the schema is clearly distinguishable from theimage. Thus, if I place five points one afteranother …. this is an image of the number five.On the other hand, if I only think a number ingeneral, which may be either five or a hundred,this thought is rather the representation of amethod of representing in an image a sum (e.g.,a thousand) in conformity with a conception,than the image itself, an image which I shouldfind some little difficulty in reviewing, andcomparing with the conception. Now this re-presentation of a general procedure of the ima-gination to present its image to a conception, Icall the schema of this conception.

In truth, it is not images of objects, but schema-ta, which lie at the foundation of our pure sen-suous conceptions. No image could ever be

adequate to our conception of a triangle in ge-neral. For the generalness of the conception itnever could attain to, as this includes underitself all triangles, whether right-angled, acute-angled, etc., whilst the image would always belimited to a single part of this sphere. Theschema of the triangle can exist nowhere elsethan in thought, and it indicates a rule of thesynthesis of the imagination in regard to purefigures in space. Still less is an object of expe-rience, or an image of the object, ever to theempirical conception. On the contrary, the con-ception always relates immediately to theschema of the imagination, as a rule for thedetermination of our intuition, in conformitywith a certain general conception. The concep-tion of a dog indicates a rule, according towhich my imagination can delineate the figureof a four-footed animal in general, withoutbeing limited to any particular individual formwhich experience presents to me, or indeed toany possible image that I can represent to my-

self in concreto. This schematism of our unders-tanding in regard to phenomena and their mereform, is an art, hidden in the depths of thehuman soul, whose true modes of action weshall only with difficulty discover and unveil.Thus much only can we say: "The image is aproduct of the empirical faculty of the produc-tive imagination—the schema of sensuous con-ceptions (of figures in space, for example) is aproduct, and, as it were, a monogram of thepure imagination a priori, whereby and accor-ding to which images first become possible,which, however, can be connected with theconception only mediately by means of theschema which they indicate, and are in them-selves never fully adequate to it." On the otherhand, the schema of a pure conception of theunderstanding is something that cannot be re-duced into any image—it is nothing else thanthe pure synthesis expressed by the category,conformably, to a rule of unity according toconceptions. It is a transcendental product of

the imagination, a product which concerns thedetermination of the internal sense, accordingto conditions of its form (time) in respect to allrepresentations, in so far as these representa-tions must be conjoined a priori in one concep-tion, conformably to the unity of apperception.

Without entering upon a dry and tedious ana-lysis of the essential requisites of transcenden-tal schemata of the pure conceptions of the un-derstanding, we shall rather proceed at once togive an explanation of them according to theorder of the categories, and in connection the-rewith.

For the external sense the pure image of allquantities (quantorum) is space; the pure imageof all objects of sense in general, is time. But thepure schema of quantity (quantitatis) as a con-ception of the understanding, is number, a re-presentation which comprehends the successi-ve addition of one to one (homogeneous quan-tities). Thus, number is nothing else than the

unity of the synthesis of the manifold in ahomogeneous intuition, by means of my gene-rating time itself in my apprehension of theintuition.

Reality, in the pure conception of the unders-tanding, is that which corresponds to a sensa-tion in general; that, consequently, the concep-tion of which indicates a being (in time). Nega-tion is that the conception of which represents anot-being (in time). The opposition of these twoconsists therefore in the difference of one andthe same time, as a time filled or a time empty.Now as time is only the form of intuition, con-sequently of objects as phenomena, that whichin objects corresponds to sensation is the trans-cendental matter of all objects as things inthemselves (Sachheit, reality). Now every sen-sation has a degree or quantity by which it canfill time, that is to say, the internal sense in res-pect of the representation of an object, more orless, until it vanishes into nothing (= 0 = nega-

tio). Thus there is a relation and connectionbetween reality and negation, or rather a transi-tion from the former to the latter, which makesevery reality representable to us as a quantum;and the schema of a reality as the quantity ofsomething in so far as it fills time, is exactly thiscontinuous and uniform generation of the reali-ty in time, as we descend in time from the sen-sation which has a certain degree, down to thevanishing thereof, or gradually ascend fromnegation to the quantity thereof.

The schema of substance is the permanence ofthe real in time; that is, the representation of itas a substratum of the empirical determinationof time; a substratum which therefore remains,whilst all else changes. (Time passes not, but init passes the existence of the changeable. Totime, therefore, which is itself unchangeableand permanent, corresponds that which in thephenomenon is unchangeable in existence, thatis, substance, and it is only by it that the succes-

sion and coexistence of phenomena can be de-termined in regard to time.)

The schema of cause and of the causality of athing is the real which, when posited, is alwaysfollowed by something else. It consists, therefo-re, in the succession of the manifold, in so far asthat succession is subjected to a rule.

The schema of community (reciprocity of actionand reaction), or the reciprocal causality ofsubstances in respect of their accidents, is thecoexistence of the determinations of the onewith those of the other, according to a generalrule.

The schema of possibility is the accordance ofthe synthesis of different representations withthe conditions of time in general (as, for exam-ple, opposites cannot exist together at the sametime in the same thing, but only after each ot-her), and is therefore the determination of therepresentation of a thing at any time.

The schema of reality is existence in a determi-ned time.

The schema of necessity is the existence of anobject in all time.

It is clear, from all this, that the schema of thecategory of quantity contains and representsthe generation (synthesis) of time itself, in thesuccessive apprehension of an object; the sche-ma of quality the synthesis of sensation withthe representation of time, or the filling up oftime; the schema of relation the relation of per-ceptions to each other in all time (that is, accor-ding to a rule of the determination of time): andfinally, the schema of modality and its catego-ries, time itself, as the correlative of the deter-mination of an object—whether it does belongto time, and how. The schemata, therefore, arenothing but a priori determinations of timeaccording to rules, and these, in regard to allpossible objects, following the arrangement ofthe categories, relate to the series in time, the

content in time, the order in time, and finally,to the complex or totality in time.

Hence it is apparent that the schematism of theunderstanding, by means of the transcendentalsynthesis of the imagination, amounts to not-hing else than the unity of the manifold of in-tuition in the internal sense, and thus indirectlyto the unity of apperception, as a function co-rresponding to the internal sense (a receptivi-ty). Thus, the schemata of the pure conceptionsof the understanding are the true and onlyconditions whereby our understanding receivesan application to objects, and consequently sig-nificance. Finally, therefore, the categories areonly capable of empirical use, inasmuch as theyserve merely to subject phenomena to the uni-versal rules of synthesis, by means of an a prio-ri necessary unity (on account of the necessaryunion of all consciousness in one original ap-perception); and so to render them susceptibleof a complete connection in one experience. But

within this whole of possible experience lie allour cognitions, and in the universal relation tothis experience consists transcendental truth,which antecedes all empirical truth, and ren-ders the latter possible.

It is, however, evident at first sight, that alt-hough the schemata of sensibility are the soleagents in realizing the categories, they do, ne-vertheless, also restrict them, that is, they limitthe categories by conditions which lie beyondthe sphere of understanding— namely, in sen-sibility. Hence the schema is properly only thephenomenon, or the sensuous conception of anobject in harmony with the category. (Numerusest quantitas phaenomenon—sensatio realitasphaenomenon; constans et perdurabile rerumsubstantia phaenomenon— aeternitas, necessi-tas, phaenomena, etc.) Now, if we remove arestrictive condition, we thereby amplify, itappears, the formerly limited conception. Inthis way, the categories in their pure significa-

tion, free from all conditions of sensibility,ought to be valid of things as they are, and not,as the schemata represent them, merely as theyappear; and consequently the categories musthave a significance far more extended, andwholly independent of all schemata. In truth,there does always remain to the pure concep-tions of the understanding, after abstractingevery sensuous condition, a value and signifi-cance, which is, however, merely logical. But inthis case, no object is given them, and thereforethey have no meaning sufficient to afford us aconception of an object. The notion of substan-ce, for example, if we leave out the sensuousdetermination of permanence, would meannothing more than a something which can becogitated as subject, without the possibility ofbecoming a predicate to anything else. Of thisrepresentation I can make nothing, inasmuch asit does not indicate to me what determinationsthe thing possesses which must thus be valid aspremier subject. Consequently, the categories,

without schemata are merely functions of theunderstanding for the production of concep-tions, but do not represent any object. This sig-nificance they derive from sensibility, which atthe same time realizes the understanding andrestricts it.

CHAPTER II. System of all Principles ofthe Pure Understanding.

In the foregoing chapter we have merely consi-dered the general conditions under which alo-ne the transcendental faculty of judgement isjustified in using the pure conceptions of theunderstanding for synthetical judgements. Ourduty at present is to exhibit in systematic con-nection those judgements which the unders-tanding really produces a priori. For this pur-pose, our table of the categories will certainlyafford us the natural and safe guidance. For it is

precisely the categories whose application topossible experience must constitute all pure apriori cognition of the understanding; and therelation of which to sensibility will, on that ve-ry account, present us with a complete and sys-tematic catalogue of all the transcendental prin-ciples of the use of the understanding.

Principles a priori are so called, not merely be-cause they contain in themselves the groundsof other judgements, but also because theythemselves are not grounded in higher andmore general cognitions. This peculiarity,however, does not raise them altogether abovethe need of a proof. For although there could befound no higher cognition, and therefore noobjective proof, and although such a principlerather serves as the foundation for all cognitionof the object, this by no means hinders us fromdrawing a proof from the subjective sources ofthe possibility of the cognition of an object.Such a proof is necessary, moreover, because

without it the principle might be liable to theimputation of being a mere gratuitous asser-tion.

In the second place, we shall limit our investi-gations to those principles which relate to thecategories. For as to the principles of transcen-dental aesthetic, according to which space andtime are the conditions of the possibility ofthings as phenomena, as also the restriction ofthese principles, namely, that they cannot beapplied to objects as things in themselves—these, of course, do not fall within the scope ofour present inquiry. In like manner, the princi-ples of mathematical science form no part ofthis system, because they are all drawn fromintuition, and not from the pure conception ofthe understanding. The possibility of theseprinciples, however, will necessarily be consi-dered here, inasmuch as they are syntheticaljudgements a priori, not indeed for the purposeof proving their accuracy and apodeictic cer-

tainty, which is unnecessary, but merely torender conceivable and deduce the possibilityof such evident a priori cognitions.

But we shall have also to speak of the principleof analytical judgements, in opposition to synt-hetical judgements, which is the proper subjectof our inquiries, because this very oppositionwill free the theory of the latter from all ambi-guity, and place it clearly before our eyes in itstrue nature.

SYSTEM OF THE PRINCIPLES OF THEPURE UNDERSTANDING.

SECTION I. Of the Supreme Principle of allAnalytical Judgements.

Whatever may be the content of our cognition,and in whatever manner our cognition may be

related to its object, the universal, althoughonly negative conditions of all our judgementsis that they do not contradict themselves; ot-herwise these judgements are in themselves(even without respect to the object) nothing.But although there may exist no contradictionin our judgement, it may nevertheless connectconceptions in such a manner that they do notcorrespond to the object, or without anygrounds either a priori or a posteriori for arri-ving at such a judgement, and thus, withoutbeing self-contradictory, a judgement may ne-vertheless be either false or groundless.

Now, the proposition: "No subject can have apredicate that contradicts it," is called the prin-ciple of contradiction, and is a universal butpurely negative criterion of all truth. But it be-longs to logic alone, because it is valid of cogni-tions, merely as cognitions and without respectto their content, and declares that the contradic-tion entirely nullifies them. We can also, howe-

ver, make a positive use of this principle, thatis, not merely to banish falsehood and error (inso far as it rests upon contradiction), but alsofor the cognition of truth. For if the judgementis analytical, be it affirmative or negative, itstruth must always be recognizable by means ofthe principle of contradiction. For the contraryof that which lies and is cogitated as conceptionin the cognition of the object will be alwaysproperly negatived, but the conception itselfmust always be affirmed of the object, inas-much as the contrary thereof would be in con-tradiction to the object.

We must therefore hold the principle of contra-diction to be the universal and fully sufficientPrinciple of all analytical cognition. But as asufficient criterion of truth, it has no furtherutility or authority. For the fact that no cogni-tion can be at variance with this principle wit-hout nullifying itself, constitutes this principlethe sine qua non, but not the determining

ground of the truth of our cognition. As ourbusiness at present is properly with the synt-hetical part of our knowledge only, we shallalways be on our guard not to transgress thisinviolable principle; but at the same time not toexpect from it any direct assistance in the esta-blishment of the truth of any synthetical propo-sition.

There exists, however, a formula of this celebra-ted principle—a principle merely formal andentirely without content—which contains asynthesis that has been inadvertently and quiteunnecessarily mixed up with it. It is this: "It isimpossible for a thing to be and not to be at thesame time." Not to mention the superfluous-ness of the addition of the word impossible toindicate the apodeictic certainty, which oughtto be self-evident from the proposition itself,the proposition is affected by the condition oftime, and as it were says: "A thing = A, which issomething = B, cannot at the same time be non-

B." But both, B as well as non-B, may quite wellexist in succession. For example, a man who isyoung cannot at the same time be old; but thesame man can very well be at one time young,and at another not young, that is, old. Now theprinciple of contradiction as a merely logicalproposition must not by any means limit itsapplication merely to relations of time, andconsequently a formula like the preceding isquite foreign to its true purpose. The misun-derstanding arises in this way. We first of allseparate a predicate of a thing from the concep-tion of the thing, and afterwards connect withthis predicate its opposite, and hence do notestablish any contradiction with the subject, butonly with its predicate, which has been conjoi-ned with the subject synthetically— a contra-diction, moreover, which obtains only when thefirst and second predicate are affirmed in thesame time. If I say: "A man who is ignorant isnot learned," the condition "at the same time"must be added, for he who is at one time igno-

rant, may at another be learned. But if I say:"No ignorant man is a learned man," the propo-sition is analytical, because the characteristicignorance is now a constituent part of the con-ception of the subject; and in this case the nega-tive proposition is evident immediately fromthe proposition of contradiction, without thenecessity of adding the condition "the sametime." This is the reason why I have altered theformula of this principle—an alteration whichshows very clearly the nature of an analyticalproposition.

SECTION II. Of the Supreme Principle ofall Synthetical Judgements.

The explanation of the possibility of syntheticaljudgements is a task with which general logichas nothing to do; indeed she needs not evenbe acquainted with its name. But in transcen-

dental logic it is the most important matter tobe dealt with—indeed the only one, if the ques-tion is of the possibility of synthetical judge-ments a priori, the conditions and extent oftheir validity. For when this question is fullydecided, it can reach its aim with perfect ease,the determination, to wit, of the extent and li-mits of the pure understanding.

In an analytical judgement I do not go beyondthe given conception, in order to arrive at somedecision respecting it. If the judgement is affir-mative, I predicate of the conception only thatwhich was already cogitated in it; if negative, Imerely exclude from the conception its contra-ry. But in synthetical judgements, I must gobeyond the given conception, in order to cogi-tate, in relation with it, something quite diffe-rent from that which was cogitated in it, a rela-tion which is consequently never one either ofidentity or contradiction, and by means ofwhich the truth or error of the judgement can-

not be discerned merely from the judgementitself.

Granted, then, that we must go out beyond agiven conception, in order to compare it synt-hetically with another, a third thing is necessa-ry, in which alone the synthesis of two concep-tions can originate. Now what is this tertiumquid that is to be the medium of all syntheticaljudgements? It is only a complex in which allour representations are contained, the internalsense to wit, and its form a priori, time.

The synthesis of our representations rests uponthe imagination; their synthetical unity (whichis requisite to a judgement), upon the unity ofapperception. In this, therefore, is to be soughtthe possibility of synthetical judgements, andas all three contain the sources of a priori repre-sentations, the possibility of pure syntheticaljudgements also; nay, they are necessary uponthese grounds, if we are to possess a knowled-

ge of objects, which rests solely upon the synt-hesis of representations.

If a cognition is to have objective reality, that is,to relate to an object, and possess sense andmeaning in respect to it, it is necessary that theobject be given in some way or another. Wit-hout this, our conceptions are empty, and wemay indeed have thought by means of them,but by such thinking we have not, in fact, cog-nized anything, we have merely played withrepresentation. To give an object, if this expres-sion be understood in the sense of "to present"the object, not mediately but immediately inintuition, means nothing else than to apply therepresentation of it to experience, be that expe-rience real or only possible. Space and timethemselves, pure as these conceptions are fromall that is empirical, and certain as it is that theyare represented fully a priori in the mind,would be completely without objective validity,and without sense and significance, if their ne-

cessary use in the objects of experience werenot shown. Nay, the representation of them is amere schema, that always relates to the repro-ductive imagination, which calls up the objectsof experience, without which they have nomeaning. And so it is with all conceptions wit-hout distinction.

The possibility of experience is, then, thatwhich gives objective reality to all our a prioricognitions. Now experience depends upon thesynthetical unity of phenomena, that is, upon asynthesis according to conceptions of the objectof phenomena in general, a synthesis withoutwhich experience never could become know-ledge, but would be merely a rhapsody of per-ceptions, never fitting together into any connec-ted text, according to rules of a thoroughly uni-ted (possible) consciousness, and therefore ne-ver subjected to the transcendental and neces-sary unity of apperception. Experience has the-refore for a foundation, a priori principles of its

form, that is to say, general rules of unity in thesynthesis of phenomena, the objective reality ofwhich rules, as necessary conditions even of thepossibility of experience can which rules, asnecessary conditions—even of the possibility ofexperience—can always be shown in experien-ce. But apart from this relation, a priori synt-hetical propositions are absolutely impossible,because they have no third term, that is, nopure object, in which the synthetical unity canexhibit the objective reality of its conceptions.

Although, then, respecting space, or the formswhich productive imagination describes the-rein, we do cognize much a priori in syntheticaljudgements, and are really in no need of expe-rience for this purpose, such knowledge wouldnevertheless amount to nothing but a busy tri-fling with a mere chimera, were not space to beconsidered as the condition of the phenomenawhich constitute the material of external expe-rience. Hence those pure synthetical judge-

ments do relate, though but mediately, to pos-sible experience, or rather to the possibility ofexperience, and upon that alone is founded theobjective validity of their synthesis.

While then, on the one hand, experience, asempirical synthesis, is the only possible modeof cognition which gives reality to all othersynthesis; on the other hand, this latter synt-hesis, as cognition a priori, possesses truth, thatis, accordance with its object, only in so far as itcontains nothing more than what is necessaryto the synthetical unity of experience.

Accordingly, the supreme principle of all synt-hetical judgements is: "Every object is subject tothe necessary conditions of the synthetical uni-ty of the manifold of intuition in a possible ex-perience."

A priori synthetical judgements are possiblewhen we apply the formal conditions of the apriori intuition, the synthesis of the imagina-

tion, and the necessary unity of that synthesisin a transcendental apperception, to a possiblecognition of experience, and say: "The condi-tions of the possibility of experience in generalare at the same time conditions of the possibili-ty of the objects of experience, and have, forthat reason, objective validity in an a priorisynthetical judgement."

SECTION III. Systematic Representationof all Synthetical Principles of the Pure Un-derstanding.

That principles exist at all is to be ascribed sole-ly to the pure understanding, which is not onlythe faculty of rules in regard to that which hap-pens, but is even the source of principles accor-ding to which everything that can be presentedto us as an object is necessarily subject to rules,because without such rules we never could

attain to cognition of an object. Even the laws ofnature, if they are contemplated as principles ofthe empirical use of the understanding, possessalso a characteristic of necessity, and we maytherefore at least expect them to be determinedupon grounds which are valid a priori and an-tecedent to all experience. But all laws of natu-re, without distinction, are subject to higherprinciples of the understanding, inasmuch asthe former are merely applications of the latterto particular cases of experience. These higherprinciples alone therefore give the conception,which contains the necessary condition, and, asit were, the exponent of a rule; experience, onthe other hand, gives the case which comesunder the rule.

There is no danger of our mistaking merelyempirical principles for principles of the pureunderstanding, or conversely; for the characterof necessity, according to conceptions whichdistinguish the latter, and the absence of this in

every empirical proposition, how extensivelyvalid soever it may be, is a perfect safeguardagainst confounding them. There are, however,pure principles a priori, which nevertheless Ishould not ascribe to the pure understanding—for this reason, that they are not derived frompure conceptions, but (although by the media-tion of the understanding) from pure intuitions.But understanding is the faculty of conceptions.Such principles mathematical science possesses,but their application to experience, consequen-tly their objective validity, nay the possibility ofsuch a priori synthetical cognitions (the deduc-tion thereof) rests entirely upon the pure un-derstanding.

On this account, I shall not reckon among myprinciples those of mathematics; though I shallinclude those upon the possibility and objectivevalidity a priori, of principles of the mathema-tical science, which, consequently, are to belooked upon as the principle of these, and

which proceed from conceptions to intuition,and not from intuition to conceptions.

In the application of the pure conceptions of theunderstanding to possible experience, the em-ployment of their synthesis is either mathema-tical or dynamical, for it is directed partly onthe intuition alone, partly on the existence of aphenomenon. But the a priori conditions ofintuition are in relation to a possible experienceabsolutely necessary, those of the existence ofobjects of a possible empirical intuition are inthemselves contingent. Hence the principles ofthe mathematical use of the categories will pos-sess a character of absolute necessity, that is,will be apodeictic; those, on the other hand, ofthe dynamical use, the character of an a priorinecessity indeed, but only under the conditionof empirical thought in an experience, thereforeonly mediately and indirectly. Consequentlythey will not possess that immediate evidencewhich is peculiar to the former, although their

application to experience does not, for that rea-son, lose its truth and certitude. But of thispoint we shall be better able to judge at theconclusion of this system of principles.

The table of the categories is naturally our gui-de to the table of principles, because these arenothing else than rules for the objective em-ployment of the former. Accordingly, all prin-ciples of the pure understanding are:

1 Axioms of Intuition

2 3 Anticipations Analogies of Perception of Experience 4 Postulates of Empirical Thought in general

These appellations I have chosen advisedly, inorder that we might not lose sight of the dis-tinctions in respect of the evidence and the em-ployment of these principles. It will, however,soon appear that—a fact which concerns boththe evidence of these principles, and the a prio-ri determination of phenomena—according tothe categories of quantity and quality (if weattend merely to the form of these), the princi-ples of these categories are distinguishablefrom those of the two others, in as much as theformer are possessed of an intuitive, but thelatter of a merely discursive, though in bothinstances a complete, certitude. I shall thereforecall the former mathematical, and the latterdynamical principles.* It must be observed,however, that by these terms I mean just aslittle in the one case the principles of mathema-tics as those of general (physical) dynamics inthe other. I have here in view merely the prin-ciples of the pure understanding, in their appli-cation to the internal sense (without distinction

of the representations given therein), by meansof which the sciences of mathematics and dy-namics become possible. Accordingly, I havenamed these principles rather with reference totheir application than their content; and I shallnow proceed to consider them in the order inwhich they stand in the table.

[*Footnote: All combination (conjunctio) is eit-her composition (compositio) or connection(nexus). The former is the synthesis of a mani-fold, the parts of which do not necessarily be-long to each other. For example, the two trian-gles into which a square is divided by a diago-nal, do not necessarily belong to each other,and of this kind is the synthesis of the homoge-neous in everything that can be mathematicallyconsidered. This synthesis can be divided intothose of aggregation and coalition, the formerof which is applied to extensive, the latter tointensive quantities. The second sort of combi-

nation (nexus) is the synthesis of a manifold, inso far as its parts do belong necessarily to eachother; for example, the accident to a substance,or the effect to the cause. Consequently it is asynthesis of that which though heterogeneous,is represented as connected a priori. This com-bination—not an arbitrary one—I entitle dy-namical because it concerns the connection ofthe existence of the manifold. This, again, maybe divided into the physical synthesis, of thephenomena divided among each other, and themetaphysical synthesis, or the connection ofphenomena a priori in the faculty of cognition.]

1. AXIOMS OF INTUITION.

The principle of these is: All Intuitions are Ex-tensive Quantities.

PROOF.

All phenomena contain, as regards their form,an intuition in space and time, which lies apriori at the foundation of all without excep-tion. Phenomena, therefore, cannot be appre-hended, that is, received into empirical cons-ciousness otherwise than through the synthesisof a manifold, through which the representa-tions of a determinate space or time are genera-ted; that is to say, through the composition ofthe homogeneous and the consciousness of thesynthetical unity of this manifold (homoge-neous). Now the consciousness of a homoge-neous manifold in intuition, in so far as therebythe representation of an object is rendered pos-sible, is the conception of a quantity (quanti).Consequently, even the perception of an objectas phenomenon is possible only through thesame synthetical unity of the manifold of thegiven sensuous intuition, through which the

unity of the composition of the homogeneousmanifold in the conception of a quantity is cogi-tated; that is to say, all phenomena are quanti-ties, and extensive quantities, because as intui-tions in space or time they must be representedby means of the same synthesis through whichspace and time themselves are determined.

An extensive quantity I call that wherein therepresentation of the parts renders possible(and therefore necessarily antecedes) the repre-sentation of the whole. I cannot represent tomyself any line, however small, without dra-wing it in thought, that is, without generatingfrom a point all its parts one after another, andin this way alone producing this intuition. Pre-cisely the same is the case with every, even thesmallest, portion of time. I cogitate therein onlythe successive progress from one moment toanother, and hence, by means of the differentportions of time and the addition of them, adeterminate quantity of time is produced. As

the pure intuition in all phenomena is eithertime or space, so is every phenomenon in itscharacter of intuition an extensive quantity,inasmuch as it can only be cognized in our ap-prehension by successive synthesis (from partto part). All phenomena are, accordingly, to beconsidered as aggregates, that is, as a collectionof previously given parts; which is not the casewith every sort of quantities, but only with tho-se which are represented and apprehended byus as extensive.

On this successive synthesis of the productiveimagination, in the generation of figures, isfounded the mathematics of extension, or geo-metry, with its axioms, which express the con-ditions of sensuous intuition a priori, underwhich alone the schema of a pure conception ofexternal intuition can exist; for example, "between two points only one straight line is pos-sible," "two straight lines cannot enclose a spa-

ce," etc. These are the axioms which properlyrelate only to quantities (quanta) as such.

But, as regards the quantity of a thing (quanti-tas), that is to say, the answer to the question:"How large is this or that object?" although, inrespect to this question, we have various pro-positions synthetical and immediately certain(indemonstrabilia); we have, in the proper sen-se of the term, no axioms. For example, thepropositions: "If equals be added to equals, thewholes are equal"; "If equals be taken fromequals, the remainders are equal"; are analyti-cal, because I am immediately conscious of theidentity of the production of the one quantitywith the production of the other; whereasaxioms must be a priori synthetical proposi-tions. On the other hand, the self-evident pro-positions as to the relation of numbers, are cer-tainly synthetical but not universal, like thoseof geometry, and for this reason cannot be ca-lled axioms, but numerical formulae. That 7 + 5

= 12 is not an analytical proposition. For neitherin the representation of seven, nor of five, norof the composition of the two numbers, do Icogitate the number twelve. (Whether I cogitatethe number in the addition of both, is not atpresent the question; for in the case of an analy-tical proposition, the only point is whether Ireally cogitate the predicate in the representa-tion of the subject.) But although the proposi-tion is synthetical, it is nevertheless only a sin-gular proposition. In so far as regard is herehad merely to the synthesis of the homoge-neous (the units), it cannot take place except inone manner, although our use of these numbersis afterwards general. If I say: "A triangle can beconstructed with three lines, any two of whichtaken together are greater than the third," Iexercise merely the pure function of the pro-ductive imagination, which may draw the lineslonger or shorter and construct the angles at itspleasure. On the contrary, the number seven ispossible only in one manner, and so is likewise

the number twelve, which results from thesynthesis of seven and five. Such propositions,then, cannot be termed axioms (for in that casewe should have an infinity of these), but nume-rical formulae.

This transcendental principle of the mathema-tics of phenomena greatly enlarges our a prioricognition. For it is by this principle alone thatpure mathematics is rendered applicable in allits precision to objects of experience, and wit-hout it the validity of this application wouldnot be so self-evident; on the contrary, contra-dictions and confusions have often arisen onthis very point. Phenomena are not things inthemselves. Empirical intuition is possible onlythrough pure intuition (of space and time); con-sequently, what geometry affirms of the latter,is indisputably valid of the former. All eva-sions, such as the statement that objects of sen-se do not conform to the rules of constructionin space (for example, to the rule of the infinite

divisibility of lines or angles), must fall to theground. For, if these objections hold good, wedeny to space, and with it to all mathematics,objective validity, and no longer know where-fore, and how far, mathematics can be appliedto phenomena. The synthesis of spaces andtimes as the essential form of all intuition, isthat which renders possible the apprehensionof a phenomenon, and therefore every externalexperience, consequently all cognition of theobjects of experience; and whatever mathema-tics in its pure use proves of the former, mustnecessarily hold good of the latter. All objec-tions are but the chicaneries of an ill-instructedreason, which erroneously thinks to liberate theobjects of sense from the formal conditions ofour sensibility, and represents these, althoughmere phenomena, as things in themselves, pre-sented as such to our understanding. But in thiscase, no a priori synthetical cognition of themcould be possible, consequently not throughpure conceptions of space and the science

which determines these conceptions, that is tosay, geometry, would itself be impossible.

2. ANTICIPATIONS OF PERCEPTION.

The principle of these is: In all phenomena theReal, that which is an object of sensation, hasIntensive Quantity, that is, has a Degree.

PROOF.

Perception is empirical consciousness, that is tosay, a consciousness which contains an elementof sensation. Phenomena as objects of percep-tion are not pure, that is, merely formal intui-tions, like space and time, for they cannot beperceived in themselves. [Footnote: They can beperceived only as phenomena, and some part

of them must always belong to the non-ego;whereas pure intuitions are entirely the pro-ducts of the mind itself, and as such are cogui-zed IN THEMSELVES.—Tr] They contain, then,over and above the intuition, the materials foran object (through which is represented somet-hing existing in space or time), that is to say,they contain the real of sensation, as a represen-tation merely subjective, which gives us merelythe consciousness that the subject is affected,and which we refer to some external object.Now, a gradual transition from empirical cons-ciousness to pure consciousness is possible,inasmuch as the real in this consciousness enti-rely vanishes, and there remains a merely for-mal consciousness (a priori) of the manifold intime and space; consequently there is possible asynthesis also of the production of the quantityof a sensation from its commencement, that is,from the pure intuition = 0 onwards up to acertain quantity of the sensation. Now as sensa-tion in itself is not an objective representation,

and in it is to be found neither the intuition ofspace nor of time, it cannot possess any exten-sive quantity, and yet there does belong to it aquantity (and that by means of its apprehen-sion, in which empirical consciousness can wit-hin a certain time rise from nothing = 0 up to itsgiven amount), consequently an intensivequantity. And thus we must ascribe intensivequantity, that is, a degree of influence on senseto all objects of perception, in so far as this per-ception contains sensation.

All cognition, by means of which I am enabledto cognize and determine a priori what belongsto empirical cognition, may be called an antici-pation; and without doubt this is the sense inwhich Epicurus employed his expression prho-lepsis. But as there is in phenomena somethingwhich is never cognized a priori, which on thisaccount constitutes the proper difference bet-ween pure and empirical cognition, that is tosay, sensation (as the matter of perception), it

follows, that sensation is just that element incognition which cannot be at all anticipated. Onthe other hand, we might very well term thepure determinations in space and time, as wellin regard to figure as to quantity, anticipationsof phenomena, because they represent a priorithat which may always be given a posteriori inexperience. But suppose that in every sensa-tion, as sensation in general, without any parti-cular sensation being thought of, there existedsomething which could be cognized a priori,this would deserve to be called anticipation in aspecial sense—special, because it may seemsurprising to forestall experience, in that whichconcerns the matter of experience, and whichwe can only derive from itself. Yet such reallyis the case here.

Apprehension*, by means of sensation alone,fills only one moment, that is, if I do not takeinto consideration a succession of many sensa-tions. As that in the phenomenon, the appre-

hension of which is not a successive synthesisadvancing from parts to an entire representa-tion, sensation has therefore no extensive quan-tity; the want of sensation in a moment of timewould represent it as empty, consequently = 0.That which in the empirical intuition corres-ponds to sensation is reality (realitas phaeno-menon); that which corresponds to the absenceof it, negation = 0. Now every sensation is ca-pable of a diminution, so that it can decrease,and thus gradually disappear. Therefore, bet-ween reality in a phenomenon and negation,there exists a continuous concatenation of ma-ny possible intermediate sensations, the diffe-rence of which from each other is always sma-ller than that between the given sensation andzero, or complete negation. That is to say, thereal in a phenomenon has always a quantity,which however is not discoverable in appre-hension, inasmuch as apprehension take placeby means of mere sensation in one instant, andnot by the successive synthesis of many sensa-

tions, and therefore does not progress fromparts to the whole. Consequently, it has a quan-tity, but not an extensive quantity.

[*Footnote: Apprehension is the Kantian wordfor preception, in the largest sense in which weemploy that term. It is the genus which inclu-des under i, as species, perception proper andsensation proper—Tr]

Now that quantity which is apprehended onlyas unity, and in which plurality can be repre-sented only by approximation to negation = O,I term intensive quantity. Consequently, realityin a phenomenon has intensive quantity, thatis, a degree. If we consider this reality as cause(be it of sensation or of another reality in thephenomenon, for example, a change), we callthe degree of reality in its character of cause amomentum, for example, the momentum ofweight; and for this reason, that the degree onlyindicates that quantity the apprehension ofwhich is not successive, but instantaneous.

This, however, I touch upon only in passing,for with causality I have at present nothing todo.

Accordingly, every sensation, consequentlyevery reality in phenomena, however small itmay be, has a degree, that is, an intensive quan-tity, which may always be lessened, and bet-ween reality and negation there exists a conti-nuous connection of possible realities, and pos-sible smaller perceptions. Every colour— forexample, red—has a degree, which, be it everso small, is never the smallest, and so is it al-ways with heat, the momentum of weight, etc.

This property of quantities, according to whichno part of them is the smallest possible (no partsimple), is called their continuity. Space andtime are quanta continua, because no part ofthem can be given, without enclosing it withinboundaries (points and moments), consequen-tly, this given part is itself a space or a time.Space, therefore, consists only of spaces, and

time of times. Points and moments are onlyboundaries, that is, the mere places or positionsof their limitation. But places always presuppo-se intuitions which are to limit or determinethem; and we cannot conceive either space ortime composed of constituent parts which aregiven before space or time. Such quantities mayalso be called flowing, because synthesis (of theproductive imagination) in the production ofthese quantities is a progression in time, thecontinuity of which we are accustomed to indi-cate by the expression flowing.

All phenomena, then, are continuous quanti-ties, in respect both to intuition and mere per-ception (sensation, and with it reality). In theformer case they are extensive quantities; in thelatter, intensive. When the synthesis of the ma-nifold of a phenomenon is interrupted, thereresults merely an aggregate of several pheno-mena, and not properly a phenomenon as aquantity, which is not produced by the mere

continuation of the productive synthesis of acertain kind, but by the repetition of a synthesisalways ceasing. For example, if I call thirteendollars a sum or quantity of money, I employthe term quite correctly, inasmuch as I unders-tand by thirteen dollars the value of a mark instandard silver, which is, to be sure, a conti-nuous quantity, in which no part is the sma-llest, but every part might constitute a piece ofmoney, which would contain material for stillsmaller pieces. If, however, by the words thir-teen dollars I understand so many coins (betheir value in silver what it may), it would bequite erroneous to use the expression a quanti-ty of dollars; on the contrary, I must call themaggregate, that is, a number of coins. And as inevery number we must have unity as the foun-dation, so a phenomenon taken as unity is aquantity, and as such always a continuousquantity (quantum continuum).

Now, seeing all phenomena, whether conside-red as extensive or intensive, are continuousquantities, the proposition: "All change (transi-tion of a thing from one state into another) iscontinuous," might be proved here easily, andwith mathematical evidence, were it not thatthe causality of a change lies, entirely beyondthe bounds of a transcendental philosophy, andpresupposes empirical principles. For of thepossibility of a cause which changes the condi-tion of things, that is, which determines them tothe contrary to a certain given state, the unders-tanding gives us a priori no knowledge; notmerely because it has no insight into the possi-bility of it (for such insight is absent in several apriori cognitions), but because the notion ofchange concerns only certain determinations ofphenomena, which experience alone can ac-quaint us with, while their cause lies in the un-changeable. But seeing that we have nothingwhich we could here employ but the pure fun-damental conceptions of all possible experien-

ce, among which of course nothing empiricalcan be admitted, we dare not, without injuringthe unity of our system, anticipate general phy-sical science, which is built upon certain fun-damental experiences.

Nevertheless, we are in no want of proofs of thegreat influence which the principle above deve-loped exercises in the anticipation of percep-tions, and even in supplying the want of them,so far as to shield us against the false conclu-sions which otherwise we might rashly draw.

If all reality in perception has a degree, bet-ween which and negation there is an endlesssequence of ever smaller degrees, and if, ne-vertheless, every sense must have a determina-te degree of receptivity for sensations; no per-ception, and consequently no experience is pos-sible, which can prove, either immediately ormediately, an entire absence of all reality in aphenomenon; in other words, it is impossibleever to draw from experience a proof of the

existence of empty space or of empty time. Forin the first place, an entire absence of reality ina sensuous intuition cannot of course be anobject of perception; secondly, such absencecannot be deduced from the contemplation ofany single phenomenon, and the difference ofthe degrees in its reality; nor ought it ever to beadmitted in explanation of any phenomenon.For if even the complete intuition of a determi-nate space or time is thoroughly real, that is, ifno part thereof is empty, yet because every rea-lity has its degree, which, with the extensivequantity of the phenomenon unchanged, candiminish through endless gradations down tonothing (the void), there must be infinitely gra-duated degrees, with which space or time isfilled, and the intensive quantity in differentphenomena may be smaller or greater, alt-hough the extensive quantity of the intuitionremains equal and unaltered.

We shall give an example of this. Almost allnatural philosophers, remarking a great diffe-rence in the quantity of the matter of differentkinds in bodies with the same volume (partlyon account of the momentum of gravity orweight, partly on account of the momentum ofresistance to other bodies in motion), concludeunanimously that this volume (extensive quan-tity of the phenomenon) must be void in allbodies, although in different proportion. Butwho would suspect that these for the most partmathematical and mechanical inquirers intonature should ground this conclusion solely ona metaphysical hypothesis—a sort of hypot-hesis which they profess to disparage andavoid? Yet this they do, in assuming that thereal in space (I must not here call it impenetra-bility or weight, because these are empiricalconceptions) is always identical, and can onlybe distinguished according to its extensivequantity, that is, multiplicity. Now to this pre-supposition, for which they can have no

ground in experience, and which consequentlyis merely metaphysical, I oppose a transcen-dental demonstration, which it is true will notexplain the difference in the filling up of spaces,but which nevertheless completely does awaywith the supposed necessity of the above-mentioned presupposition that we cannot ex-plain the said difference otherwise than by thehypothesis of empty spaces. This demonstra-tion, moreover, has the merit of setting the un-derstanding at liberty to conceive this distinc-tion in a different manner, if the explanation ofthe fact requires any such hypothesis. For weperceive that although two equal spaces may becompletely filled by matters altogether diffe-rent, so that in neither of them is there left asingle point wherein matter is not present, ne-vertheless, every reality has its degree (of resis-tance or of weight), which, without diminutionof the extensive quantity, can become less andless ad infinitum, before it passes into nothing-ness and disappears. Thus an expansion which

fills a space—for example, caloric, or any otherreality in the phenomenal world—can decreasein its degrees to infinity, yet without leavingthe smallest part of the space empty; on thecontrary, filling it with those lesser degrees ascompletely as another phenomenon could withgreater. My intention here is by no means tomaintain that this is really the case with thedifference of matters, in regard to their specificgravity; I wish only to prove, from a principleof the pure understanding, that the nature ofour perceptions makes such a mode of explana-tion possible, and that it is erroneous to regardthe real in a phenomenon as equal quoad itsdegree, and different only quoad its aggrega-tion and extensive quantity, and this, too, onthe pretended authority of an a priori principleof the understanding.

Nevertheless, this principle of the anticipationof perception must somewhat startle an inqui-rer whom initiation into transcendental philo-

sophy has rendered cautious. We must natura-lly entertain some doubt whether or not theunderstanding can enounce any such syntheti-cal proposition as that respecting the degree ofall reality in phenomena, and consequently thepossibility of the internal difference of sensa-tion itself—abstraction being made of its empi-rical quality. Thus it is a question not unworthyof solution: "How the understanding can pro-nounce synthetically and a priori respectingphenomena, and thus anticipate these, even inthat which is peculiarly and merely empirical,that, namely, which concerns sensation itself?"

The quality of sensation is in all cases merelyempirical, and cannot be represented a priori(for example, colours, taste, etc.). But the real—that which corresponds to sensation—in oppo-sition to negation = 0, only represents somet-hing the conception of which in itself contains abeing (ein seyn), and signifies nothing but thesynthesis in an empirical consciousness. That is

to say, the empirical consciousness in the inter-nal sense can be raised from 0 to every higherdegree, so that the very same extensive quanti-ty of intuition, an illuminated surface, forexample, excites as great a sensation as an ag-gregate of many other surfaces less illuminated.We can therefore make complete abstraction ofthe extensive quantity of a phenomenon, andrepresent to ourselves in the mere sensation ina certain momentum, a synthesis of homoge-neous ascension from 0 up to the given empiri-cal consciousness, All sensations therefore assuch are given only a posteriori, but this pro-perty thereof, namely, that they have a degree,can be known a priori. It is worthy of remark,that in respect to quantities in general, we cancognize a priori only a single quality, namely,continuity; but in respect to all quality (the realin phenomena), we cannot cognize a priorianything more than the intensive quantity the-reof, namely, that they have a degree. All else isleft to experience.

3. ANALOGIES OF EXPERIENCE.

The principle of these is: Experience is possibleonly through the representation of a necessaryconnection of Perceptions.

PROOF.

Experience is an empirical cognition; that is tosay, a cognition which determines an object bymeans of perceptions. It is therefore a synthesisof perceptions, a synthesis which is not itselfcontained in perception, but which contains thesynthetical unity of the manifold of perceptionin a consciousness; and this unity constitutesthe essential of our cognition of objects of thesenses, that is, of experience (not merely of in-tuition or sensation). Now in experience ourperceptions come together contingently, so thatno character of necessity in their connection

appears, or can appear from the perceptionsthemselves, because apprehension is only aplacing together of the manifold of empiricalintuition, and no representation of a necessityin the connected existence of the phenomenawhich apprehension brings together, is to bediscovered therein. But as experience is a cogni-tion of objects by means of perceptions, it fo-llows that the relation of the existence of theexistence of the manifold must be representedin experience not as it is put together in time,but as it is objectively in time. And as time itselfcannot be perceived, the determination of theexistence of objects in time can only take placeby means of their connection in time in general,consequently only by means of a priori connec-ting conceptions. Now as these conceptionsalways possess the character of necessity, expe-rience is possible only by means of a represen-tation of the necessary connection of percep-tion.

The three modi of time are permanence, succes-sion, and coexistence. Accordingly, there arethree rules of all relations of time in phenome-na, according to which the existence of everyphenomenon is determined in respect of theunity of all time, and these antecede all expe-rience and render it possible.

The general principle of all three analogies restson the necessary unity of apperception in rela-tion to all possible empirical consciousness(perception) at every time, consequently, as thisunity lies a priori at the foundation of all men-tal operations, the principle rests on the synt-hetical unity of all phenomena according totheir relation in time. For the original appercep-tion relates to our internal sense (the complexof all representations), and indeed relates apriori to its form, that is to say, the relation ofthe manifold empirical consciousness in time.Now this manifold must be combined in origi-nal apperception according to relations of ti-

me—a necessity imposed by the a priori trans-cendental unity of apperception, to which issubjected all that can belong to my (i.e., myown) cognition, and therefore all that can be-come an object for me. This synthetical and apriori determined unity in relation of percep-tions in time is therefore the rule: "All empiricaldeterminations of time must be subject to rulesof the general determination of time"; and theanalogies of experience, of which we are nowabout to treat, must be rules of this nature.

These principles have this peculiarity, that theydo not concern phenomena, and the synthesisof the empirical intuition thereof, but merelythe existence of phenomena and their relationto each other in regard to this existence. Nowthe mode in which we apprehend a thing in aphenomenon can be determined a priori insuch a manner that the rule of its synthesis cangive, that is to say, can produce this a prioriintuition in every empirical example. But the

existence of phenomena cannot be known apriori, and although we could arrive by thispath at a conclusion of the fact of some existen-ce, we could not cognize that existence deter-minately, that is to say, we should be incapableof anticipating in what respect the empiricalintuition of it would be distinguishable fromthat of others.

The two principles above mentioned, which Icalled mathematical, in consideration of the factof their authorizing the application of mat-hematic phenomena, relate to these phenomenaonly in regard to their possibility, and instructus how phenomena, as far as regards their in-tuition or the real in their perception, can begenerated according to the rules of a mathema-tical synthesis. Consequently, numerical quan-tities, and with them the determination of aphenomenon as a quantity, can be employed inthe one case as well as in the other. Thus, forexample, out of 200,000 illuminations by the

moon, I might compose and give a priori, thatis construct, the degree of our sensations of thesun-light.* We may therefore entitle these twoprinciples constitutive.

[*Footnote: Kant's meaning is: The two princi-ples enunciated under the heads of "Axioms ofIntuition," and "Anticipations of Perception,"authorize the application to phenomena of de-terminations of size and number, that is of mat-hematic. For exampkle, I may compute the lightof the sun, and say that its quantity is a certainnumber of times greater than that of the moon.In the same way, heat is measured by the com-parison of its different effects on water, &c.,and on mercury in a thermometer.—Tr]

The case is very different with those principleswhose province it is to subject the existence ofphenomena to rules a priori. For as existencedoes not admit of being constructed, it is clearthat they must only concern the relations ofexistence and be merely regulative principles.

In this case, therefore, neither axioms nor anti-cipations are to be thought of. Thus, if a percep-tion is given us, in a certain relation of time toother (although undetermined) perceptions, wecannot then say a priori, what and how great(in quantity) the other perception necessarilyconnected with the former is, but only how it isconnected, quoad its existence, in this givenmodus of time. Analogies in philosophy meansomething very different from that which theyrepresent in mathematics. In the latter they areformulae, which enounce the equality of tworelations of quantity, and are always constituti-ve, so that if two terms of the proportion aregiven, the third is also given, that is, can beconstructed by the aid of these formulae. But inphilosophy, analogy is not the equality of twoquantitative but of two qualitative relations. Inthis case, from three given terms, I can give apriori and cognize the relation to a fourthmember, but not this fourth term itself, alt-hough I certainly possess a rule to guide me in

the search for this fourth term in experience,and a mark to assist me in discovering it. Ananalogy of experience is therefore only a ruleaccording to which unity of experience mustarise out of perceptions in respect to objects(phenomena) not as a constitutive, but merelyas a regulative principle. The same holds goodalso of the postulates of empirical thought ingeneral, which relate to the synthesis of mereintuition (which concerns the form of pheno-mena), the synthesis of perception (which con-cerns the matter of phenomena), and the synt-hesis of experience (which concerns the relationof these perceptions). For they are only regula-tive principles, and clearly distinguishable fromthe mathematical, which are constitutive, notindeed in regard to the certainty which bothpossess a priori, but in the mode of evidencethereof, consequently also in the manner ofdemonstration.

But what has been observed of all syntheticalpropositions, and must be particularly remar-ked in this place, is this, that these analogiespossess significance and validity, not as princi-ples of the transcendental, but only as princi-ples of the empirical use of the understanding,and their truth can therefore be proved only assuch, and that consequently the phenomenamust not be subjoined directly under the cate-gories, but only under their schemata. For if theobjects to which those principles must be ap-plied were things in themselves, it would bequite impossible to cognize aught concerningthem synthetically a priori. But they are not-hing but phenomena; a complete knowledge ofwhich—a knowledge to which all principles apriori must at last relate—is the only possibleexperience. It follows that these principles canhave nothing else for their aim than the condi-tions of the empirical cognition in the unity ofsynthesis of phenomena. But this synthesis iscogitated only in the schema of the pure con-

ception of the understanding, of whose unity,as that of a synthesis in general, the categorycontains the function unrestricted by any sen-suous condition. These principles will thereforeauthorize us to connect phenomena accordingto an analogy, with the logical and universalunity of conceptions, and consequently to em-ploy the categories in the principles themselves;but in the application of them to experience, weshall use only their schemata, as the key to theirproper application, instead of the categories, orrather the latter as restricting conditions, underthe title of "formulae" of the former.

A. FIRST ANALOGY.

Principle of the Permanence of Substance.

In all changes of phenomena, substance is per-manent, and the quantum thereof in nature isneither increased nor diminished.

PROOF.

All phenomena exist in time, wherein alone assubstratum, that is, as the permanent form ofthe internal intuition, coexistence and succes-sion can be represented. Consequently time, inwhich all changes of phenomena must be cogi-tated, remains and changes not, because it isthat in which succession and coexistence can berepresented only as determinations thereof.Now, time in itself cannot be an object of per-ception. It follows that in objects of perception,that is, in phenomena, there must be found asubstratum which represents time in general,and in which all change or coexistence can beperceived by means of the relation of pheno-

mena to it. But the substratum of all reality, thatis, of all that pertains to the existence of things,is substance; all that pertains to existence can becogitated only as a determination of substance.Consequently, the permanent, in relation towhich alone can all relations of time in pheno-mena be determined, is substance in the worldof phenomena, that is, the real in phenomena,that which, as the substratum of all change,remains ever the same. Accordingly, as thiscannot change in existence, its quantity in natu-re can neither be increased nor diminished.

Our apprehension of the manifold in a pheno-menon is always successive, is Consequentlyalways changing. By it alone we could, therefo-re, never determine whether this manifold, asan object of experience, is coexistent or succes-sive, unless it had for a foundation somethingfixed and permanent, of the existence of whichall succession and coexistence are nothing butso many modes (modi of time). Only in the

permanent, then, are relations of time possible(for simultaneity and succession are the onlyrelations in time); that is to say, the permanentis the substratum of our empirical representa-tion of time itself, in which alone all determina-tion of time is possible. Permanence is, in fact,just another expression for time, as the abidingcorrelate of all existence of phenomena, and ofall change, and of all coexistence. For changedoes not affect time itself, but only the pheno-mena in time (just as coexistence cannot be re-garded as a modus of time itself, seeing that intime no parts are coexistent, but all successive).If we were to attribute succession to time itself,we should be obliged to cogitate another time,in which this succession would be possible. It isonly by means of the permanent that existencein different parts of the successive series of timereceives a quantity, which we entitle duration.For in mere succession, existence is perpetuallyvanishing and recommencing, and thereforenever has even the least quantity. Without the

permanent, then, no relation in time is possible.Now, time in itself is not an object of percep-tion; consequently the permanent in phenome-na must be regarded as the substratum of alldetermination of time, and consequently also asthe condition of the possibility of all syntheticalunity of perceptions, that is, of experience; andall existence and all change in time can only beregarded as a mode in the existence of thatwhich abides unchangeably. Therefore, in allphenomena, the permanent is the object in it-self, that is, the substance (phenomenon); butall that changes or can change belongs only tothe mode of the existence of this substance orsubstances, consequently to its determinations.

I find that in all ages not only the philosopher,but even the common understanding, has pre-posited this permanence as a substratum of allchange in phenomena; indeed, I am compelledto believe that they will always accept this as anindubitable fact. Only the philosopher expres-

ses himself in a more precise and definite man-ner, when he says: "In all changes in the world,the substance remains, and the accidents aloneare changeable." But of this decidedly syntheti-cal proposition, I nowhere meet with even anattempt at proof; nay, it very rarely has thegood fortune to stand, as it deserves to do, atthe head of the pure and entirely a priori lawsof nature. In truth, the statement that substanceis permanent, is tautological. For this verypermanence is the ground on which we applythe category of substance to the phenomenon;and we should have been obliged to prove thatin all phenomena there is something perma-nent, of the existence of which the changeable isnothing but a determination. But because aproof of this nature cannot be dogmatical, thatis, cannot be drawn from conceptions, inas-much as it concerns a synthetical proposition apriori, and as philosophers never reflected thatsuch propositions are valid only in relation topossible experience, and therefore cannot be

proved except by means of a deduction of thepossibility of experience, it is no wonder thatwhile it has served as the foundation of all ex-perience (for we feel the need of it in empiricalcognition), it has never been supported byproof.

A philosopher was asked: "What is the weightof smoke?" He answered: "Subtract from theweight of the burnt wood the weight of theremaining ashes, and you will have the weightof the smoke." Thus he presumed it to be incon-trovertible that even in fire the matter (substan-ce) does not perish, but that only the form of itundergoes a change. In like manner was thesaying: "From nothing comes nothing," onlyanother inference from the principle or perma-nence, or rather of the ever-abiding existence ofthe true subject in phenomena. For if that in thephenomenon which we call substance is to bethe proper substratum of all determination oftime, it follows that all existence in past as well

as in future time, must be determinable bymeans of it alone. Hence we are entitled to ap-ply the term substance to a phenomenon, onlybecause we suppose its existence in all time, anotion which the word permanence does notfully express, as it seems rather to be referableto future time. However, the internal necessityperpetually to be, is inseparably connectedwith the necessity always to have been, and sothe expression may stand as it is. "Gigni denihilo nihil; in nihilum nil posse reverti,"* aretwo propositions which the ancients never par-ted, and which people nowadays sometimesmistakenly disjoin, because they imagine thatthe propositions apply to objects as things inthemselves, and that the former might be ini-mical to the dependence (even in respect of itssubstance also) of the world upon a supremecause. But this apprehension is entirely need-less, for the question in this case is only of phe-nomena in the sphere of experience, the unityof which never could be possible, if we admit-

ted the possibility that new things (in respect oftheir substance) should arise. For in that case,we should lose altogether that which alone canrepresent the unity of time, to wit, the identityof the substratum, as that through which aloneall change possesses complete and thoroughunity. This permanence is, however, nothingbut the manner in which we represent to our-selves the existence of things in the phenome-nal world.

[*Footnote: Persius, Satirae, iii.83-84.]

The determinations of a substance, which areonly particular modes of its existence, are ca-lled accidents. They are always real, becausethey concern the existence of substance (nega-tions are only determinations, which expressthe non-existence of something in the substan-ce). Now, if to this real in the substance we as-cribe a particular existence (for example, tomotion as an accident of matter), this existenceis called inherence, in contradistinction to the

existence of substance, which we call subsisten-ce. But hence arise many misconceptions, and itwould be a more accurate and just mode ofexpression to designate the accident only as themode in which the existence of a substance ispositively determined. Meanwhile, by reason ofthe conditions of the logical exercise of our un-derstanding, it is impossible to avoid separa-ting, as it were, that which in the existence of asubstance is subject to change, whilst the subs-tance remains, and regarding it in relation tothat which is properly permanent and radical.On this account, this category of substancestands under the title of relation, rather becauseit is the condition thereof than because it con-tains in itself any relation.

Now, upon this notion of permanence rests theproper notion of the conception change. Originand extinction are not changes of that whichoriginates or becomes extinct. Change is but amode of existence, which follows on another

mode of existence of the same object; hence allthat changes is permanent, and only the condi-tion thereof changes. Now since this mutationaffects only determinations, which can have abeginning or an end, we may say, employingan expression which seems somewhat para-doxical: "Only the permanent (substance) issubject to change; the mutable suffers no chan-ge, but rather alternation, that is, when certaindeterminations cease, others begin."

Change, when, cannot be perceived by us ex-cept in substances, and origin or extinction inan absolute sense, that does not concern merelya determination of the permanent, cannot be apossible perception, for it is this very notion ofthe permanent which renders possible the re-presentation of a transition from one state intoanother, and from non-being to being, which,consequently, can be empirically cognized onlyas alternating determinations of that which ispermanent. Grant that a thing absolutely begins

to be; we must then have a point of time inwhich it was not. But how and by what can wefix and determine this point of time, unless bythat which already exists? For a void time—preceding—is not an object of perception; but ifwe connect this beginning with objects whichexisted previously, and which continue to existtill the object in question in question begins tobe, then the latter can only be a determinationof the former as the permanent. The same holdsgood of the notion of extinction, for this pre-supposes the empirical representation of a time,in which a phenomenon no longer exists.

Substances (in the world of phenomena) are thesubstratum of all determinations of time. Thebeginning of some, and the ceasing to be ofother substances, would utterly do away withthe only condition of the empirical unity oftime; and in that case phenomena would relateto two different times, in which, side by side,existence would pass; which is absurd. For the-

re is only one time in which all different timesmust be placed, not as coexistent, but as succes-sive.

Accordingly, permanence is a necessary condi-tion under which alone phenomena, as thingsor objects, are determinable in a possible expe-rience. But as regards the empirical criterion ofthis necessary permanence, and with it of thesubstantiality of phenomena, we shall find suf-ficient opportunity to speak in the sequel.

B. SECOND ANALOGY.

Principle of the Succession of Time Accordingto the Law of Causality. All changes take placeaccording to the law of the connection of Causeand Effect.

PROOF.

(That all phenomena in the succession of timeare only changes, that is, a successive being andnon-being of the determinations of substance,which is permanent; consequently that a beingof substance itself which follows on the non-being thereof, or a non-being of substancewhich follows on the being thereof, in otherwords, that the origin or extinction of substanceitself, is impossible—all this has been fully es-tablished in treating of the foregoing principle.This principle might have been expressed asfollows: "All alteration (succession) of pheno-mena is merely change"; for the changes ofsubstance are not origin or extinction, becausethe conception of change presupposes the samesubject as existing with two opposite determi-nations, and consequently as permanent. Afterthis premonition, we shall proceed to theproof.)

I perceive that phenomena succeed one anot-her, that is to say, a state of things exists at onetime, the opposite of which existed in a formerstate. In this case, then, I really connect togethertwo perceptions in time. Now connection is notan operation of mere sense and intuition, but isthe product of a synthetical faculty of imagina-tion, which determines the internal sense inrespect of a relation of time. But imaginationcan connect these two states in two ways, sothat either the one or the other may antecede intime; for time in itself cannot be an object ofperception, and what in an object precedes andwhat follows cannot be empirically determinedin relation to it. I am only conscious, then, thatmy imagination places one state before and theother after; not that the one state antecedes theother in the object. In other words, the objectiverelation of the successive phenomena remainsquite undetermined by means of mere percep-tion. Now in order that this relation may becognized as determined, the relation between

the two states must be so cogitated that it isthereby determined as necessary, which ofthem must be placed before and which after,and not conversely. But the conception whichcarries with it a necessity of synthetical unity,can be none other than a pure conception of theunderstanding which does not lie in mere per-ception; and in this case it is the conception of"the relation of cause and effect," the former ofwhich determines the latter in time, as its ne-cessary consequence, and not as somethingwhich might possibly antecede (or which mightin some cases not be perceived to follow). Itfollows that it is only because we subject thesequence of phenomena, and consequently allchange, to the law of causality, that experienceitself, that is, empirical cognition of phenome-na, becomes possible; and consequently, thatphenomena themselves, as objects of experien-ce, are possible only by virtue of this law.

Our apprehension of the manifold of pheno-mena is always successive. The representationsof parts succeed one another. Whether theysucceed one another in the object also, is a se-cond point for reflection, which was not contai-ned in the former. Now we may certainly givethe name of object to everything, even to everyrepresentation, so far as we are conscious the-reof; but what this word may mean in the caseof phenomena, not merely in so far as they (asrepresentations) are objects, but only in so faras they indicate an object, is a question requi-ring deeper consideration. In so far as they,regarded merely as representations, are at thesame time objects of consciousness, they are notto be distinguished from apprehension, that is,reception into the synthesis of imagination, andwe must therefore say: "The manifold of phe-nomena is always produced successively in themind." If phenomena were things in themsel-ves, no man would be able to conjecture fromthe succession of our representations how this

manifold is connected in the object; for we haveto do only with our representations. Howthings may be in themselves, without regard tothe representations through which they affectus, is utterly beyond the sphere of our cogni-tion. Now although phenomena are not thingsin themselves, and are nevertheless the onlything given to us to be cognized, it is my dutyto show what sort of connection in time belongsto the manifold in phenomena themselves, whi-le the representation of this manifold in appre-hension is always successive. For example, theapprehension of the manifold in the phenome-non of a house which stands before me, is suc-cessive. Now comes the question whether themanifold of this house is in itself successive—which no one will be at all willing to grant. But,so soon as I raise my conception of an object tothe transcendental signification thereof, I findthat the house is not a thing in itself, but only aphenomenon, that is, a representation, thetranscendental object of which remains utterly

unknown. What then am I to understand by thequestion: "How can the manifold be connectedin the phenomenon itself—not considered as athing in itself, but merely as a phenomenon?"Here that which lies in my successive appre-hension is regarded as representation, whilstthe phenomenon which is given me, notwiths-tanding that it is nothing more than a complexof these representations, is regarded as the ob-ject thereof, with which my conception, drawnfrom the representations of apprehension, mustharmonize. It is very soon seen that, as accor-dance of the cognition with its object constitu-tes truth, the question now before us can onlyrelate to the formal conditions of empiricaltruth; and that the phenomenon, in oppositionto the representations of apprehension, canonly be distinguished therefrom as the object ofthem, if it is subject to a rule which distinguis-hes it from every other apprehension, andwhich renders necessary a mode of connectionof the manifold. That in the phenomenon which

contains the condition of this necessary rule ofapprehension, is the object.

Let us now proceed to our task. That somethinghappens, that is to say, that something or somestate exists which before was not, cannot beempirically perceived, unless a phenomenonprecedes, which does not contain in itself thisstate. For a reality which should follow upon avoid time, in other words, a beginning, whichno state of things precedes, can just as little beapprehended as the void time itself. Every ap-prehension of an event is therefore a perceptionwhich follows upon another perception. But asthis is the case with all synthesis of apprehen-sion, as I have shown above in the example of ahouse, my apprehension of an event is not yetsufficiently distinguished from other apprehen-sions. But I remark also that if in a phenome-non which contains an occurrence, I call theantecedent state of my perception, A, and thefollowing state, B, the perception B can only

follow A in apprehension, and the perception Acannot follow B, but only precede it. For exam-ple, I see a ship float down the stream of a ri-ver. My perception of its place lower downfollows upon my perception of its place higherup the course of the river, and it is impossiblethat, in the apprehension of this phenomenon,the vessel should be perceived first below andafterwards higher up the stream. Here, therefo-re, the order in the sequence of perceptions inapprehension is determined; and by this orderapprehension is regulated. In the former exam-ple, my perceptions in the apprehension of ahouse might begin at the roof and end at thefoundation, or vice versa; or I might apprehendthe manifold in this empirical intuition, bygoing from left to right, and from right to left.Accordingly, in the series of these perceptions,there was no determined order, which necessi-tated my beginning at a certain point, in orderempirically to connect the manifold. But thisrule is always to be met with in the perception

of that which happens, and it makes the orderof the successive perceptions in the apprehen-sion of such a phenomenon necessary.

I must, therefore, in the present case, deducethe subjective sequence of apprehension fromthe objective sequence of phenomena, for ot-herwise the former is quite undetermined, andone phenomenon is not distinguishable fromanother. The former alone proves nothing as tothe connection of the manifold in an object, forit is quite arbitrary. The latter must consist inthe order of the manifold in a phenomenon,according to which order the apprehension ofone thing (that which happens) follows that ofanother thing (which precedes), in conformitywith a rule. In this way alone can I be authori-zed to say of the phenomenon itself, and notmerely of my own apprehension, that a certainorder or sequence is to be found therein. Thatis, in other words, I cannot arrange my appre-hension otherwise than in this order.

In conformity with this rule, then, it is necessa-ry that in that which antecedes an event therebe found the condition of a rule, according towhich in this event follows always and neces-sarily; but I cannot reverse this and go backfrom the event, and determine (by apprehen-sion) that which antecedes it. For no phenome-non goes back from the succeeding point oftime to the preceding point, although it doescertainly relate to a preceding point of time;from a given time, on the other hand, there isalways a necessary progression to the determi-ned succeeding time. Therefore, because therecertainly is something that follows, I must ofnecessity connect it with something else, whichantecedes, and upon which it follows, in con-formity with a rule, that is necessarily, so thatthe event, as conditioned, affords certain indi-cation of a condition, and this condition deter-mines the event.

Let us suppose that nothing precedes an event,upon which this event must follow in conformi-ty with a rule. All sequence of perceptionwould then exist only in apprehension, that isto say, would be merely subjective, and it couldnot thereby be objectively determined whatthing ought to precede, and what ought to fo-llow in perception. In such a case, we shouldhave nothing but a play of representations,which would possess no application to anyobject. That is to say, it would not be possiblethrough perception to distinguish one pheno-menon from another, as regards relations oftime; because the succession in the act of ap-prehension would always be of the same sort,and therefore there would be nothing in thephenomenon to determine the succession, andto render a certain sequence objectively neces-sary. And, in this case, I cannot say that twostates in a phenomenon follow one upon theother, but only that one apprehension followsupon another. But this is merely subjective, and

does not determine an object, and consequentlycannot be held to be cognition of an object—noteven in the phenomenal world.

Accordingly, when we know in experience thatsomething happens, we always presupposethat something precedes, whereupon it followsin conformity with a rule. For otherwise I couldnot say of the object that it follows; because themere succession in my apprehension, if it benot determined by a rule in relation to somet-hing preceding, does not authorize successionin the object. Only, therefore, in reference to arule, according to which phenomena are de-termined in their sequence, that is, as they hap-pen, by the preceding state, can I make my sub-jective synthesis (of apprehension) objective,and it is only under this presupposition thateven the experience of an event is possible.

No doubt it appears as if this were in thoroughcontradiction to all the notions which peoplehave hitherto entertained in regard to the pro-

cedure of the human understanding. Accordingto these opinions, it is by means of the percep-tion and comparison of similar consequencesfollowing upon certain antecedent phenomenathat the understanding is led to the discoveryof a rule, according to which certain events al-ways follow certain phenomena, and it is onlyby this process that we attain to the conceptionof cause. Upon such a basis, it is clear that thisconception must be merely empirical, and therule which it furnishes us with—"Everythingthat happens must have a cause"—would bejust as contingent as experience itself. The uni-versality and necessity of the rule or law wouldbe perfectly spurious attributes of it. Indeed, itcould not possess universal validity, inasmuchas it would not in this case be a priori, butfounded on deduction. But the same is the casewith this law as with other pure a priori repre-sentations (e.g., space and time), which we candraw in perfect clearness and completenessfrom experience, only because we had already

placed them therein, and by that means, and bythat alone, had rendered experience possible.Indeed, the logical clearness of this representa-tion of a rule, determining the series of events,is possible only when we have made use the-reof in experience. Nevertheless, the recogni-tion of this rule, as a condition of the syntheti-cal unity of phenomena in time, was theground of experience itself and consequentlypreceded it a priori.

It is now our duty to show by an example thatwe never, even in experience, attribute to anobject the notion of succession or effect (of anevent—that is, the happening of something thatdid not exist before), and distinguish it fromthe subjective succession of apprehension, un-less when a rule lies at the foundation, whichcompels us to observe this order of perceptionin preference to any other, and that, indeed, itis this necessity which first renders possible therepresentation of a succession in the object.

We have representations within us, of whichalso we can be conscious. But, however widelyextended, however accurate and thorough-going this consciousness may be, these repre-sentations are still nothing more than represen-tations, that is, internal determinations of themind in this or that relation of time. Now howhappens it that to these representations weshould set an object, or that, in addition to theirsubjective reality, as modifications, we shouldstill further attribute to them a certain unk-nown objective reality? It is clear that objectivesignificancy cannot consist in a relation toanother representation (of that which we desireto term object), for in that case the questionagain arises: "How does this other representa-tion go out of itself, and obtain objective signi-ficancy over and above the subjective, which isproper to it, as a determination of a state ofmind?" If we try to discover what sort of newproperty the relation to an object gives to oursubjective representations, and what new im-

portance they thereby receive, we shall findthat this relation has no other effect than that ofrendering necessary the connection of our re-presentations in a certain manner, and of sub-jecting them to a rule; and that conversely, it isonly because a certain order is necessary in therelations of time of our representations, thatobjective significancy is ascribed to them.

In the synthesis of phenomena, the manifold ofour representations is always successive. Nowhereby is not represented an object, for bymeans of this succession, which is common toall apprehension, no one thing is distinguishedfrom another. But so soon as I perceive or as-sume that in this succession there is a relationto a state antecedent, from which the represen-tation follows in accordance with a rule, sosoon do I represent something as an event, oras a thing that happens; in other words, I cog-nize an object to which I must assign a certaindeterminate position in time, which cannot be

altered, because of the preceding state in theobject. When, therefore, I perceive that somet-hing happens, there is contained in this repre-sentation, in the first place, the fact, that somet-hing antecedes; because, it is only in relation tothis that the phenomenon obtains its properrelation of time, in other words, exists after anantecedent time, in which it did not exist. But itcan receive its determined place in time only bythe presupposition that something existed inthe foregoing state, upon which it follows ine-vitably and always, that is, in conformity with arule. From all this it is evident that, in the firstplace, I cannot reverse the order of succession,and make that which happens precede thatupon which it follows; and that, in the secondplace, if the antecedent state be posited, a cer-tain determinate event inevitably and necessa-rily follows. Hence it follows that there exists acertain order in our representations, wherebythe present gives a sure indication of some pre-viously existing state, as a correlate, though still

undetermined, of the existing event which isgiven—a correlate which itself relates to theevent as its consequence, conditions it, andconnects it necessarily with itself in the series oftime.

If then it be admitted as a necessary law of sen-sibility, and consequently a formal condition ofall perception, that the preceding necessarilydetermines the succeeding time (inasmuch as Icannot arrive at the succeeding except throughthe preceding), it must likewise be an indispen-sable law of empirical representation of theseries of time that the phenomena of the pastdetermine all phenomena in the succeedingtime, and that the latter, as events, cannot takeplace, except in so far as the former determinetheir existence in time, that is to say, establish itaccording to a rule. For it is of course only inphenomena that we can empirically cognizethis continuity in the connection of times.

For all experience and for the possibility of ex-perience, understanding is indispensable, andthe first step which it takes in this sphere is notto render the representation of objects clear, butto render the representation of an object in ge-neral, possible. It does this by applying the or-der of time to phenomena, and their existence.In other words, it assigns to each phenomenon,as a consequence, a place in relation to prece-ding phenomena, determined a priori in time,without which it could not harmonize withtime itself, which determines a place a priori toall its parts. This determination of place cannotbe derived from the relation of phenomena toabsolute time (for it is not an object of percep-tion); but, on the contrary, phenomena mustreciprocally determine the places in time of oneanother, and render these necessary in the or-der of time. In other words, whatever followsor happens, must follow in conformity with auniversal rule upon that which was containedin the foregoing state. Hence arises a series of

phenomena, which, by means of the unders-tanding, produces and renders necessary exac-tly the same order and continuous connectionin the series of our possible perceptions, as isfound a priori in the form of internal intuition(time), in which all our perceptions must haveplace.

That something happens, then, is a perceptionwhich belongs to a possible experience, whichbecomes real only because I look upon the phe-nomenon as determined in regard to its place intime, consequently as an object, which can al-ways be found by means of a rule in the con-nected series of my perceptions. But this rule ofthe determination of a thing according to suc-cession in time is as follows: "In what precedesmay be found the condition, under which anevent always (that is, necessarily) follows."From all this it is obvious that the principle ofcause and effect is the principle of possible ex-perience, that is, of objective cognition of phe-

nomena, in regard to their relations in the suc-cession of time.

The proof of this fundamental proposition restsentirely on the following momenta of argu-ment. To all empirical cognition belongs thesynthesis of the manifold by the imagination, asynthesis which is always successive, that is, inwhich the representations therein always fo-llow one another. But the order of succession inimagination is not determined, and the series ofsuccessive representations may be taken retro-gressively as well as progressively. But if thissynthesis is a synthesis of apprehension (of themanifold of a given phenomenon), then theorder is determined in the object, or to speakmore accurately, there is therein an order ofsuccessive synthesis which determines an ob-ject, and according to which something neces-sarily precedes, and when this is posited, so-mething else necessarily follows. If, then, myperception is to contain the cognition of an

event, that is, of something which really hap-pens, it must be an empirical judgement, whe-rein we think that the succession is determined;that is, it presupposes another phenomenon,upon which this event follows necessarily, or inconformity with a rule. If, on the contrary,when I posited the antecedent, the event didnot necessarily follow, I should be obliged toconsider it merely as a subjective play of myimagination, and if in this I represented to my-self anything as objective, I must look upon itas a mere dream. Thus, the relation of pheno-mena (as possible perceptions), according towhich that which happens is, as to its existence,necessarily determined in time by somethingwhich antecedes, in conformity with a rule—inother words, the relation of cause and effect—isthe condition of the objective validity of ourempirical judgements in regard to the sequenceof perceptions, consequently of their empiricaltruth, and therefore of experience. The princi-ple of the relation of causality in the succession

of phenomena is therefore valid for all objectsof experience, because it is itself the ground ofthe possibility of experience.

Here, however, a difficulty arises, which mustbe resolved. The principle of the connection ofcausality among phenomena is limited in ourformula to the succession thereof, although inpractice we find that the principle applies alsowhen the phenomena exist together in the sametime, and that cause and effect may be simulta-neous. For example, there is heat in a room,which does not exist in the open air. I lookabout for the cause, and find it to be the fire,Now the fire as the cause is simultaneous withits effect, the heat of the room. In this case,then, there is no succession as regards time,between cause and effect, but they are simulta-neous; and still the law holds good. The greaterpart of operating causes in nature are simulta-neous with their effects, and the succession intime of the latter is produced only because the

cause cannot achieve the total of its effect inone moment. But at the moment when the ef-fect first arises, it is always simultaneous withthe causality of its cause, because, if the causehad but a moment before ceased to be, the ef-fect could not have arisen. Here it must be spe-cially remembered that we must consider theorder of time and not the lapse thereof. Therelation remains, even though no time has elap-sed. The time between the causality of the cau-se and its immediate effect may entirely vanish,and the cause and effect be thus simultaneous,but the relation of the one to the other remainsalways determinable according to time. If, forexample, I consider a leaden ball, which liesupon a cushion and makes a hollow in it, as acause, then it is simultaneous with the effect.But I distinguish the two through the relationof time of the dynamical connection of both.For if I lay the ball upon the cushion, then thehollow follows upon the before smooth surface;but supposing the cushion has, from some cau-

se or another, a hollow, there does not thereu-pon follow a leaden ball.

Thus, the law of succession of time is in all ins-tances the only empirical criterion of effect inrelation to the causality of the antecedent cause.The glass is the cause of the rising of the waterabove its horizontal surface, although the twophenomena are contemporaneous. For, as soonas I draw some water with the glass from a lar-ger vessel, an effect follows thereupon, namely,the change of the horizontal state which thewater had in the large vessel into a concave,which it assumes in the glass.

This conception of causality leads us to the con-ception of action; that of action, to the concep-tion of force; and through it, to the conceptionof substance. As I do not wish this critical es-say, the sole purpose of which is to treat of thesources of our synthetical cognition a priori, tobe crowded with analyses which merely ex-plain, but do not enlarge the sphere of our con-

ceptions, I reserve the detailed explanation ofthe above conceptions for a future system ofpure reason. Such an analysis, indeed, executedwith great particularity, may already be foundin well-known works on this subject. But I can-not at present refrain from making a few re-marks on the empirical criterion of a substance,in so far as it seems to be more evident andmore easily recognized through the conceptionof action than through that of the permanenceof a phenomenon.

Where action (consequently activity and force)exists, substance also must exist, and in it alonemust be sought the seat of that fruitful sourceof phenomena. Very well. But if we are calledupon to explain what we mean by substance,and wish to avoid the vice of reasoning in acircle, the answer is by no means so easy. Howshall we conclude immediately from the actionto the permanence of that which acts, this beingnevertheless an essential and peculiar criterion

of substance (phenomenon)? But after what hasbeen said above, the solution of this questionbecomes easy enough, although by the com-mon mode of procedure—merely analysing ourconceptions—it would be quite impossible. Theconception of action indicates the relation of thesubject of causality to the effect. Now becauseall effect consists in that which happens, there-fore in the changeable, the last subject thereof isthe permanent, as the substratum of all thatchanges, that is, substance. For according to theprinciple of causality, actions are always thefirst ground of all change in phenomena and,consequently, cannot be a property of a subjectwhich itself changes, because if this were thecase, other actions and another subject wouldbe necessary to determine this change. From allthis it results that action alone, as an empiricalcriterion, is a sufficient proof of the presence ofsubstantiality, without any necessity on mypart of endeavouring to discover the perma-nence of substance by a comparison. Besides,

by this mode of induction we could not attainto the completeness which the magnitude andstrict universality of the conception requires.For that the primary subject of the causality ofall arising and passing away, all origin andextinction, cannot itself (in the sphere of phe-nomena) arise and pass away, is a sound andsafe conclusion, a conclusion which leads us tothe conception of empirical necessity and per-manence in existence, and consequently to theconception of a substance as phenomenon.

When something happens, the mere fact of theoccurrence, without regard to that which oc-curs, is an object requiring investigation. Thetransition from the non-being of a state into theexistence of it, supposing that this state con-tains no quality which previously existed in thephenomenon, is a fact of itself demanding in-quiry. Such an event, as has been shown in No.A, does not concern substance (for substancedoes not thus originate), but its condition or

state. It is therefore only change, and not originfrom nothing. If this origin be regarded as theeffect of a foreign cause, it is termed creation,which cannot be admitted as an event amongphenomena, because the very possibility of itwould annihilate the unity of experience. If,however, I regard all things not as phenomena,but as things in themselves and objects of un-derstanding alone, they, although substances,may be considered as dependent, in respect oftheir existence, on a foreign cause. But thiswould require a very different meaning in thewords, a meaning which could not apply tophenomena as objects of possible experience.

How a thing can be changed, how it is possiblethat upon one state existing in one point of ti-me, an opposite state should follow in anotherpoint of time—of this we have not the smallestconception a priori. There is requisite for thisthe knowledge of real powers, which can onlybe given empirically; for example, knowledge

of moving forces, or, in other words, of certainsuccessive phenomena (as movements) whichindicate the presence of such forces. But theform of every change, the condition underwhich alone it can take place as the coming intoexistence of another state (be the content of thechange, that is, the state which is changed,what it may), and consequently the successionof the states themselves can very well be consi-dered a priori, in relation to the law of causalityand the conditions of time.*

[*Footnote: It must be remarked that I do notspeak of the change of certain relations, but ofthe change of the state. Thus, when a body mo-ves in a uniform manner, it does not change itsstate (of motion); but only when all motion in-creases or decreases.]

When a substance passes from one state, a, intoanother state, b, the point of time in which thelatter exists is different from, and subsequent tothat in which the former existed. In like man-

ner, the second state, as reality (in the pheno-menon), differs from the first, in which the rea-lity of the second did not exist, as b from zero.That is to say, if the state, b, differs from thestate, a, only in respect to quantity, the changeis a coming into existence of b - a, which in theformer state did not exist, and in relation towhich that state is = O.

Now the question arises how a thing passesfrom one state = a, into another state = b. Bet-ween two moments there is always a certaintime, and between two states existing in thesemoments there is always a difference having acertain quantity (for all parts of phenomena arein their turn quantities). Consequently, everytransition from one state into another is alwayseffected in a time contained between two mo-ments, of which the first determines the statewhich leaves, and the second determines thestate into the thing passes. The thing leaves,and the second determines the state into which

the thing Both moments, then, are limitations ofthe time of a change, consequently of the in-termediate state between both, and as such theybelong to the total of the change. Now everychange has a cause, which evidences its causali-ty in the whole time during which the chargetakes place. The cause, therefore, does not pro-duce the change all at once or in one moment,but in a time, so that, as the time gradually in-creases from the commencing instant, a, to itscompletion at b, in like manner also, the quanti-ty of the reality (b - a) is generated through thelesser degrees which are contained between thefirst and last. All change is therefore possibleonly through a continuous action of the causali-ty, which, in so far as it is uniform, we call amomentum. The change does not consist ofthese momenta, but is generated or producedby them as their effect.

Such is the law of the continuity of all change,the ground of which is that neither time itself

nor any phenomenon in time consists of partswhich are the smallest possible, but that, not-withstanding, the state of a thing passes in theprocess of a change through all these parts, aselements, to its second state. There is no sma-llest degree of reality in a phenomenon, just asthere is no smallest degree in the quantity oftime; and so the new state of reality grows upout of the former state, through all the infinitedegrees thereof, the differences of which onefrom another, taken all together, are less thanthe difference between o and a.

It is not our business to inquire here into theutility of this principle in the investigation ofnature. But how such a proposition, which ap-pears so greatly to extend our knowledge ofnature, is possible completely a priori, is indeeda question which deserves investigation, alt-hough the first view seems to demonstrate thetruth and reality of the principle, and the ques-tion, how it is possible, may be considered su-

perfluous. For there are so many groundlesspretensions to the enlargement of our know-ledge by pure reason that we must take it as ageneral rule to be mistrustful of all such, andwithout a thoroughgoing and radical deduc-tion, to believe nothing of the sort even on theclearest dogmatical evidence.

Every addition to our empirical knowledge,and every advance made in the exercise of ourperception, is nothing more than an extensionof the determination of the internal sense, thatis to say, a progression in time, be objects them-selves what they may, phenomena, or pureintuitions. This progression in time determineseverything, and is itself determined by nothingelse. That is to say, the parts of the progressionexist only in time, and by means of the synt-hesis thereof, and are not given antecedently toit. For this reason, every transition in percep-tion to anything which follows upon another intime, is a determination of time by means of the

production of this perception. And as this de-termination of time is, always and in all itsparts, a quantity, the perception produced is tobe considered as a quantity which proceedsthrough all its degrees—no one of which is thesmallest possible—from zero up to its determi-ned degree. From this we perceive the possibili-ty of cognizing a priori a law of changes—alaw, however, which concerns their form mere-ly. We merely anticipate our own apprehen-sion, the formal condition of which, inasmuchas it is itself to be found in the mind anteceden-tly to all given phenomena, must certainly becapable of being cognized a priori.

Thus, as time contains the sensuous condition apriori of the possibility of a continuous pro-gression of that which exists to that which fo-llows it, the understanding, by virtue of theunity of apperception, contains the condition apriori of the possibility of a continuous deter-mination of the position in time of all pheno-

mena, and this by means of the series of causesand effects, the former of which necessitate thesequence of the latter, and thereby render uni-versally and for all time, and by consequence,objectively, valid the empirical cognition of therelations of time.

C. THIRD ANALOGY.

Principle of Coexistence, According to the Lawof Reciprocity or Community.

All substances, in so far as they can be percei-ved in space at the same time, exist in a state ofcomplete reciprocity of action.

PROOF.

Things are coexistent, when in empirical intui-tion the perception of the one can follow uponthe perception of the other, and vice versa—which cannot occur in the succession of phe-nomena, as we have shown in the explanationof the second principle. Thus I can perceive themoon and then the earth, or conversely, firstthe earth and then the moon; and for the reasonthat my perceptions of these objects can reci-procally follow each other, I say, they exist con-temporaneously. Now coexistence is the exis-tence of the manifold in the same time. But timeitself is not an object of perception; and therefo-re we cannot conclude from the fact that thingsare placed in the same time, the other fact, thatthe perception of these things can follow eachother reciprocally. The synthesis of the imagi-nation in apprehension would only present tous each of these perceptions as present in the

subject when the other is not present, and con-trariwise; but would not show that the objectsare coexistent, that is to say, that, if the oneexists, the other also exists in the same time,and that this is necessarily so, in order that theperceptions may be capable of following eachother reciprocally. It follows that a conceptionof the understanding or category of the recipro-cal sequence of the determinations of pheno-mena (existing, as they do, apart from each ot-her, and yet contemporaneously), is requisite tojustify us in saying that the reciprocal succes-sion of perceptions has its foundation in theobject, and to enable us to represent coexistenceas objective. But that relation of substances inwhich the one contains determinations theground of which is in the other substance, is therelation of influence. And, when this influenceis reciprocal, it is the relation of community orreciprocity. Consequently the coexistence ofsubstances in space cannot be cognized in ex-perience otherwise than under the precondition

of their reciprocal action. This is therefore thecondition of the possibility of things themselvesas objects of experience.

Things are coexistent, in so far as they exist inone and the same time. But how can we knowthat they exist in one and the same time? Onlyby observing that the order in the synthesis ofapprehension of the manifold is arbitrary and amatter of indifference, that is to say, that it canproceed from A, through B, C, D, to E, or con-trariwise from E to A. For if they were succes-sive in time (and in the order, let us suppose,which begins with A), it is quite impossible forthe apprehension in perception to begin with Eand go backwards to A, inasmuch as A belongsto past time and, therefore, cannot be an objectof apprehension.

Let us assume that in a number of substancesconsidered as phenomena each is completelyisolated, that is, that no one acts upon another.Then I say that the coexistence of these cannot

be an object of possible perception and that theexistence of one cannot, by any mode of empi-rical synthesis, lead us to the existence of anot-her. For we imagine them in this case to be se-parated by a completely void space, and thusperception, which proceeds from the one to theother in time, would indeed determine theirexistence by means of a following perception,but would be quite unable to distinguish whet-her the one phenomenon follows objectivelyupon the first, or is coexistent with it.

Besides the mere fact of existence, then, theremust be something by means of which A de-termines the position of B in time and, conver-sely, B the position of A; because only underthis condition can substances be empiricallyrepresented as existing contemporaneously.Now that alone determines the position ofanother thing in time which is the cause of it orof its determinations. Consequently every subs-tance (inasmuch as it can have succession pre-

dicated of it only in respect of its determina-tions) must contain the causality of certain de-terminations in another substance, and at thesame time the effects of the causality of the ot-her in itself. That is to say, substances muststand (mediately or immediately) in dynamicalcommunity with each other, if coexistence is tobe cognized in any possible experience. But, inregard to objects of experience, that is absolute-ly necessary without which the experience ofthese objects would itself be impossible. Con-sequently it is absolutely necessary that allsubstances in the world of phenomena, in so faras they are coexistent, stand in a relation ofcomplete community of reciprocal action toeach other.

The word community has in our language[Footnote: German] two meanings, and con-tains the two notions conveyed in the Latincommunio and commercium. We employ it inthis place in the latter sense—that of a dynami-

cal community, without which even the com-munity of place (communio spatii) could not beempirically cognized. In our experiences it iseasy to observe that it is only the continuousinfluences in all parts of space that can conductour senses from one object to another; that thelight which plays between our eyes and theheavenly bodies produces a mediating com-munity between them and us, and thereby evi-dences their coexistence with us; that we can-not empirically change our position (perceivethis change), unless the existence of matterthroughout the whole of space rendered possi-ble the perception of the positions we occupy;and that this perception can prove the contem-poraneous existence of these places onlythrough their reciprocal influence, and therebyalso the coexistence of even the most remoteobjects—although in this case the proof is onlymediate. Without community, every perception(of a phenomenon in space) is separated fromevery other and isolated, and the chain of em-

pirical representations, that is, of experience,must, with the appearance of a new object, be-gin entirely de novo, without the least connec-tion with preceding representations, and wit-hout standing towards these even in the rela-tion of time. My intention here is by no meansto combat the notion of empty space; for it mayexist where our perceptions cannot exist, inas-much as they cannot reach thereto, and where,therefore, no empirical perception of coexisten-ce takes place. But in this case it is not an objectof possible experience.

The following remarks may be useful in theway of explanation. In the mind, all phenome-na, as contents of a possible experience, mustexist in community (communio) of appercep-tion or consciousness, and in so far as it is re-quisite that objects be represented as coexistentand connected, in so far must they reciprocallydetermine the position in time of each otherand thereby constitute a whole. If this subjecti-

ve community is to rest upon an objective basis,or to be applied to substances as phenomena,the perception of one substance must renderpossible the perception of another, and conver-sely. For otherwise succession, which is alwaysfound in perceptions as apprehensions, wouldbe predicated of external objects, and their re-presentation of their coexistence be thus impos-sible. But this is a reciprocal influence, that is tosay, a real community (commercium) of subs-tances, without which therefore the empiricalrelation of coexistence would be a notion be-yond the reach of our minds. By virtue of thiscommercium, phenomena, in so far as they areapart from, and nevertheless in connection witheach other, constitute a compositum reale. Suchcomposita are possible in many different ways.The three dynamical relations then, from whichall others spring, are those of inherence, conse-quence, and composition.

These, then, are the three analogies of experien-ce. They are nothing more than principles ofthe determination of the existence of phenome-na in time, according to the three modi of thisdetermination; to wit, the relation to time itselfas a quantity (the quantity of existence, that is,duration), the relation in time as a series or suc-cession, finally, the relation in time as the com-plex of all existence (simultaneity). This unityof determination in regard to time is thorough-ly dynamical; that is to say, time is not conside-red as that in which experience determinesimmediately to every existence its position; forthis is impossible, inasmuch as absolute time isnot an object of perception, by means of whichphenomena can be connected with each other.On the contrary, the rule of the understanding,through which alone the existence of pheno-mena can receive synthetical unity as regardsrelations of time, determines for every pheno-menon its position in time, and consequently apriori, and with validity for all and every time.

By nature, in the empirical sense of the word,we understand the totality of phenomena con-nected, in respect of their existence, accordingto necessary rules, that is, laws. There are there-fore certain laws (which are moreover a priori)which make nature possible; and all empiricallaws can exist only by means of experience, andby virtue of those primitive laws throughwhich experience itself becomes possible. Thepurpose of the analogies is therefore to repre-sent to us the unity of nature in the connectionof all phenomena under certain exponents, theonly business of which is to express the relationof time (in so far as it contains all existence initself) to the unity of apperception, which canexist in synthesis only according to rules. Thecombined expression of all is this: "All pheno-mena exist in one nature, and must so exist,inasmuch as without this a priori unity, no uni-ty of experience, and consequently no determi-nation of objects in experience, is possible."

As regards the mode of proof which we haveemployed in treating of these transcendentallaws of nature, and the peculiar character of wemust make one remark, which will at the sametime be important as a guide in every otherattempt to demonstrate the truth of intellectualand likewise synthetical propositions a priori.Had we endeavoured to prove these analogiesdogmatically, that is, from conceptions; that isto say, had we employed this method in at-tempting to show that everything which exists,exists only in that which is permanent—thatevery thing or event presupposes the existenceof something in a preceding state, upon whichit follows in conformity with a rule—lastly, thatin the manifold, which is coexistent, the statescoexist in connection with each other accordingto a rule, all our labour would have been utter-ly in vain. For more conceptions of things, ana-lyse them as we may, cannot enable us to con-clude from the existence of one object to theexistence of another. What other course was left

for us to pursue? This only, to demonstrate thepossibility of experience as a cognition in whichat last all objects must be capable of being pre-sented to us, if the representation of them is topossess any objective reality. Now in this third,this mediating term, the essential form of whichconsists in the synthetical unity of the apper-ception of all phenomena, we found a prioriconditions of the universal and necessary de-termination as to time of all existences in theworld of phenomena, without which the empi-rical determination thereof as to time woulditself be impossible, and we also discoveredrules of synthetical unity a priori, by means ofwhich we could anticipate experience. For wantof this method, and from the fancy that it waspossible to discover a dogmatical proof of thesynthetical propositions which are requisite inthe empirical employment of the understan-ding, has it happened that a proof of the prin-ciple of sufficient reason has been so often at-tempted, and always in vain. The other two

analogies nobody has ever thought of, althoughthey have always been silently employed bythe mind,* because the guiding thread furnis-hed by the categories was wanting, the guidewhich alone can enable us to discover everyhiatus, both in the system of conceptions and ofprinciples.

[*Footnote: The unity of the universe, in whichall phenomena to be connected, is evidently amere consequence of the admitted principle ofthe community of all substances which are coe-xistent. For were substances isolated, theycould not as parts constitute a whole, and weretheir connection (reciprocal action of the mani-fold) not necessary from the very fact of coexis-tence, we could not conclude from the fact ofthe latter as a merely ideal relation to the for-mer as a real one. We have, however, shown inits place that community is the proper groundof the possibility of an empirical cognition ofcoexistence, and that we may therefore proper-

ly reason from the latter to the former as itscondition.]

4. THE POSTULATES OF EMPIRICALTHOUGHT.

1. That which agrees with the formal conditions(intuition and conception) of experience, is pos-sible.

2. That which coheres with the material condi-tions of experience (sensation), is real.

3. That whose coherence with the real is deter-mined according to universal conditions of ex-perience is (exists) necessary.

Explanation.

The categories of modality possess this peculia-rity, that they do not in the least determine theobject, or enlarge the conception to which theyare annexed as predicates, but only express itsrelation to the faculty of cognition. Though myconception of a thing is in itself complete, I amstill entitled to ask whether the object of it ismerely possible, or whether it is also real, or, ifthe latter, whether it is also necessary. But here-by the object itself is not more definitely deter-mined in thought, but the question is only inwhat relation it, including all its determina-tions, stands to the understanding and its em-ployment in experience, to the empirical facultyof judgement, and to the reason of its applica-tion to experience.

For this very reason, too, the categories of mo-dality are nothing more than explanations ofthe conceptions of possibility, reality, and ne-cessity, as employed in experience, and at thesame time, restrictions of all the categories to

empirical use alone, not authorizing the trans-cendental employment of them. For if they areto have something more than a merely logicalsignificance, and to be something more than amere analytical expression of the form ofthought, and to have a relation to things andtheir possibility, reality, or necessity, they mustconcern possible experience and its syntheticalunity, in which alone objects of cognition canbe given.

The postulate of the possibility of things requi-res also, that the conception of the things agreewith the formal conditions of our experience ingeneral. But this, that is to say, the objectiveform of experience, contains all the kinds ofsynthesis which are requisite for the cognitionof objects. A conception which contains a synt-hesis must be regarded as empty and, withoutreference to an object, if its synthesis does notbelong to experience—either as borrowed fromit, and in this case it is called an empirical con-

ception, or such as is the ground and a prioricondition of experience (its form), and in thiscase it is a pure conception, a conception whichnevertheless belongs to experience, inasmuchas its object can be found in this alone. Forwhere shall we find the criterion or character ofthe possibility of an object which is cogitated bymeans of an a priori synthetical conception, ifnot in the synthesis which constitutes the formof empirical cognition of objects? That in such aconception no contradiction exists is indeed anecessary logical condition, but very far frombeing sufficient to establish the objective realityof the conception, that is, the possibility of suchan object as is thought in the conception. Thus,in the conception of a figure which is containedwithin two straight lines, there is no contradic-tion, for the conceptions of two straight linesand of their junction contain no negation of afigure. The impossibility in such a case does notrest upon the conception in itself, but upon theconstruction of it in space, that is to say, upon

the conditions of space and its determinations.But these have themselves objective reality, thatis, they apply to possible things, because theycontain a priori the form of experience in gene-ral.

And now we shall proceed to point out the ex-tensive utility and influence of this postulate ofpossibility. When I represent to myself a thingthat is permanent, so that everything in itwhich changes belongs merely to its state orcondition, from such a conception alone I nevercan cognize that such a thing is possible. Or, if Irepresent to myself something which is so cons-tituted that, when it is posited, something elsefollows always and infallibly, my thought con-tains no self-contradiction; but whether such aproperty as causality is to be found in any pos-sible thing, my thought alone affords no meansof judging. Finally, I can represent to myselfdifferent things (substances) which are so cons-tituted that the state or condition of one causes

a change in the state of the other, and reciproca-lly; but whether such a relation is a property ofthings cannot be perceived from these concep-tions, which contain a merely arbitrary synt-hesis. Only from the fact, therefore, that theseconceptions express a priori the relations ofperceptions in every experience, do we knowthat they possess objective reality, that is, trans-cendental truth; and that independent of expe-rience, though not independent of all relation toform of an experience in general and its synt-hetical unity, in which alone objects can be em-pirically cognized.

But when we fashion to ourselves new concep-tions of substances, forces, action, and reaction,from the material presented to us by percep-tion, without following the example of expe-rience in their connection, we create mere chi-meras, of the possibility of which we cannotdiscover any criterion, because we have nottaken experience for our instructress, though

we have borrowed the conceptions from her.Such fictitious conceptions derive their charac-ter of possibility not, like the categories, a prio-ri, as conceptions on which all experience de-pends, but only, a posteriori, as conceptionsgiven by means of experience itself, and theirpossibility must either be cognized a posterioriand empirically, or it cannot be cognized at all.A substance which is permanently present inspace, yet without filling it (like that tertiumquid between matter and the thinking subjectwhich some have tried to introduce into me-taphysics), or a peculiar fundamental power ofthe mind of intuiting the future by anticipation(instead of merely inferring from past and pre-sent events), or, finally, a power of the mind toplace itself in community of thought with othermen, however distant they may be—these areconceptions the possibility of which has noground to rest upon. For they are not basedupon experience and its known laws; and, wit-hout experience, they are a merely arbitrary

conjunction of thoughts, which, though contai-ning no internal contradiction, has no claim toobjective reality, neither, consequently, to thepossibility of such an object as is thought inthese conceptions. As far as concerns reality, itis self-evident that we cannot cogitate such apossibility in concreto without the aid of expe-rience; because reality is concerned only withsensation, as the matter of experience, and notwith the form of thought, with which we canno doubt indulge in shaping fancies.

But I pass by everything which derives its pos-sibility from reality in experience, and I purpo-se treating here merely of the possibility ofthings by means of a priori conceptions. I main-tain, then, that the possibility of things is notderived from such conceptions per se, but onlywhen considered as formal and objective condi-tions of an experience in general.

It seems, indeed, as if the possibility of a trian-gle could be cognized from the conception of it

alone (which is certainly independent of expe-rience); for we can certainly give to the concep-tion a corresponding object completely a priori,that is to say, we can construct it. But as a trian-gle is only the form of an object, it must remaina mere product of the imagination, and the pos-sibility of the existence of an object correspon-ding to it must remain doubtful, unless we candiscover some other ground, unless we knowthat the figure can be cogitated under the con-ditions upon which all objects of experiencerest. Now, the facts that space is a formal condi-tion a priori of external experience, that theformative synthesis, by which we construct atriangle in imagination, is the very same as thatwe employ in the apprehension of a phenome-non for the purpose of making an empiricalconception of it, are what alone connect thenotion of the possibility of such a thing, withthe conception of it. In the same manner, thepossibility of continuous quantities, indeed ofquantities in general, for the conceptions of

them are without exception synthetical, is neverevident from the conceptions in themselves, butonly when they are considered as the formalconditions of the determination of objects inexperience. And where, indeed, should we lookfor objects to correspond to our conceptions, ifnot in experience, by which alone objects arepresented to us? It is, however, true that wit-hout antecedent experience we can cognize andcharacterize the possibility of things, relativelyto the formal conditions, under which somet-hing is determined in experience as an object,consequently, completely a priori. But still thisis possible only in relation to experience andwithin its limits.

The postulate concerning the cognition of thereality of things requires perception, conse-quently conscious sensation, not indeed imme-diately, that is, of the object itself, whose exis-tence is to be cognized, but still that the objecthave some connection with a real perception, in

accordance with the analogies of experience,which exhibit all kinds of real connection inexperience.

From the mere conception of a thing it is im-possible to conclude its existence. For, let theconception be ever so complete, and containinga statement of all the determinations of thething, the existence of it has nothing to do withall this, but only with thew question whethersuch a thing is given, so that the perception of itcan in every case precede the conception. Forthe fact that the conception of it precedes theperception, merely indicates the possibility ofits existence; it is perception which presentsmatter to the conception, that is the sole crite-rion of reality. Prior to the perception of thething, however, and therefore comparatively apriori, we are able to cognize its existence, pro-vided it stands in connection with some percep-tions according to the principles of the empiri-cal conjunction of these, that is, in conformity

with the analogies of perception. For, in thiscase, the existence of the supposed thing isconnected with our perception in a possibleexperience, and we are able, with the guidanceof these analogies, to reason in the series ofpossible perceptions from a thing which we doreally perceive to the thing we do not perceive.Thus, we cognize the existence of a magneticmatter penetrating all bodies from the percep-tion of the attraction of the steel-filings by themagnet, although the constitution of our organsrenders an immediate perception of this matterimpossible for us. For, according to the laws ofsensibility and the connected context of ourperceptions, we should in an experience comealso on an immediate empirical intuition of thismatter, if our senses were more acute—but thisobtuseness has no influence upon and cannotalter the form of possible experience in general.Our knowledge of the existence of things rea-ches as far as our perceptions, and what may beinferred from them according to empirical

laws, extend. If we do not set out from expe-rience, or do not proceed according to the lawsof the empirical connection of phenomena, ourpretensions to discover the existence of a thingwhich we do not immediately perceive arevain. Idealism, however, brings forward po-werful objections to these rules for provingexistence mediately. This is, therefore, the pro-per place for its refutation.

REFUTATION OF IDEALISM.

Idealism—I mean material idealism—is thetheory which declares the existence of objects inspace without us to be either () doubtful andindemonstrable, or (2) false and impossible.The first is the problematical idealism of Des-cartes, who admits the undoubted certainty ofonly one empirical assertion (assertio), to wit, "Iam." The second is the dogmatical idealism of

Berkeley, who maintains that space, togetherwith all the objects of which it is the inseparablecondition, is a thing which is in itself impossi-ble, and that consequently the objects in spaceare mere products of the imagination. Thedogmatical theory of idealism is unavoidable, ifwe regard space as a property of things inthemselves; for in that case it is, with all towhich it serves as condition, a nonentity. Butthe foundation for this kind of idealism wehave already destroyed in the transcendentalaesthetic. Problematical idealism, which makesno such assertion, but only alleges our incapaci-ty to prove the existence of anything besidesourselves by means of immediate experience, isa theory rational and evidencing a thoroughand philosophical mode of thinking, for it ob-serves the rule not to form a decisive judge-ment before sufficient proof be shown. Thedesired proof must therefore demonstrate thatwe have experience of external things, and notmere fancies. For this purpose, we must prove,

that our internal and, to Descartes, indubitableexperience is itself possible only under the pre-vious assumption of external experience.

THEOREM.

The simple but empirically determined cons-ciousness of my own existence proves the exis-tence of external objects in space.

PROOF

I am conscious of my own existence as deter-mined in time. All determination in regard totime presupposes the existence of somethingpermanent in perception. But this permanentsomething cannot be something in me, for thevery reason that my existence in time is itself

determined by this permanent something. Itfollows that the perception of this permanentexistence is possible only through a thing wit-hout me and not through the mere representa-tion of a thing without me. Consequently, thedetermination of my existence in time is possi-ble only through the existence of real thingsexternal to me. Now, consciousness in time isnecessarily connected with the consciousness ofthe possibility of this determination in time.Hence it follows that consciousness in time isnecessarily connected also with the existence ofthings without me, inasmuch as the existence ofthese things is the condition of determination intime. That is to say, the consciousness of myown existence is at the same time an immediateconsciousness of the existence of other thingswithout me.

Remark I. The reader will observe, that in theforegoing proof the game which idealism playsis retorted upon itself, and with more justice. It

assumed that the only immediate experience isinternal and that from this we can only infer theexistence of external things. But, as alwayshappens, when we reason from given effects todetermined causes, idealism has reasoned withtoo much haste and uncertainty, for it is quitepossible that the cause of our representationsmay lie in ourselves, and that we ascribe it fal-sely to external things. But our proof showsthat external experience is properly immedia-te,* that only by virtue of it—not, indeed, theconsciousness of our own existence, but cer-tainly the determination of our existence intime, that is, internal experience—is possible. Itis true, that the representation "I am," which isthe expression of the consciousness which canaccompany all my thoughts, is that which im-mediately includes the existence of a subject.But in this representation we cannot find anyknowledge of the subject, and therefore also noempirical knowledge, that is, experience. Forexperience contains, in addition to the thought

of something existing, intuition, and in this caseit must be internal intuition, that is, time, inrelation to which the subject must be determi-ned. But the existence of external things is abso-lutely requisite for this purpose, so that it fo-llows that internal experience is itself possibleonly mediately and through external experien-ce.

[*Footnote: The immediate consciousness of theexistence of external things is, in the precedingtheorem, not presupposed, but proved, by thepossibility of this consciousness understood byus or not. The question as to the possibility of itwould stand thus: "Have we an internal sense,but no external sense, and is our belief in exter-nal perception a mere delusion?" But it is evi-dent that, in order merely to fancy to ourselvesanything as external, that is, to present it to thesense in intuition we must already possess anexternal sense, and must thereby distinguishimmediately the mere receptivity of an external

intuition from the spontaneity which characte-rizes every act of imagination. For merely toimagine also an external sense, would annihila-te the faculty of intuition itself which is to bedetermined by the imagination.]

Remark II. Now with this view all empiricaluse of our faculty of cognition in the determina-tion of time is in perfect accordance. Its truth issupported by the fact that it is possible to per-ceive a determination of time only by means ofa change in external relations (motion) to thepermanent in space (for example, we becomeaware of the sun's motion by observing thechanges of his relation to the objects of thisearth). But this is not all. We find that we pos-sess nothing permanent that can correspondand be submitted to the conception of a subs-tance as intuition, except matter. This idea ofpermanence is not itself derived from externalexperience, but is an a priori necessary condi-tion of all determination of time, consequently

also of the internal sense in reference to ourown existence, and that through the existenceof external things. In the representation "I," theconsciousness of myself is not an intuition, buta merely intellectual representation producedby the spontaneous activity of a thinking sub-ject. It follows, that this "I" has not any predica-te of intuition, which, in its character of perma-nence, could serve as correlate to the determi-nation of time in the internal sense—in the sa-me way as impenetrability is the correlate ofmatter as an empirical intuition.

Remark III. From the fact that the existence ofexternal things is a necessary condition of thepossibility of a determined consciousness ofourselves, it does not follow that every intuitiverepresentation of external things involves theexistence of these things, for their representa-tions may very well be the mere products of theimagination (in dreams as well as in madness);though, indeed, these are themselves created

by the reproduction of previous external per-ceptions, which, as has been shown, are possi-ble only through the reality of external objects.The sole aim of our remarks has, however, beento prove that internal experience in general ispossible only through external experience ingeneral. Whether this or that supposed expe-rience be purely imaginary must be discoveredfrom its particular determinations and by com-paring these with the criteria of all real expe-rience.

Finally, as regards the third postulate, it appliesto material necessity in existence, and not tomerely formal and logical necessity in the con-nection of conceptions. Now as we cannot cog-nize completely a priori the existence of anyobject of sense, though we can do so compara-tively a priori, that is, relatively to some otherpreviously given existence—a cognition, howe-ver, which can only be of such an existence asmust be contained in the complex of experien-

ce, of which the previously given perception isa part—the necessity of existence can never becognized from conceptions, but always, on thecontrary, from its connection with that which isan object of perception. But the only existencecognized, under the condition of other givenphenomena, as necessary, is the existence ofeffects from given causes in conformity withthe laws of causality. It is consequently not thenecessity of the existence of things (as substan-ces), but the necessity of the state of things thatwe cognize, and that not immediately, but bymeans of the existence of other states given inperception, according to empirical laws of cau-sality. Hence it follows that the criterion of ne-cessity is to be found only in the law of possibleexperience—that everything which happens isdetermined a priori in the phenomenon by itscause. Thus we cognize only the necessity ofeffects in nature, the causes of which are givenus. Moreover, the criterion of necessity in exis-tence possesses no application beyond the field

of possible experience, and even in this it is notvalid of the existence of things as substances,because these can never be considered as empi-rical effects, or as something that happens andhas a beginning. Necessity, therefore, regardsonly the relations of phenomena according tothe dynamical law of causality, and the possibi-lity grounded thereon, of reasoning from somegiven existence (of a cause) a priori to anotherexistence (of an effect). "Everything that hap-pens is hypothetically necessary," is a principlewhich subjects the changes that take place inthe world to a law, that is, to a rule of necessaryexistence, without which nature herself couldnot possibly exist. Hence the proposition, "Not-hing happens by blind chance (in mundo nondatur casus)," is an a priori law of nature. Thecase is the same with the proposition, "Necessi-ty in nature is not blind," that is, it is conditio-ned, consequently intelligible necessity (nondatur fatum). Both laws subject the play ofchange to "a nature of things (as phenomena),"

or, which is the same thing, to the unity of theunderstanding, and through the understandingalone can changes belong to an experience, asthe synthetical unity of phenomena. Both be-long to the class of dynamical principles. Theformer is properly a consequence of the princi-ple of causality—one of the analogies of expe-rience. The latter belongs to the principles ofmodality, which to the determination of causa-lity adds the conception of necessity, which isitself, however, subject to a rule of the unders-tanding. The principle of continuity forbids anyleap in the series of phenomena regarded aschanges (in mundo non datur saltus); and li-kewise, in the complex of all empirical intui-tions in space, any break or hiatus between twophenomena (non datur hiatus)—for we can soexpress the principle, that experience can admitnothing which proves the existence of a va-cuum, or which even admits it as a part of anempirical synthesis. For, as regards a vacuumor void, which we may cogitate as out and be-

yond the field of possible experience (theworld), such a question cannot come before thetribunal of mere understanding, which decidesonly upon questions that concern the employ-ment of given phenomena for the constructionof empirical cognition. It is rather a problem forideal reason, which passes beyond the sphereof a possible experience and aims at forming ajudgement of that which surrounds and cir-cumscribes it, and the proper place for the con-sideration of it is the transcendental dialectic.These four propositions, "In mundo non daturhiatus, non datur saltus, non datur casus, nondatur fatum," as well as all principles of trans-cendental origin, we could very easily exhibitin their proper order, that is, in conformity withthe order of the categories, and assign to eachits proper place. But the already practised rea-der will do this for himself, or discover the clueto such an arrangement. But the combined re-sult of all is simply this, to admit into the empi-rical synthesis nothing which might cause a

break in or be foreign to the understanding andthe continuous connection of all phenomena,that is, the unity of the conceptions of the un-derstanding. For in the understanding alone isthe unity of experience, in which all percep-tions must have their assigned place, possible.

Whether the field of possibility be greater thanthat of reality, and whether the field of the lat-ter be itself greater than that of necessity, areinteresting enough questions, and quite capableof synthetic solution, questions, however,which come under the jurisdiction of reasonalone. For they are tantamount to asking whet-her all things as phenomena do without excep-tion belong to the complex and connected who-le of a single experience, of which every givenperception is a part which therefore cannot beconjoined with any other phenomena—or,whether my perceptions can belong to morethan one possible experience? The understan-ding gives to experience, according to the sub-

jective and formal conditions, of sensibility aswell as of apperception, the rules which alonemake this experience possible. Other forms ofintuition besides those of space and time, otherforms of understanding besides the discursiveforms of thought, or of cognition by means ofconceptions, we can neither imagine nor makeintelligible to ourselves; and even if we could,they would still not belong to experience,which is the only mode of cognition by whichobjects are presented to us. Whether other per-ceptions besides those which belong to the totalof our possible experience, and consequentlywhether some other sphere of matter exists, theunderstanding has no power to decide, its pro-per occupation being with the synthesis of thatwhich is given. Moreover, the poverty of theusual arguments which go to prove the existen-ce of a vast sphere of possibility, of which allthat is real (every object of experience) is but asmall part, is very remarkable. "All real is pos-sible"; from this follows naturally, according to

the logical laws of conversion, the particularproposition: "Some possible is real." Now thisseems to be equivalent to: "Much is possiblethat is not real." No doubt it does seem as if weought to consider the sum of the possible to begreater than that of the real, from the fact thatsomething must be added to the former toconstitute the latter. But this notion of addingto the possible is absurd. For that which is notin the sum of the possible, and consequentlyrequires to be added to it, is manifestly impos-sible. In addition to accordance with the formalconditions of experience, the understandingrequires a connection with some perception;but that which is connected with this percep-tion is real, even although it is not immediatelyperceived. But that another series of phenome-na, in complete coherence with that which isgiven in perception, consequently more thanone all-embracing experience is possible, is aninference which cannot be concluded from thedata given us by experience, and still less wit-

hout any data at all. That which is possible onlyunder conditions which are themselves merelypossible, is not possible in any respect. And yetwe can find no more certain ground on whichto base the discussion of the question whetherthe sphere of possibility is wider than that ofexperience.

I have merely mentioned these questions, thatin treating of the conception of the understan-ding, there might be no omission of anythingthat, in the common opinion, belongs to them.In reality, however, the notion of absolute pos-sibility (possibility which is valid in every res-pect) is not a mere conception of the unders-tanding, which can be employed empirically,but belongs to reason alone, which passes thebounds of all empirical use of the understan-ding. We have, therefore, contented ourselveswith a merely critical remark, leaving the sub-ject to be explained in the sequel.

Before concluding this fourth section, and atthe same time the system of all principles of thepure understanding, it seems proper to men-tion the reasons which induced me to term theprinciples of modality postulates. This expres-sion I do not here use in the sense which somemore recent philosophers, contrary to its mea-ning with mathematicians, to whom the wordproperly belongs, attach to it—that of a propo-sition, namely, immediately certain, requiringneither deduction nor proof. For if, in the caseof synthetical propositions, however evidentthey may be, we accord to them without deduc-tion, and merely on the strength of their ownpretensions, unqualified belief, all critique ofthe understanding is entirely lost; and, as thereis no want of bold pretensions, which thecommon belief (though for the philosopher thisis no credential) does not reject, the understan-ding lies exposed to every delusion and conceit,without the power of refusing its assent to tho-se assertions, which, though illegitimate, de-

mand acceptance as veritable axioms. When,therefore, to the conception of a thing an a prio-ri determination is synthetically added, such aproposition must obtain, if not a proof, at leasta deduction of the legitimacy of its assertion.

The principles of modality are, however, notobjectively synthetical, for the predicates ofpossibility, reality, and necessity do not in theleast augment the conception of that of whichthey are affirmed, inasmuch as they contributenothing to the representation of the object. Butas they are, nevertheless, always synthetical,they are so merely subjectively. That is to say,they have a reflective power, and apply to theconception of a thing, of which, in other res-pects, they affirm nothing, the faculty of cogni-tion in which the conception originates and hasits seat. So that if the conception merely agreewith the formal conditions of experience, itsobject is called possible; if it is in connectionwith perception, and determined thereby, the

object is real; if it is determined according toconceptions by means of the connection of per-ceptions, the object is called necessary. Theprinciples of modality therefore predicate of aconception nothing more than the procedure ofthe faculty of cognition which generated it.Now a postulate in mathematics is a practicalproposition which contains nothing but thesynthesis by which we present an object to our-selves, and produce the conception of it, forexample—"With a given line, to describe a cir-cle upon a plane, from a given point"; and sucha proposition does not admit of proof, becausethe procedure, which it requires, is exactly thatby which alone it is possible to generate theconception of such a figure. With the sameright, accordingly, can we postulate the princi-ples of modality, because they do not augment*the conception of a thing but merely indicatethe manner in which it is connected with thefaculty of cognition.

[*Footnote: When I think the reality of a thing, Ido really think more than the possibility, butnot in the thing; for that can never contain morein reality than was contained in its completepossibility. But while the notion of possibility ismerely the notion of a position of thing in rela-tion to the understanding (its empirical use),reality is the conjunction of the thing with per-ception.]

GENERAL REMARK ON THE SYSTEMOF PRINCIPLES.

It is very remarkable that we cannot perceivethe possibility of a thing from the category alo-ne, but must always have an intuition, bywhich to make evident the objective reality ofthe pure conception of the understanding. Ta-ke, for example, the categories of relation. How(1) a thing can exist only as a subject, and not as

a mere determination of other things, that is,can be substance; or how (2), because somet-hing exists, some other thing must exist, conse-quently how a thing can be a cause; or how (3),when several things exist, from the fact that oneof these things exists, some consequence to theothers follows, and reciprocally, and in thisway a community of substances can be possi-ble—are questions whose solution cannot beobtained from mere conceptions. The very sa-me is the case with the other categories; forexample, how a thing can be of the same sortwith many others, that is, can be a quantity,and so on. So long as we have not intuition wecannot know whether we do really think anobject by the categories, and where an objectcan anywhere be found to cohere with them,and thus the truth is established, that the cate-gories are not in themselves cognitions, butmere forms of thought for the construction ofcognitions from given intuitions. For the samereason is it true that from categories alone no

synthetical proposition can be made. Forexample: "In every existence there is substan-ce," that is, something that can exist only as asubject and not as mere predicate; or, "Everyt-hing is a quantity"—to construct propositionssuch as these, we require something to enableus to go out beyond the given conception andconnect another with it. For the same reasonthe attempt to prove a synthetical propositionby means of mere conceptions, for example:"Everything that exists contingently has a cau-se," has never succeeded. We could never getfurther than proving that, without this relationto conceptions, we could not conceive the exis-tence of the contingent, that is, could not a prio-ri through the understanding cognize the exis-tence of such a thing; but it does not hence fo-llow that this is also the condition of the possi-bility of the thing itself that is said to be contin-gent. If, accordingly; we look back to our proofof the principle of causality, we shall find thatwe were able to prove it as valid only of objects

of possible experience, and, indeed, only asitself the principle of the possibility of expe-rience, Consequently of the cognition of an ob-ject given in empirical intuition, and not frommere conceptions. That, however, the proposi-tion: "Everything that is contingent must have acause," is evident to every one merely fromconceptions, is not to be denied. But in this casethe conception of the contingent is cogitated asinvolving not the category of modality (as thatthe non-existence of which can be conceived)but that of relation (as that which can exist onlyas the consequence of something else), and so itis really an identical proposition: "That whichcan exist only as a consequence, has a cause." Infact, when we have to give examples of contin-gent existence, we always refer to changes, andnot merely to the possibility of conceiving theopposite.* But change is an event, which, assuch, is possible only through a cause, and con-sidered per se its non-existence is thereforepossible, and we become cognizant of its con-

tingency from the fact that it can exist only asthe effect of a cause. Hence, if a thing is assu-med to be contingent, it is an analytical propo-sition to say, it has a cause.

[*Footnote: We can easily conceive the non-existence of matter; but the ancients did notthence infer its contingency. But even the alter-nation of the existence and non-existence of agiven state in a thing, in which all change con-sists, by no means proves the contingency ofthat state—the ground of proof being the reali-ty of its opposite. For example, a body is in astate of rest after motion, but we cannot inferthe contingency of the motion from the fact thatthe former is the opposite of the latter. For thisopposite is merely a logical and not a real op-posite to the other. If we wish to demonstratethe contingency of the motion, what we oughtto prove is that, instead of the motion whichtook place in the preceding point of time, it waspossible for the body to have been then in rest,

not, that it is afterwards in rest; for in this case,both opposites are perfectly consistent witheach other.]

But it is still more remarkable that, to unders-tand the possibility of things according to thecategories and thus to demonstrate the objecti-ve reality of the latter, we require not merelyintuitions, but external intuitions. If, for exam-ple, we take the pure conceptions of relation,we find that (1) for the purpose of presenting tothe conception of substance something perma-nent in intuition corresponding thereto andthus of demonstrating the objective reality ofthis conception, we require an intuition (of mat-ter) in space, because space alone is permanentand determines things as such, while time, andwith it all that is in the internal sense, is in astate of continual flow; (2) in order to representchange as the intuition corresponding to theconception of causality, we require the repre-sentation of motion as change in space; in fact,

it is through it alone that changes, the possibili-ty of which no pure understanding can percei-ve, are capable of being intuited. Change is theconnection of determinations contradictorilyopposed to each other in the existence of oneand the same thing. Now, how it is possiblethat out of a given state one quite opposite to itin the same thing should follow, reason wit-hout an example can not only not conceive, butcannot even make intelligible without intuition;and this intuition is the motion of a point inspace; the existence of which in different spaces(as a consequence of opposite determinations)alone makes the intuition of change possible.For, in order to make even internal change cog-nitable, we require to represent time, as theform of the internal sense, figuratively by aline, and the internal change by the drawing ofthat line (motion), and consequently are obli-ged to employ external intuition to be able torepresent the successive existence of ourselvesin different states. The proper ground of this

fact is that all change to be perceived as changepresupposes something permanent in intuition,while in the internal sense no permanent intui-tion is to be found. Lastly, the objective possibi-lity of the category of community cannot beconceived by mere reason, and consequently itsobjective reality cannot be demonstrated wit-hout an intuition, and that external in space.For how can we conceive the possibility ofcommunity, that is, when several substancesexist, that some effect on the existence of theone follows from the existence of the other, andreciprocally, and therefore that, because somet-hing exists in the latter, something else mustexist in the former, which could not be unders-tood from its own existence alone? For this isthe very essence of community—which is in-conceivable as a property of things which areperfectly isolated. Hence, Leibnitz, in attribu-ting to the substances of the world—as cogita-ted by the understanding alone—a community,required the mediating aid of a divinity; for,

from their existence, such a property seemed tohim with justice inconceivable. But we can veryeasily conceive the possibility of community (ofsubstances as phenomena) if we represent themto ourselves as in space, consequently in exter-nal intuition. For external intuition contains initself a priori formal external relations, as theconditions of the possibility of the real relationsof action and reaction, and therefore of the pos-sibility of community. With the same ease can itbe demonstrated, that the possibility of thingsas quantities, and consequently the objectivereality of the category of quantity, can begrounded only in external intuition, and that byits means alone is the notion of quantity appro-priated by the internal sense. But I must avoidprolixity, and leave the task of illustrating thisby examples to the reader's own reflection.

The above remarks are of the greatest impor-tance, not only for the confirmation of our pre-vious confutation of idealism, but still more

when the subject of self-cognition by mere in-ternal consciousness and the determination ofour own nature without the aid of external em-pirical intuitions is under discussion, for theindication of the grounds of the possibility ofsuch a cognition.

The result of the whole of this part of the analy-tic of principles is, therefore: "All principles ofthe pure understanding are nothing more thana priori principles of the possibility of experien-ce, and to experience alone do all a priori synt-hetical propositions apply and relate"; indeed,their possibility itself rests entirely on this rela-tion.

CHAPTER III Of the Ground of the Divi-sion of all Objects into Phenomena and Nou-mena.

We have now not only traversed the region ofthe pure understanding and carefully surveyedevery part of it, but we have also measured it,and assigned to everything therein its properplace. But this land is an island, and enclosedby nature herself within unchangeable limits. Itis the land of truth (an attractive word), su-rrounded by a wide and stormy ocean, the re-gion of illusion, where many a fog-bank, manyan iceberg, seems to the mariner, on his voyageof discovery, a new country, and, while cons-tantly deluding him with vain hopes, engageshim in dangerous adventures, from which henever can desist, and which yet he never canbring to a termination. But before venturingupon this sea, in order to explore it in its wholeextent, and to arrive at a certainty whetheranything is to be discovered there, it will not be

without advantage if we cast our eyes upon thechart of the land that we are about to leave, andto ask ourselves, firstly, whether we cannot restperfectly contented with what it contains, orwhether we must not of necessity be contentedwith it, if we can find nowhere else a solidfoundation to build upon; and, secondly, bywhat title we possess this land itself, and howwe hold it secure against all hostile claims?Although, in the course of our analytic, wehave already given sufficient answers to thesequestions, yet a summary recapitulation of the-se solutions may be useful in strengthening ourconviction, by uniting in one point the momen-ta of the arguments.

We have seen that everything which the un-derstanding draws from itself, without borro-wing from experience, it nevertheless possessesonly for the behoof and use of experience. Theprinciples of the pure understanding, whetherconstitutive a priori (as the mathematical prin-

ciples), or merely regulative (as the dynamical),contain nothing but the pure schema, as it we-re, of possible experience. For experience pos-sesses its unity from the synthetical unitywhich the understanding, originally and fromitself, imparts to the synthesis of the imagina-tion in relation to apperception, and in a priorirelation to and agreement with which pheno-mena, as data for a possible cognition, muststand. But although these rules of the unders-tanding are not only a priori true, but the verysource of all truth, that is, of the accordance ofour cognition with objects, and on this ground,that they contain the basis of the possibility ofexperience, as the ensemble of all cognition, itseems to us not enough to propound what istrue—we desire also to be told what we want toknow. If, then, we learn nothing more by thiscritical examination than what we should havepractised in the merely empirical use of theunderstanding, without any such subtle inqui-ry, the presumption is that the advantage we

reap from it is not worth the labour bestowedupon it. It may certainly be answered that norash curiosity is more prejudicial to the enlar-gement of our knowledge than that which mustknow beforehand the utility of this or that pieceof information which we seek, before we haveentered on the needful investigations, and befo-re one could form the least conception of itsutility, even though it were placed before oureyes. But there is one advantage in such trans-cendental inquiries which can be made com-prehensible to the dullest and most reluctantlearner—this, namely, that the understandingwhich is occupied merely with empirical exer-cise, and does not reflect on the sources of itsown cognition, may exercise its functions verywell and very successfully, but is quite unableto do one thing, and that of very great impor-tance, to determine, namely, the bounds thatlimit its employment, and to know what lieswithin or without its own sphere. This purposecan be obtained only by such profound investi-

gations as we have instituted. But if it cannotdistinguish whether certain questions lie withinits horizon or not, it can never be sure either asto its claims or possessions, but must lay itsaccount with many humiliating corrections,when it transgresses, as it unavoidably will, thelimits of its own territory, and loses itself infanciful opinions and blinding illusions.

That the understanding, therefore, cannot makeof its a priori principles, or even of its concep-tions, other than an empirical use, is a proposi-tion which leads to the most important results.A transcendental use is made of a conception ina fundamental proposition or principle, when itis referred to things in general and consideredas things in themselves; an empirical use, whenit is referred merely to phenomena, that is, toobjects of a possible experience. That the latteruse of a conception is the only admissible one isevident from the reasons following. For everyconception are requisite, firstly, the logical form

of a conception (of thought) general; and, se-condly, the possibility of presenting to this anobject to which it may apply. Failing this latter,it has no sense, and utterly void of content, alt-hough it may contain the logical function forconstructing a conception from certain data.Now, object cannot be given to a conceptionotherwise than by intuition, and, even if a pureintuition antecedent to the object is a priori pos-sible, this pure intuition can itself obtain objec-tive validity only from empirical intuition, ofwhich it is itself but the form. All conceptions,therefore, and with them all principles, howe-ver high the degree of their a priori possibility,relate to empirical intuitions, that is, to datatowards a possible experience. Without thisthey possess no objective validity, but are mereplay of imagination or of understanding withimages or notions. Let us take, for example, theconceptions of mathematics, and first in its pu-re intuitions. "Space has three dimensions"—"Between two points there can be only one

straight line," etc. Although all these principles,and the representation of the object with whichthis science occupies itself, are generated in themind entirely a priori, they would neverthelesshave no significance if we were not always ableto exhibit their significance in and by means ofphenomena (empirical objects). Hence it is re-quisite that an abstract conception be madesensuous, that is, that an object correspondingto it in intuition be forthcoming, otherwise theconception remains, as we say, without sense,that is, without meaning. Mathematics fulfilsthis requirement by the construction of the fi-gure, which is a phenomenon evident to thesenses. The same science finds support andsignificance in number; this in its turn finds itin the fingers, or in counters, or in lines andpoints. The conception itself is always produ-ced a priori, together with the synthetical prin-ciples or formulas from such conceptions; butthe proper employment of them, and their ap-plication to objects, can exist nowhere but in

experience, the possibility of which, as regardsits form, they contain a priori.

That this is also the case with all of the catego-ries and the principles based upon them is evi-dent from the fact that we cannot render inte-lligible the possibility of an object correspon-ding to them without having recourse to theconditions of sensibility, consequently, to theform of phenomena, to which, as their onlyproper objects, their use must therefore be con-fined, inasmuch as, if this condition is removed,all significance, that is, all relation to an object,disappears, and no example can be found tomake it comprehensible what sort of things weought to think under such conceptions.

The conception of quantity cannot be explainedexcept by saying that it is the determination ofa thing whereby it can be cogitated how manytimes one is placed in it. But this "how manytimes" is based upon successive repetition, con-sequently upon time and the synthesis of the

homogeneous therein. Reality, in contradistinc-tion to negation, can be explained only by cogi-tating a time which is either filled therewith oris void. If I leave out the notion of permanence(which is existence in all time), there remains inthe conception of substance nothing but thelogical notion of subject, a notion of which Iendeavour to realize by representing to myselfsomething that can exist only as a subject. Butnot only am I perfectly ignorant of any condi-tions under which this logical prerogative canbelong to a thing, I can make nothing out of thenotion, and draw no inference from it, becauseno object to which to apply the conception isdetermined, and we consequently do not knowwhether it has any meaning at all. In like man-ner, if I leave out the notion of time, in whichsomething follows upon some other thing inconformity with a rule, I can find nothing in thepure category, except that there is a somethingof such a sort that from it a conclusion may bedrawn as to the existence of some other thing.

But in this case it would not only be impossibleto distinguish between a cause and an effect,but, as this power to draw conclusions requiresconditions of which I am quite ignorant, theconception is not determined as to the mode inwhich it ought to apply to an object. The so-called principle: "Everything that is contingenthas a cause," comes with a gravity and self-assumed authority that seems to require nosupport from without. But, I ask, what is meantby contingent? The answer is that the non-existence of which is possible. But I should likevery well to know by what means this possibili-ty of non-existence is to be cognized, if we donot represent to ourselves a succession in theseries of phenomena, and in this succession anexistence which follows a non-existence, orconversely, consequently, change. For to say,that the non-existence of a thing is not self-contradictory is a lame appeal to a logical con-dition, which is no doubt a necessary conditionof the existence of the conception, but is far

from being sufficient for the real objective pos-sibility of non-existence. I can annihilate inthought every existing substance without self-contradiction, but I cannot infer from this theirobjective contingency in existence, that is tosay, the possibility of their non-existence initself. As regards the category of community, itmay easily be inferred that, as the pure catego-ries of substance and causality are incapable ofa definition and explanation sufficient to de-termine their object without the aid of intuition,the category of reciprocal causality in the rela-tion of substances to each other (commercium)is just as little susceptible thereof. Possibility,existence, and necessity nobody has ever yetbeen able to explain without being guilty ofmanifest tautology, when the definition hasbeen drawn entirely from the pure understan-ding. For the substitution of the logical possibi-lity of the conception—the condition of whichis that it be not self-contradictory, for the trans-cendental possibility of things—the condition

of which is that there be an object correspon-ding to the conception, is a trick which can onlydeceive the inexperienced.*

[*Footnote: In one word, to none of these con-ceptions belongs a corresponding object, andconsequently their real possibility cannot bedemonstrated, if we take away sensuous intui-tion—the only intuition which we possess—and there then remains nothing but the logicalpossibility, that is, the fact that the conceptionor thought is possible—which, however, is notthe question; what we want to know being,whether it relates to an object and thus posses-ses any meaning.]

It follows incontestably, that the pure concep-tions of the understanding are incapable oftranscendental, and must always be of empiri-cal use alone, and that the principles of the pu-re understanding relate only to the generalconditions of a possible experience, to objects of

the senses, and never to things in general, apartfrom the mode in which we intuite them.

Transcendental analytic has accordingly thisimportant result, to wit, that the understandingis competent' effect nothing a priori, except theanticipation of the form of a possible experiencein general, and that, as that which is not phe-nomenon cannot be an object of experience, itcan never overstep the limits of sensibility, wit-hin which alone objects are presented to us. Itsprinciples are merely principles of the exposi-tion of phenomena, and the proud name of anontology, which professes to present syntheti-cal cognitions a priori of things in general in asystematic doctrine, must give place to the mo-dest title of analytic of the pure understanding.

Thought is the act of referring a given intuitionto an object. If the mode of this intuition is unk-nown to us, the object is merely transcendental,and the conception of the understanding is em-ployed only transcendentally, that is, to produ-

ce unity in the thought of a manifold in general.Now a pure category, in which all conditions ofsensuous intuition—as the only intuition wepossess—are abstracted, does not determine anobject, but merely expresses the thought of anobject in general, according to different modes.Now, to employ a conception, the function ofjudgement is required, by which an object issubsumed under the conception, consequentlythe at least formal condition, under which so-mething can be given in intuition. Failing thiscondition of judgement (schema), subsumptionis impossible; for there is in such a case nothinggiven, which may be subsumed under the con-ception. The merely transcendental use of thecategories is therefore, in fact, no use at all andhas no determined, or even, as regards its form,determinable object. Hence it follows that thepure category is incompetent to establish asynthetical a priori principle, and that the prin-ciples of the pure understanding are only ofempirical and never of transcendental use, and

that beyond the sphere of possible experienceno synthetical a priori principles are possible.

It may be advisable, therefore, to express our-selves thus. The pure categories, apart from theformal conditions of sensibility, have a merelytranscendental meaning, but are neverthelessnot of transcendental use, because this is initself impossible, inasmuch as all the conditionsof any employment or use of them (in judge-ments) are absent, to wit, the formal conditionsof the subsumption of an object under theseconceptions. As, therefore, in the character ofpure categories, they must be employed empi-rically, and cannot be employed transcendenta-lly, they are of no use at all, when separatedfrom sensibility, that is, they cannot be appliedto an object. They are merely the pure form ofthe employment of the understanding in res-pect of objects in general and of thought, wit-hout its being at the same time possible to thinkor to determine any object by their means. But

there lurks at the foundation of this subject anillusion which it is very difficult to avoid. Thecategories are not based, as regards their origin,upon sensibility, like the forms of intuition,space, and time; they seem, therefore, to be ca-pable of an application beyond the sphere ofsensuous objects. But this is not the case. Theyare nothing but mere forms of thought, whichcontain only the logical faculty of uniting apriori in consciousness the manifold given inintuition. Apart, then, from the only intuitionpossible for us, they have still less meaningthan the pure sensuous forms, space and time,for through them an object is at least given,while a mode of connection of the manifold,when the intuition which alone gives the mani-fold is wanting, has no meaning at all. At thesame time, when we designate certain objectsas phenomena or sensuous existences, thusdistinguishing our mode of intuiting them fromtheir own nature as things in themselves, it isevident that by this very distinction we as it

were place the latter, considered in this theirown nature, although we do not so intuitethem, in opposition to the former, or, on theother hand, we do so place other possiblethings, which are not objects of our senses, butare cogitated by the understanding alone, andcall them intelligible existences (noumena).Now the question arises whether the pure con-ceptions of our understanding do possess signi-ficance in respect of these latter, and may pos-sibly be a mode of cognizing them.

But we are met at the very commencement withan ambiguity, which may easily occasion greatmisapprehension. The understanding, when itterms an object in a certain relation phenome-non, at the same time forms out of this relationa representation or notion of an object in itself,and hence believes that it can form also concep-tions of such objects. Now as the understandingpossesses no other fundamental conceptionsbesides the categories, it takes for granted that

an object considered as a thing in itself must becapable of being thought by means of thesepure conceptions, and is thereby led to hold theperfectly undetermined conception of an inte-lligible existence, a something out of the sphereof our sensibility, for a determinate conceptionof an existence which we can cognize in someway or other by means of the understanding.

If, by the term noumenon, we understand athing so far as it is not an object of our sensuousintuition, thus making abstraction of our modeof intuiting it, this is a noumenon in the negati-ve sense of the word. But if we understand by itan object of a non-sensuous intuition, we in thiscase assume a peculiar mode of intuition, anintellectual intuition, to wit, which does not,however, belong to us, of the very possibility ofwhich we have no notion—and this is a nou-menon in the positive sense.

The doctrine of sensibility is also the doctrine ofnoumena in the negative sense, that is, of

things which the understanding is obliged tocogitate apart from any relation to our mode ofintuition, consequently not as mere phenome-na, but as things in themselves. But the unders-tanding at the same time comprehends that itcannot employ its categories for the considera-tion of things in themselves, because these pos-sess significance only in relation to the unity ofintuitions in space and time, and that they arecompetent to determine this unity by means ofgeneral a priori connecting conceptions only onaccount of the pure ideality of space and time.Where this unity of time is not to be met with,as is the case with noumena, the whole use,indeed the whole meaning of the categories isentirely lost, for even the possibility of things tocorrespond to the categories is in this case in-comprehensible. On this point, I need only referthe reader to what I have said at the commen-cement of the General Remark appended to theforegoing chapter. Now, the possibility of athing can never be proved from the fact that the

conception of it is not self-contradictory, butonly by means of an intuition corresponding tothe conception. If, therefore, we wish to applythe categories to objects which cannot be regar-ded as phenomena, we must have an intuitiondifferent from the sensuous, and in this case theobjects would be a noumena in the positivesense of the word. Now, as such an intuition,that is, an intellectual intuition, is no part of ourfaculty of cognition, it is absolutely impossiblefor the categories to possess any applicationbeyond the limits of experience. It may be truethat there are intelligible existences to whichour faculty of sensuous intuition has no rela-tion, and cannot be applied, but our concep-tions of the understanding, as mere forms ofthought for our sensuous intuition, do not ex-tend to these. What, therefore, we call noume-non must be understood by us as such in a ne-gative sense.

If I take away from an empirical intuition allthought (by means of the categories), there re-mains no cognition of any object; for by meansof mere intuition nothing is cogitated, and,from the existence of such or such an affectionof sensibility in me, it does not follow that thisaffection or representation has any relation toan object without me. But if I take away all in-tuition, there still remains the form of thought,that is, the mode of determining an object forthe manifold of a possible intuition. Thus thecategories do in some measure really extendfurther than sensuous intuition, inasmuch asthey think objects in general, without regard tothe mode (of sensibility) in which these objectsare given. But they do not for this reason applyto and determine a wider sphere of objects,because we cannot assume that such can begiven, without presupposing the possibility ofanother than the sensuous mode of intuition, asupposition we are not justified in making.

I call a conception problematical which con-tains in itself no contradiction, and which isconnected with other cognitions as a limitationof given conceptions, but whose objective reali-ty cannot be cognized in any manner. The con-ception of a noumenon, that is, of a thing whichmust be cogitated not as an object of sense, butas a thing in itself (solely through the pure un-derstanding), is not self-contradictory, for weare not entitled to maintain that sensibility isthe only possible mode of intuition. Nay, furt-her, this conception is necessary to restrain sen-suous intuition within the bounds of phenome-na, and thus to limit the objective validity ofsensuous cognition; for things in themselves,which lie beyond its province, are called nou-mena for the very purpose of indicating thatthis cognition does not extend its application toall that the understanding thinks. But, after all,the possibility of such noumena is quite incom-prehensible, and beyond the sphere of pheno-mena, all is for us a mere void; that is to say, we

possess an understanding whose province doesproblematically extend beyond this sphere, butwe do not possess an intuition, indeed, noteven the conception of a possible intuition, bymeans of which objects beyond the region ofsensibility could be given us, and in referenceto which the understanding might be employedassertorically. The conception of a noumenon istherefore merely a limitative conception andtherefore only of negative use. But it is not anarbitrary or fictitious notion, but is connectedwith the limitation of sensibility, without,however, being capable of presenting us withany positive datum beyond this sphere.

The division of objects into phenomena andnoumena, and of the world into a mundus sen-sibilis and intelligibilis is therefore quite inad-missible in a positive sense, although concep-tions do certainly admit of such a division; forthe class of noumena have no determinate ob-ject corresponding to them, and cannot therefo-

re possess objective validity. If we abandon thesenses, how can it be made conceivable that thecategories (which are the only conceptions thatcould serve as conceptions for noumena) haveany sense or meaning at all, inasmuch as so-mething more than the mere unity of thought,namely, a possible intuition, is requisite fortheir application to an object? The conception ofa noumenon, considered as merely problemati-cal, is, however, not only admissible, but, as alimitative conception of sensibility, absolutelynecessary. But, in this case, a noumenon is not aparticular intelligible object for our understan-ding; on the contrary, the kind of understan-ding to which it could belong is itself a pro-blem, for we cannot form the most distant con-ception of the possibility of an understandingwhich should cognize an object, not discursive-ly by means of categories, but intuitively in anon-sensuous intuition. Our understandingattains in this way a sort of negative extension.That is to say, it is not limited by, but rather

limits, sensibility, by giving the name of nou-mena to things, not considered as phenomena,but as things in themselves. But it at the sametime prescribes limits to itself, for it confessesitself unable to cognize these by means of thecategories, and hence is compelled to cogitatethem merely as an unknown something.

I find, however, in the writings of modern aut-hors, an entirely different use of the expres-sions, mundus sensibilis and intelligibilis,which quite departs from the meaning of theancients—an acceptation in which, indeed, the-re is to be found no difficulty, but which at thesame time depends on mere verbal quibbling.According to this meaning, some have chosento call the complex of phenomena, in so far as itis intuited, mundus sensibilis, but in so far asthe connection thereof is cogitated according togeneral laws of thought, mundus intelligibilis.Astronomy, in so far as we mean by the wordthe mere observation of the starry heaven, may

represent the former; a system of astronomy,such as the Copernican or Newtonian, the lat-ter. But such twisting of words is a mere sop-histical subterfuge, to avoid a difficult question,by modifying its meaning to suit our own con-venience. To be sure, understanding and reasonare employed in the cognition of phenomena;but the question is, whether these can be ap-plied when the object is not a phenomenon andin this sense we regard it if it is cogitated asgiven to the understanding alone, and not tothe senses. The question therefore is whether,over and above the empirical use of the unders-tanding, a transcendental use is possible, whichapplies to the noumenon as an object. Thisquestion we have answered in the negative.

When therefore we say, the senses representobjects as they appear, the understanding asthey are, the latter statement must not be un-derstood in a transcendental, but only in anempirical signification, that is, as they must be

represented in the complete connection of phe-nomena, and not according to what they maybe, apart from their relation to possible expe-rience, consequently not as objects of the pureunderstanding. For this must ever remain unk-nown to us. Nay, it is also quite unknown to uswhether any such transcendental or extraordi-nary cognition is possible under any circums-tances, at least, whether it is possible by meansof our categories. Understanding and sensibili-ty, with us, can determine objects only in con-junction. If we separate them, we have intui-tions without conceptions, or conceptions wit-hout intuitions; in both cases, representations,which we cannot apply to any determinate ob-ject.

If, after all our inquiries and explanations, anyone still hesitates to abandon the mere trans-cendental use of the categories, let him attemptto construct with them a synthetical proposi-tion. It would, of course, be unnecessary for

this purpose to construct an analytical proposi-tion, for that does not extend the sphere of theunderstanding, but, being concerned onlyabout what is cogitated in the conception itself,it leaves it quite undecided whether the concep-tion has any relation to objects, or merely indi-cates the unity of thought—complete abstrac-tion being made of the modi in which an objectmay be given: in such a proposition, it is suffi-cient for the understanding to know what liesin the conception—to what it applies is to itindifferent. The attempt must therefore be ma-de with a synthetical and so-called transcen-dental principle, for example: "Everything thatexists, exists as substance," or, "Everything thatis contingent exists as an effect of some otherthing, viz., of its cause." Now I ask, whence canthe understanding draw these synthetical pro-positions, when the conceptions contained the-rein do not relate to possible experience but tothings in themselves (noumena)? Where is to befound the third term, which is always requisite

PURE site in a synthetical proposition, whichmay connect in the same proposition concep-tions which have no logical (analytical) connec-tion with each other? The proposition neverwill be demonstrated, nay, more, the possibilityof any such pure assertion never can be shown,without making reference to the empirical useof the understanding, and thus, ipso facto,completely renouncing pure and non-sensuousjudgement. Thus the conception of pure andmerely intelligible objects is completely void ofall principles of its application, because wecannot imagine any mode in which they mightbe given, and the problematical thought whichleaves a place open for them serves only, like avoid space, to limit the use of empirical princi-ples, without containing at the same time anyother object of cognition beyond their sphere.

APPENDIX.

Of the Equivocal Nature or Amphiboly of theConceptions of Reflection from the Confusionof the Transcendental with the Empirical use ofthe Understanding.

Reflection (reflexio) is not occupied about ob-jects themselves, for the purpose of directlyobtaining conceptions of them, but is that stateof the mind in which we set ourselves to disco-ver the subjective conditions under which weobtain conceptions. It is the consciousness ofthe relation of given representations to the dif-ferent sources or faculties of cognition, bywhich alone their relation to each other can berightly determined. The first question whichoccurs in considering our representations is towhat faculty of cognition do they belong? Tothe understanding or to the senses? Many jud-gements are admitted to be true from mere

habit or inclination; but, because reflectionneither precedes nor follows, it is held to be ajudgement that has its origin in the understan-ding. All judgements do not require examina-tion, that is, investigation into the grounds oftheir truth. For, when they are immediatelycertain (for example: "Between two points therecan be only one straight line"), no better or lessmediate test of their truth can be found thanthat which they themselves contain and expre-ss. But all judgement, nay, all comparisons re-quire reflection, that is, a distinction of the fa-culty of cognition to which the given concep-tions belong. The act whereby I compare myrepresentations with the faculty of cognitionwhich originates them, and whereby I distin-guish whether they are compared with eachother as belonging to the pure understandingor to sensuous intuition, I term transcendentalreflection. Now, the relations in which concep-tions can stand to each other are those of identi-ty and difference, agreement and opposition, of

the internal and external, finally, of the deter-minable and the determining (matter andform). The proper determination of these rela-tions rests on the question, to what faculty ofcognition they subjectively belong, whether tosensibility or understanding? For, on the man-ner in which we solve this question dependsthe manner in which we must cogitate theserelations.

Before constructing any objective judgement,we compare the conceptions that are to be pla-ced in the judgement, and observe whetherthere exists identity (of many representations inone conception), if a general judgement is to beconstructed, or difference, if a particular; whet-her there is agreement when affirmative; andopposition when negative judgements are to beconstructed, and so on. For this reason weought to call these conceptions, conceptions ofcomparison (conceptus comparationis). But as,when the question is not as to the logical form,

but as to the content of conceptions, that is tosay, whether the things themselves are identicalor different, in agreement or opposition, and soon, the things can have a twofold relation toour faculty of cognition, to wit, a relation eitherto sensibility or to the understanding, and as onthis relation depends their relation to each ot-her, transcendental reflection, that is, the rela-tion of given representations to one or the otherfaculty of cognition, can alone determine thislatter relation. Thus we shall not be able to dis-cover whether the things are identical or diffe-rent, in agreement or opposition, etc., from themere conception of the things by means ofcomparison (comparatio), but only by distin-guishing the mode of cognition to which theybelong, in other words, by means of transcen-dental reflection. We may, therefore, with justi-ce say, that logical reflection is mere compari-son, for in it no account is taken of the facultyof cognition to which the given conceptionsbelong, and they are consequently, as far as

regards their origin, to be treated as homoge-neous; while transcendental reflection (whichapplies to the objects themselves) contains theground of the possibility of objective compari-son of representations with each other, and istherefore very different from the former, becau-se the faculties of cognition to which they be-long are not even the same. Transcendentalreflection is a duty which no one can neglectwho wishes to establish an a priori judgementupon things. We shall now proceed to fulfil thisduty, and thereby throw not a little light on thequestion as to the determination of the properbusiness of the understanding.

1. Identity and Difference. When an object ispresented to us several times, but always withthe same internal determinations (qualitas etquantitas), it, if an object of pure understan-ding, is always the same, not several things, butonly one thing (numerica identitas); but if aphenomenon, we do not concern ourselves

with comparing the conception of the thingwith the conception of some other, but, alt-hough they may be in this respect perfectly thesame, the difference of place at the same time isa sufficient ground for asserting the numericaldifference of these objects (of sense). Thus, inthe case of two drops of water, we may makecomplete abstraction of all internal difference(quality and quantity), and, the fact that theyare intuited at the same time in different places,is sufficient to justify us in holding them to benumerically different. Leibnitz regarded phe-nomena as things in themselves, consequentlyas intelligibilia, that is, objects of pure unders-tanding (although, on account of the confusednature of their representations, he gave themthe name of phenomena), and in this case hisprinciple of the indiscernible (principium iden-tatis indiscernibilium) is not to be impugned.But, as phenomena are objects of sensibility,and, as the understanding, in respect of them,must be employed empirically and not purely

or transcendentally, plurality and numericaldifference are given by space itself as the condi-tion of external phenomena. For one part ofspace, although it may be perfectly similar andequal to another part, is still without it, and forthis reason alone is different from the latter,which is added to it in order to make up a grea-ter space. It follows that this must hold good ofall things that are in the different parts of spaceat the same time, however similar and equalone may be to another.

2. Agreement and Opposition. When reality isrepresented by the pure understanding (realitasnoumenon), opposition between realities isincogitable—such a relation, that is, that whenthese realities are connected in one subject, theyannihilate the effects of each other and may berepresented in the formula 3 - 3 = 0. On theother hand, the real in a phenomenon (realitasphaenomenon) may very well be in mutualopposition, and, when united in the same sub-

ject, the one may completely or in part annihila-te the effect or consequence of the other; as inthe case of two moving forces in the samestraight line drawing or impelling a point inopposite directions, or in the case of a pleasurecounterbalancing a certain amount of pain.

3. The Internal and External. In an object of thepure understanding, only that is internal whichhas no relation (as regards its existence) toanything different from itself. On the otherhand, the internal determinations of a substan-tia phaenomenon in space are nothing but rela-tions, and it is itself nothing more than a com-plex of mere relations. Substance in space weare cognizant of only through forces operativein it, either drawing others towards itself (at-traction), or preventing others from forcing intoitself (repulsion and impenetrability). We knowno other properties that make up the concep-tion of substance phenomenal in space, andwhich we term matter. On the other hand, as an

object of the pure understanding, every subs-tance must have internal determination andforces. But what other internal attributes ofsuch an object can I think than those which myinternal sense presents to me? That, to wit,which in either itself thought, or somethinganalogous to it. Hence Leibnitz, who lookedupon things as noumena, after denying themeverything like external relation, and thereforealso composition or combination, declared thatall substances, even the component parts ofmatter, were simple substances with powers ofrepresentation, in one word, monads.

4. Matter and Form. These two conceptions lieat the foundation of all other reflection, so inse-parably are they connected with every mode ofexercising the understanding. The former deno-tes the determinable in general, the second itsdetermination, both in a transcendental sense,abstraction being made of every difference inthat which is given, and of the mode in which it

is determined. Logicians formerly termed theuniversal, matter, the specific difference of thisor that part of the universal, form. In a judge-ment one may call the given conceptions logicalmatter (for the judgement), the relation of theseto each other (by means of the copula), the formof the judgement. In an object, the compositeparts thereof (essentialia) are the matter; themode in which they are connected in the object,the form. In respect to things in general, unlimi-ted reality was regarded as the matter of allpossibility, the limitation thereof (negation) asthe form, by which one thing is distinguishedfrom another according to transcendental con-ceptions. The understanding demands that so-mething be given (at least in the conception), inorder to be able to determine it in a certainmanner. Hence, in a conception of the pureunderstanding, the matter precedes the form,and for this reason Leibnitz first assumed theexistence of things (monads) and of an internalpower of representation in them, in order to

found upon this their external relation and thecommunity their state (that is, of their represen-tations). Hence, with him, space and time werepossible—the former through the relation ofsubstances, the latter through the connection oftheir determinations with each other, as causesand effects. And so would it really be, if thepure understanding were capable of an imme-diate application to objects, and if space andtime were determinations of things in themsel-ves. But being merely sensuous intuitions, inwhich we determine all objects solely as phe-nomena, the form of intuition (as a subjectiveproperty of sensibility) must antecede all mat-ter (sensations), consequently space and timemust antecede all phenomena and all data ofexperience, and rather make experience itselfpossible. But the intellectual philosopher couldnot endure that the form should precede thethings themselves and determine their possibi-lity; an objection perfectly correct, if we assumethat we intuite things as they are, although

with confused representation. But as sensuousintuition is a peculiar subjective condition,which is a priori at the foundation of all percep-tion, and the form of which is primitive, theform must be given per se, and so far from mat-ter (or the things themselves which appear)lying at the foundation of experience (as wemust conclude, if we judge by mere concep-tions), the very possibility of itself presupposes,on the contrary, a given formal intuition (spaceand time).

REMARK ON THE AMPHIBOLY OF THECONCEPTIONS OF REFLECTION.

Let me be allowed to term the position whichwe assign to a conception either in the sensibili-ty or in the pure understanding, the transcen-dental place. In this manner, the appointmentof the position which must be taken by each

conception according to the difference in itsuse, and the directions for determining thisplace to all conceptions according to rules,would be a transcendental topic, a doctrinewhich would thoroughly shield us from thesurreptitious devices of the pure understandingand the delusions which thence arise, as itwould always distinguish to what faculty ofcognition each conception properly belonged.Every conception, every title, under which ma-ny cognitions rank together, may be called alogical place. Upon this is based the logical to-pic of Aristotle, of which teachers and rhetori-cians could avail themselves, in order, undercertain titles of thought, to observe what wouldbest suit the matter they had to treat, and thusenable themselves to quibble and talk withfluency and an appearance of profundity.

Transcendental topic, on the contrary, containsnothing more than the above-mentioned fourtitles of all comparison and distinction, which

differ from categories in this respect, that theydo not represent the object according to thatwhich constitutes its conception (quantity, rea-lity), but set forth merely the comparison ofrepresentations, which precedes our concep-tions of things. But this comparison requires aprevious reflection, that is, a determination ofthe place to which the representations of thethings which are compared belong, whether, towit, they are cogitated by the pure understan-ding, or given by sensibility.

Conceptions may be logically compared wit-hout the trouble of inquiring to what facultytheir objects belong, whether as noumena, tothe understanding, or as phenomena, to sensi-bility. If, however, we wish to employ theseconceptions in respect of objects, previoustranscendental reflection is necessary. Withoutthis reflection I should make a very unsafe useof these conceptions, and construct pretendedsynthetical propositions which critical reason

cannot acknowledge and which are based sole-ly upon a transcendental amphiboly, that is,upon a substitution of an object of pure unders-tanding for a phenomenon.

For want of this doctrine of transcendental to-pic, and consequently deceived by the amp-hiboly of the conceptions of reflection, the cele-brated Leibnitz constructed an intellectual sys-tem of the world, or rather, believed himselfcompetent to cognize the internal nature ofthings, by comparing all objects merely withthe understanding and the abstract formal con-ceptions of thought. Our table of the concep-tions of reflection gives us the unexpected ad-vantage of being able to exhibit the distinctivepeculiarities of his system in all its parts, and atthe same time of exposing the fundamentalprinciple of this peculiar mode of thought,which rested upon naught but a misconception.He compared all things with each other merelyby means of conceptions, and naturally found

no other differences than those by which theunderstanding distinguishes its pure concep-tions one from another. The conditions of sen-suous intuition, which contain in themselvestheir own means of distinction, he did not lookupon as primitive, because sensibility was tohim but a confused mode of representation andnot any particular source of representations. Aphenomenon was for him the representation ofthe thing in itself, although distinguished fromcognition by the understanding only in respectof the logical form—the former with its usualwant of analysis containing, according to him, acertain mixture of collateral representations inits conception of a thing, which it is the duty ofthe understanding to separate and distinguish.In one word, Leibnitz intellectualized pheno-mena, just as Locke, in his system of noogony(if I may be allowed to make use of such ex-pressions), sensualized the conceptions of theunderstanding, that is to say, declared them tobe nothing more than empirical or abstract con-

ceptions of reflection. Instead of seeking in theunderstanding and sensibility two differentsources of representations, which, however, canpresent us with objective judgements of thingsonly in conjunction, each of these great menrecognized but one of these faculties, which, intheir opinion, applied immediately to things inthemselves, the other having no duty but thatof confusing or arranging the representations ofthe former.

Accordingly, the objects of sense were compa-red by Leibnitz as things in general merely inthe understanding.

1st. He compares them in regard to their identi-ty or difference —as judged by the understan-ding. As, therefore, he considered merely theconceptions of objects, and not their position inintuition, in which alone objects can be given,and left quite out of sight the transcendentallocale of these conceptions—whether, that is,their object ought to be classed among pheno-

mena, or among things in themselves, it was tobe expected that he should extend the applica-tion of the principle of indiscernibles, which isvalid solely of conceptions of things in general,to objects of sense (mundus phaenomenon),and that he should believe that he had therebycontributed in no small degree to extend ourknowledge of nature. In truth, if I cognize in allits inner determinations a drop of water as athing in itself, I cannot look upon one drop asdifferent from another, if the conception of theone is completely identical with that of the ot-her. But if it is a phenomenon in space, it has aplace not merely in the understanding (amongconceptions), but also in sensuous external in-tuition (in space), and in this case, the physicallocale is a matter of indifference in regard to theinternal determinations of things, and one pla-ce, B, may contain a thing which is perfectlysimilar and equal to another in a place, A, justas well as if the two things were in every res-pect different from each other. Difference of

place without any other conditions, makes theplurality and distinction of objects as pheno-mena, not only possible in itself, but even ne-cessary. Consequently, the above so-called lawis not a law of nature. It is merely an analyticalrule for the comparison of things by means ofmere conceptions.

2nd. The principle: "Realities (as simple affir-mations) never logically contradict each other,"is a proposition perfectly true respecting therelation of conceptions, but, whether as regardsnature, or things in themselves (of which wehave not the slightest conception), is withoutany the least meaning. For real opposition, inwhich A - B is = 0, exists everywhere, an oppo-sition, that is, in which one reality united withanother in the same subject annihilates the ef-fects of the other—a fact which is constantlybrought before our eyes by the different anta-gonistic actions and operations in nature,which, nevertheless, as depending on real for-

ces, must be called realitates phaenomena. Ge-neral mechanics can even present us with theempirical condition of this opposition in an apriori rule, as it directs its attention to the op-position in the direction of forces—a conditionof which the transcendental conception of reali-ty can tell us nothing. Although M. Leibnitz didnot announce this proposition with preciselythe pomp of a new principle, he yet employedit for the establishment of new propositions,and his followers introduced it into their Leib-nitzio-Wolfian system of philosophy. Accor-ding to this principle, for example, all evils arebut consequences of the limited nature of crea-ted beings, that is, negations, because these arethe only opposite of reality. (In the mere con-ception of a thing in general this is really thecase, but not in things as phenomena.) In likemanner, the upholders of this system deem itnot only possible, but natural also, to connectand unite all reality in one being, because theyacknowledge no other sort of opposition than

that of contradiction (by which the conceptionitself of a thing is annihilated), and find them-selves unable to conceive an opposition of reci-procal destruction, so to speak, in which onereal cause destroys the effect of another, andthe conditions of whose representation we meetwith only in sensibility.

3rd. The Leibnitzian monadology has really nobetter foundation than on this philosopher'smode of falsely representing the difference ofthe internal and external solely in relation tothe understanding. Substances, in general, musthave something inward, which is therefore freefrom external relations, consequently from thatof composition also. The simple—that whichcan be represented by a unit—is therefore thefoundation of that which is internal in things inthemselves. The internal state of substancescannot therefore consist in place, shape, con-tact, or motion, determinations which are allexternal relations, and we can ascribe to them

no other than that whereby we internally de-termine our faculty of sense itself, that is to say,the state of representation. Thus, then, wereconstructed the monads, which were to formthe elements of the universe, the active force ofwhich consists in representation, the effects ofthis force being thus entirely confined to them-selves.

For the same reason, his view of the possiblecommunity of substances could not represent itbut as a predetermined harmony, and by nomeans as a physical influence. For inasmuch aseverything is occupied only internally, that is,with its own representations, the state of therepresentations of one substance could notstand in active and living connection with thatof another, but some third cause operating onall without exception was necessary to makethe different states correspond with one anot-her. And this did not happen by means of assis-tance applied in each particular case (systema

assistentiae), but through the unity of the ideaof a cause occupied and connected with allsubstances, in which they necessarily receive,according to the Leibnitzian school, their exis-tence and permanence, consequently also reci-procal correspondence, according to universallaws.

4th. This philosopher's celebrated doctrine ofspace and time, in which he intellectualizedthese forms of sensibility, originated in the sa-me delusion of transcendental reflection. If Iattempt to represent by the mere understan-ding, the external relations of things, I can doso only by employing the conception of theirreciprocal action, and if I wish to connect onestate of the same thing with another state, Imust avail myself of the notion of the order ofcause and effect. And thus Leibnitz regardedspace as a certain order in the community ofsubstances, and time as the dynamical sequen-ce of their states. That which space and time

possess proper to themselves and independentof things, he ascribed to a necessary confusionin our conceptions of them, whereby that whichis a mere form of dynamical relations is held tobe a self-existent intuition, antecedent even tothings themselves. Thus space and time werethe intelligible form of the connection of things(substances and their states) in themselves. Butthings were intelligible substances (substantiaenoumena). At the same time, he made theseconceptions valid of phenomena, because hedid not allow to sensibility a peculiar mode ofintuition, but sought all, even the empiricalrepresentation of objects, in the understanding,and left to sense naught but the despicable taskof confusing and disarranging the representa-tions of the former.

But even if we could frame any synthetical pro-position concerning things in themselves bymeans of the pure understanding (which isimpossible), it could not apply to phenomena,

which do not represent things in themselves. Insuch a case I should be obliged in transcenden-tal reflection to compare my conceptions onlyunder the conditions of sensibility, and so spa-ce and time would not be determinations ofthings in themselves, but of phenomena. Whatthings may be in themselves, I know not andneed not know, because a thing is never pre-sented to me otherwise than as a phenomenon.

I must adopt the same mode of procedure withthe other conceptions of reflection. Matter issubstantia phaenomenon. That in it which isinternal I seek to discover in all parts of spacewhich it occupies, and in all the functions andoperations it performs, and which are indeednever anything but phenomena of the externalsense. I cannot therefore find anything that isabsolutely, but only what is comparatively in-ternal, and which itself consists of external rela-tions. The absolutely internal in matter, and asit should be according to the pure understan-

ding, is a mere chimera, for matter is not anobject for the pure understanding. But thetranscendental object, which is the foundationof the phenomenon which we call matter, is amere nescio quid, the nature of which we couldnot understand, even though someone werefound able to tell us. For we can understandnothing that does not bring with it somethingin intuition corresponding to the expressionsemployed. If, by the complaint of being unableto perceive the internal nature of things, it ismeant that we do not comprehend by the pureunderstanding what the things which appear tous may be in themselves, it is a silly and unrea-sonable complaint; for those who talk thus rea-lly desire that we should be able to cognize,consequently to intuite, things without senses,and therefore wish that we possessed a facultyof cognition perfectly different from the humanfaculty, not merely in degree, but even as re-gards intuition and the mode thereof, so thatthus we should not be men, but belong to a

class of beings, the possibility of whose existen-ce, much less their nature and constitution, wehave no means of cognizing. By observationand analysis of phenomena we penetrate intothe interior of nature, and no one can say whatprogress this knowledge may make in time. Butthose transcendental questions which pass be-yond the limits of nature, we could never ans-wer, even although all nature were laid open tous, because we have not the power of observingour own mind with any other intuition thanthat of our internal sense. For herein lies themystery of the origin and source of our facultyof sensibility. Its application to an object, andthe transcendental ground of this unity of sub-jective and objective, lie too deeply concealedfor us, who cognize ourselves only through theinternal sense, consequently as phenomena, tobe able to discover in our existence anythingbut phenomena, the non-sensuous cause ofwhich we at the same time earnestly desire topenetrate to.

The great utility of this critique of conclusionsarrived at by the processes of mere reflectionconsists in its clear demonstration of the nullityof all conclusions respecting objects which arecompared with each other in the understandingalone, while it at the same time confirms whatwe particularly insisted on, namely, that, alt-hough phenomena are not included as things inthemselves among the objects of the pure un-derstanding, they are nevertheless the onlythings by which our cognition can possess ob-jective reality, that is to say, which give us in-tuitions to correspond with our conceptions.

When we reflect in a purely logical manner, wedo nothing more than compare conceptions inour understanding, to discover whether bothhave the same content, whether they are self-contradictory or not, whether anything is con-tained in either conception, which of the two isgiven, and which is merely a mode of thinkingthat given. But if I apply these conceptions to

an object in general (in the transcendental sen-se), without first determining whether it is anobject of sensuous or intellectual intuition, cer-tain limitations present themselves, which for-bid us to pass beyond the conceptions and ren-der all empirical use of them impossible. Andthus these limitations prove that the representa-tion of an object as a thing in general is not onlyinsufficient, but, without sensuous determina-tion and independently of empirical conditions,self-contradictory; that we must therefore makeabstraction of all objects, as in logic, or, admit-ting them, must think them under conditions ofsensuous intuition; that, consequently, the inte-lligible requires an altogether peculiar intuition,which we do not possess, and in the absence ofwhich it is for us nothing; while, on the otherhand phenomena cannot be objects in themsel-ves. For, when I merely think things in general,the difference in their external relations cannotconstitute a difference in the things themselves;on the contrary, the former presupposes the

latter, and if the conception of one of twothings is not internally different from that ofthe other, I am merely thinking the same thingin different relations. Further, by the additionof one affirmation (reality) to the other, the po-sitive therein is really augmented, and nothingis abstracted or withdrawn from it; hence thereal in things cannot be in contradiction with oropposition to itself—and so on.

The true use of the conceptions of reflection inthe employment of the understanding has, aswe have shown, been so misconceived by Leib-nitz, one of the most acute philosophers of eit-her ancient or modern times, that he has beenmisled into the construction of a baseless sys-tem of intellectual cognition, which professes todetermine its objects without the interventionof the senses. For this reason, the exposition ofthe cause of the amphiboly of these concep-tions, as the origin of these false principles, is of

great utility in determining with certainty theproper limits of the understanding.

It is right to say whatever is affirmed or deniedof the whole of a conception can be affirmed ordenied of any part of it (dictum de omni et nu-llo); but it would be absurd so to alter this logi-cal proposition as to say whatever is not con-tained in a general conception is likewise notcontained in the particular conceptions whichrank under it; for the latter are particular con-ceptions, for the very reason that their contentis greater than that which is cogitated in thegeneral conception. And yet the whole intellec-tual system of Leibnitz is based upon this falseprinciple, and with it must necessarily fall tothe ground, together with all the ambiguousprinciples in reference to the employment ofthe understanding which have thence origina-ted.

Leibnitz's principle of the identity of indiscer-nibles or indistinguishables is really based on

the presupposition that, if in the conception of athing a certain distinction is not to be found, itis also not to be met with in things themselves;that, consequently, all things are completelyidentical (numero eadem) which are not distin-guishable from each other (as to quality orquantity) in our conceptions of them. But, as inthe mere conception of anything abstractionhas been made of many necessary conditions ofintuition, that of which abstraction has beenmade is rashly held to be non-existent, and not-hing is attributed to the thing but what is con-tained in its conception.

The conception of a cubic foot of space, howe-ver I may think it, is in itself completely identi-cal. But two cubic feet in space are neverthelessdistinct from each other from the sole fact oftheir being in different places (they are numerodiversa); and these places are conditions of in-tuition, wherein the object of this conception isgiven, and which do not belong to the concep-

tion, but to the faculty of sensibility. In likemanner, there is in the conception of a thing nocontradiction when a negative is not connectedwith an affirmative; and merely affirmativeconceptions cannot, in conjunction, produceany negation. But in sensuous intuition, whe-rein reality (take for example, motion) is given,we find conditions (opposite directions)—ofwhich abstraction has been made in the concep-tion of motion in general—which render possi-ble a contradiction or opposition (not indeed ofa logical kind)—and which from pure positivesproduce zero = 0. We are therefore not justifiedin saying that all reality is in perfect agreementand harmony, because no contradiction is dis-coverable among its conceptions.* According tomere conceptions, that which is internal is thesubstratum of all relations or external determi-nations. When, therefore, I abstract all condi-tions of intuition, and confine myself solely tothe conception of a thing in general, I can makeabstraction of all external relations, and there

must nevertheless remain a conception of thatwhich indicates no relation, but merely internaldeterminations. Now it seems to follow that ineverything (substance) there is somethingwhich is absolutely internal and which antece-des all external determinations, inasmuch as itrenders them possible; and that therefore thissubstratum is something which does not con-tain any external relations and is consequentlysimple (for corporeal things are never anythingbut relations, at least of their parts external toeach other); and, inasmuch as we know of noother absolutely internal determinations thanthose of the internal sense, this substratum isnot only simple, but also, analogously with ourinternal sense, determined through representa-tions, that is to say, all things are properly mo-nads, or simple beings endowed with the po-wer of representation. Now all this would beperfectly correct, if the conception of a thingwere the only necessary condition of the pre-sentation of objects of external intuition. It is,

on the contrary, manifest that a permanentphenomenon in space (impenetrable extension)can contain mere relations, and nothing that isabsolutely internal, and yet be the primarysubstratum of all external perception. By mereconceptions I cannot think anything external,without, at the same time, thinking somethinginternal, for the reason that conceptions of rela-tions presuppose given things, and withoutthese are impossible. But, as an intuition thereis something (that is, space, which, with all itcontains, consists of purely formal, or, indeed,real relations) which is not found in the mereconception of a thing in general, and this pre-sents to us the substratum which could not becognized through conceptions alone, I cannotsay: because a thing cannot be represented bymere conceptions without something absolute-ly internal, there is also, in the things themsel-ves which are contained under these concep-tions, and in their intuition nothing external towhich something absolutely internal does not

serve as the foundation. For, when we havemade abstraction of all the conditions of intui-tion, there certainly remains in the mere con-ception nothing but the internal in general,through which alone the external is possible.But this necessity, which is grounded uponabstraction alone, does not obtain in the case ofthings themselves, in so far as they are given inintuition with such determinations as expressmere relations, without having anything inter-nal as their foundation; for they are not thingsof a thing of which we can neither for they arenot things in themselves, but only phenomena.What we cognize in matter is nothing but rela-tions (what we call its internal determinationsare but comparatively internal). But there aresome self-subsistent and permanent, throughwhich a determined object is given. That I,when abstraction is made of these relations,have nothing more to think, does not destroythe conception of a thing as phenomenon, northe conception of an object in abstracto, but it

does away with the possibility of an object thatis determinable according to mere conceptions,that is, of a noumenon. It is certainly startlingto hear that a thing consists solely of relations;but this thing is simply a phenomenon, andcannot be cogitated by means of the mere cate-gories: it does itself consist in the mere relationof something in general to the senses. In thesame way, we cannot cogitate relations ofthings in abstracto, if we commence with con-ceptions alone, in any other manner than thatone is the cause of determinations in the other;for that is itself the conception of the unders-tanding or category of relation. But, as in thiscase we make abstraction of all intuition, welose altogether the mode in which the manifolddetermines to each of its parts its place, that is,the form of sensibility (space); and yet this mo-de antecedes all empirical causality.

[*Footnote: If any one wishes here to have re-course to the usual subterfuge, and to say, thatat least realitates noumena cannot be in opposi-tion to each other, it will be requisite for him toadduce an example of this pure and non-sensuous reality, that it may be understoodwhether the notion represents something ornothing. But an example cannot be found ex-cept in experience, which never presents to usanything more than phenomena; and thus theproposition means nothing more than that theconception which contains only affirmativesdoes not contain anything negative—a proposi-tion nobody ever doubted.]

If by intelligible objects we understand thingswhich can be thought by means of the purecategories, without the need of the schemata ofsensibility, such objects are impossible. For thecondition of the objective use of all our concep-tions of understanding is the mode of our sen-

suous intuition, whereby objects are given; and,if we make abstraction of the latter, the formercan have no relation to an object. And even ifwe should suppose a different kind of intuitionfrom our own, still our functions of thoughtwould have no use or signification in respectthereof. But if we understand by the term, ob-jects of a non-sensuous intuition, in respect ofwhich our categories are not valid, and ofwhich we can accordingly have no knowledge(neither intuition nor conception), in this mere-ly negative sense noumena must be admitted.For this is no more than saying that our modeof intuition is not applicable to all things, butonly to objects of our senses, that consequentlyits objective validity is limited, and that room istherefore left for another kind of intuition, andthus also for things that may be objects of it.But in this sense the conception of a noumenonis problematical, that is to say, it is the notion ofthat it that it is possible, nor that it is impossi-ble, inasmuch as we do not know of any mode

of intuition besides the sensuous, or of any ot-her sort of conceptions than the categories—amode of intuition and a kind of conceptionneither of which is applicable to a non-sensuous object. We are on this account incom-petent to extend the sphere of our objects ofthought beyond the conditions of our sensibili-ty, and to assume the existence of objects ofpure thought, that is, of noumena, inasmuch asthese have no true positive signification. For itmust be confessed of the categories that theyare not of themselves sufficient for the cogni-tion of things in themselves and, without thedata of sensibility, are mere subjective forms ofthe unity of the understanding. Thought is cer-tainly not a product of the senses, and in so faris not limited by them, but it does not thereforefollow that it may be employed purely andwithout the intervention of sensibility, for itwould then be without reference to an object.And we cannot call a noumenon an object ofpure thought; for the representation thereof is

but the problematical conception of an objectfor a perfectly different intuition and a perfec-tly different understanding from ours, both ofwhich are consequently themselves problema-tical. The conception of a noumenon is therefo-re not the conception of an object, but merely aproblematical conception inseparably connec-ted with the limitation of our sensibility. That isto say, this conception contains the answer tothe question: "Are there objects quite unconnec-ted with, and independent of, our intuition?"—a question to which only an indeterminateanswer can be given. That answer is: "Inasmuchas sensuous intuition does not apply to allthings without distinction, there remains roomfor other and different objects." The existence ofthese problematical objects is therefore not ab-solutely denied, in the absence of a determinateconception of them, but, as no category is validin respect of them, neither must they be admit-ted as objects for our understanding.

Understanding accordingly limits sensibility,without at the same time enlarging its ownfield. While, moreover, it forbids sensibility toapply its forms and modes to things in them-selves and restricts it to the sphere of pheno-mena, it cogitates an object in itself, only,however, as a transcendental object, which isthe cause of a phenomenon (consequently notitself a phenomenon), and which cannot bethought either as a quantity or as reality, or assubstance (because these conceptions alwaysrequire sensuous forms in which to determinean object)—an object, therefore, of which weare quite unable to say whether it can be metwith in ourselves or out of us, whether it wouldbe annihilated together with sensibility, or, ifthis were taken away, would continue to exist.If we wish to call this object a noumenon, be-cause the representation of it is non-sensuous,we are at liberty to do so. But as we can applyto it none of the conceptions of our understan-ding, the representation is for us quite void,

and is available only for the indication of thelimits of our sensuous intuition, thereby lea-ving at the same time an empty space, whichwe are competent to fill by the aid neither ofpossible experience, nor of the pure understan-ding.

The critique of the pure understanding, accor-dingly, does not permit us to create for oursel-ves a new field of objects beyond those whichare presented to us as phenomena, and to strayinto intelligible worlds; nay, it does not evenallow us to endeavour to form so much as aconception of them. The specious error whichleads to this—and which is a perfectly excusa-ble one—lies in the fact that the employment ofthe understanding, contrary to its proper pur-pose and destination, is made transcendental,and objects, that is, possible intuitions, are ma-de to regulate themselves according to concep-tions, instead of the conceptions arrangingthemselves according to the intuitions, on

which alone their own objective validity rests.Now the reason of this again is that appercep-tion, and with it thought, antecedes all possibledeterminate arrangement of representations.Accordingly we think something in general anddetermine it on the one hand sensuously, but,on the other, distinguish the general and inabstracto represented object from this particu-lar mode of intuiting it. In this case there re-mains a mode of determining the object by me-re thought, which is really but a logical formwithout content, which, however, seems to usto be a mode of the existence of the object initself (noumenon), without regard to intuitionwhich is limited to our senses.

Before ending this transcendental analytic, wemust make an addition, which, although initself of no particular importance, seems to benecessary to the completeness of the system.The highest conception, with which a transcen-dental philosophy commonly begins, is the

division into possible and impossible. But as alldivision presupposes a divided conception, astill higher one must exist, and this is the con-ception of an object in general—problematicallyunderstood and without its being decidedwhether it is something or nothing. As the ca-tegories are the only conceptions which applyto objects in general, the distinguishing of anobject, whether it is something or nothing, mustproceed according to the order and direction ofthe categories.

1. To the categories of quantity, that is, the con-ceptions of all, many, and one, the conceptionwhich annihilates all, that is, the conception ofnone, is opposed. And thus the object of a con-ception, to which no intuition can be found tocorrespond, is = nothing. That is, it is a concep-tion without an object (ens rationis), like nou-mena, which cannot be considered possible inthe sphere of reality, though they must not the-refore be held to be impossible—or like certain

new fundamental forces in matter, the existenceof which is cogitable without contradiction,though, as examples from experience are notforthcoming, they must not be regarded as pos-sible.

2. Reality is something; negation is nothing,that is, a conception of the absence of an object,as cold, a shadow (nihil privativum).

3. The mere form of intuition, without substan-ce, is in itself no object, but the merely formalcondition of an object (as phenomenon), as pu-re space and pure time. These are certainly so-mething, as forms of intuition, but are notthemselves objects which are intuited (ens ima-ginarium).

4. The object of a conception which is self-contradictory, is nothing, because the concep-tion is nothing—is impossible, as a figure com-posed of two straight lines (nihil negativum).

The table of this division of the conception ofnothing (the corresponding division of the con-ception of something does not require specialdescription) must therefore be arranged as fo-llows:

NOTHING AS 1 As Empty Conception without object, ens rationis 2 3 Empty object of Empty intuition a conception, without object, nihil privativum ens imaginarium 4 Empty object without conception, nihil negativum

We see that the ens rationis is distinguishedfrom the nihil negativum or pure nothing bythe consideration that the former must not bereckoned among possibilities, because it is amere fiction- though not self-contradictory,while the latter is completely opposed to allpossibility, inasmuch as the conception annihi-lates itself. Both, however, are empty concep-tions. On the other hand, the nihil privativumand ens imaginarium are empty data for con-ceptions. If light be not given to the senses, wecannot represent to ourselves darkness, and ifextended objects are not perceived, we cannotrepresent space. Neither the negation, nor themere form of intuition can, without somethingreal, be an object.

TRANSCENDENTAL LOGIC. SECONDDIVISION.

TRANSCENDENTAL DIALECTIC. IN-TRODUCTION.

I. Of Transcendental Illusory Appearance.

We termed dialectic in general a logic of appea-rance. This does not signify a doctrine of pro-bability; for probability is truth, only cognizedupon insufficient grounds, and though the in-formation it gives us is imperfect, it is not the-refore deceitful. Hence it must not be separatedfrom the analytical part of logic. Still less mustphenomenon and appearance be held to beidentical. For truth or illusory appearance doesnot reside in the object, in so far as it is intuited,but in the judgement upon the object, in so faras it is thought. It is, therefore, quite correct tosay that the senses do not err, not because they

always judge correctly, but because they do notjudge at all. Hence truth and error, consequen-tly also, illusory appearance as the cause oferror, are only to be found in a judgement, thatis, in the relation of an object to our understan-ding. In a cognition which completely harmo-nizes with the laws of the understanding, noerror can exist. In a representation of the sen-ses—as not containing any judgement—there isalso no error. But no power of nature can ofitself deviate from its own laws. Hence neitherthe understanding per se (without the influenceof another cause), nor the senses per se, wouldfall into error; the former could not, because, ifit acts only according to its own laws, the effect(the judgement) must necessarily accord withthese laws. But in accordance with the laws ofthe understanding consists the formal elementin all truth. In the senses there is no judge-ment—neither a true nor a false one. But, as wehave no source of cognition besides these two,it follows that error is caused solely by the

unobserved influence of the sensibility uponthe understanding. And thus it happens thatthe subjective grounds of a judgement and areconfounded with the objective, and cause themto deviate from their proper determination,*just as a body in motion would always of itselfproceed in a straight line, but if another impe-tus gives to it a different direction, it will thenstart off into a curvilinear line of motion. Todistinguish the peculiar action of the unders-tanding from the power which mingles with it,it is necessary to consider an erroneous judge-ment as the diagonal between two forces, thatdetermine the judgement in two different direc-tions, which, as it were, form an angle, and toresolve this composite operation into the sim-ple ones of the understanding and the sensibili-ty. In pure a priori judgements this must bedone by means of transcendental reflection,whereby, as has been already shown, each re-presentation has its place appointed in the co-rresponding faculty of cognition, and conse-

quently the influence of the one faculty uponthe other is made apparent.

[*Footnote: Sensibility, subjected to the unders-tanding, as the object upon which the unders-tanding employs its functions, is the source ofreal cognitions. But, in so far as it exercises aninfluence upon the action of the understandingand determines it to judgement, sensibility isitself the cause of error.]

It is not at present our business to treat of em-pirical illusory appearance (for example, opticalillusion), which occurs in the empirical applica-tion of otherwise correct rules of the unders-tanding, and in which the judgement is misledby the influence of imagination. Our purpose isto speak of transcendental illusory appearance,which influences principles—that are not evenapplied to experience, for in this case weshould possess a sure test of their correctness—but which leads us, in disregard of all the war-nings of criticism, completely beyond the empi-

rical employment of the categories and deludesus with the chimera of an extension of thesphere of the pure understanding. We shallterm those principles the application of whichis confined entirely within the limits of possibleexperience, immanent; those, on the otherhand, which transgress these limits, we shallcall transcendent principles. But by these latterI do not understand principles of the transcen-dental use or misuse of the categories, which isin reality a mere fault of the judgement whennot under due restraint from criticism, and the-refore not paying sufficient attention to the li-mits of the sphere in which the pure unders-tanding is allowed to exercise its functions; butreal principles which exhort us to break downall those barriers, and to lay claim to a perfectlynew field of cognition, which recognizes no lineof demarcation. Thus transcendental and trans-cendent are not identical terms. The principlesof the pure understanding, which we have al-ready propounded, ought to be of empirical

and not of transcendental use, that is, they arenot applicable to any object beyond the sphereof experience. A principle which removes theselimits, nay, which authorizes us to overstepthem, is called transcendent. If our criticism cansucceed in exposing the illusion in these pre-tended principles, those which are limited intheir employment to the sphere of experiencemay be called, in opposition to the others, im-manent principles of the pure understanding.

Logical illusion, which consists merely in theimitation of the form of reason (the illusion insophistical syllogisms), arises entirely from awant of due attention to logical rules. So soonas the attention is awakened to the case beforeus, this illusion totally disappears. Transcen-dental illusion, on the contrary, does not ceaseto exist, even after it has been exposed, and itsnothingness clearly perceived by means oftranscendental criticism. Take, for example, theillusion in the proposition: "The world must

have a beginning in time." The cause of this isas follows. In our reason, subjectively conside-red as a faculty of human cognition, there existfundamental rules and maxims of its exercise,which have completely the appearance of ob-jective principles. Now from this cause it hap-pens that the subjective necessity of a certainconnection of our conceptions, is regarded asan objective necessity of the determination ofthings in themselves. This illusion it is impossi-ble to avoid, just as we cannot avoid perceivingthat the sea appears to be higher at a distancethan it is near the shore, because we see theformer by means of higher rays than the latter,or, which is a still stronger case, as even theastronomer cannot prevent himself from seeingthe moon larger at its rising than some timeafterwards, although he is not deceived by thisillusion.

Transcendental dialectic will therefore contentitself with exposing the illusory appearance in

transcendental judgements, and guarding usagainst it; but to make it, as in the case of logi-cal illusion, entirely disappear and cease to beillusion is utterly beyond its power. For wehave here to do with a natural and unavoidableillusion, which rests upon subjective principlesand imposes these upon us as objective, whilelogical dialectic, in the detection of sophisms,has to do merely with an error in the logicalconsequence of the propositions, or with anartificially constructed illusion, in imitation ofthe natural error. There is, therefore, a naturaland unavoidable dialectic of pure reason—notthat in which the bungler, from want of therequisite knowledge, involves himself, nor thatwhich the sophist devises for the purpose ofmisleading, but that which is an inseparableadjunct of human reason, and which, even afterits illusions have been exposed, does not ceaseto deceive, and continually to lead reason intomomentary errors, which it becomes necessarycontinually to remove.

II. Of Pure Reason as the Seat of TranscendentalIllusory Appearance.

A. OF REASON IN GENERAL.

All our knowledge begins with sense, proceedsthence to understanding, and ends with reason,beyond which nothing higher can be discove-red in the human mind for elaborating the mat-ter of intuition and subjecting it to the highestunity of thought. At this stage of our inquiry itis my duty to give an explanation of this, thehighest faculty of cognition, and I confess I findmyself here in some difficulty. Of reason, as ofthe understanding, there is a merely formal,that is, logical use, in which it makes abstrac-tion of all content of cognition; but there is alsoa real use, inasmuch as it contains in itself thesource of certain conceptions and principles,which it does not borrow either from the sensesor the understanding. The former faculty hasbeen long defined by logicians as the faculty of

mediate conclusion in contradistinction to im-mediate conclusions (consequentiae immedia-tae); but the nature of the latter, which itselfgenerates conceptions, is not to be understoodfrom this definition. Now as a division of rea-son into a logical and a transcendental facultypresents itself here, it becomes necessary toseek for a higher conception of this source ofcognition which shall comprehend both con-ceptions. In this we may expect, according tothe analogy of the conceptions of the unders-tanding, that the logical conception will give usthe key to the transcendental, and that the tableof the functions of the former will present uswith the clue to the conceptions of reason.

In the former part of our transcendental logic,we defined the understanding to be the facultyof rules; reason may be distinguished from un-derstanding as the faculty of principles.

The term principle is ambiguous, and common-ly signifies merely a cognition that may be em-

ployed as a principle, although it is not in itself,and as regards its proper origin, entitled to thedistinction. Every general proposition, even ifderived from experience by the process of in-duction, may serve as the major in a syllogism;but it is not for that reason a principle. Mat-hematical axioms (for example, there can beonly one straight line between two points) aregeneral a priori cognitions, and are thereforerightly denominated principles, relatively tothe cases which can be subsumed under them.But I cannot for this reason say that I cognizethis property of a straight line from princi-ples—I cognize it only in pure intuition.

Cognition from principles, then, is that cogni-tion in which I cognize the particular in thegeneral by means of conceptions. Thus everysyllogism is a form of the deduction of a cogni-tion from a principle. For the major always gi-ves a conception, through which everythingthat is subsumed under the condition thereof is

cognized according to a principle. Now as eve-ry general cognition may serve as the major in asyllogism, and the understanding presents uswith such general a priori propositions, theymay be termed principles, in respect of theirpossible use.

But if we consider these principles of the pureunderstanding in relation to their origin, weshall find them to be anything rather than cog-nitions from conceptions. For they would noteven be possible a priori, if we could not relyon the assistance of pure intuition (in mathema-tics), or on that of the conditions of a possibleexperience. That everything that happens has acause, cannot be concluded from the generalconception of that which happens; on the con-trary the principle of causality instructs us as tothe mode of obtaining from that which happensa determinate empirical conception.

Synthetical cognitions from conceptions theunderstanding cannot supply, and they alone

are entitled to be called principles. At the sametime, all general propositions may be termedcomparative principles.

It has been a long-cherished wish—that (whoknows how late), may one day, be happily ac-complished—that the principles of the endlessvariety of civil laws should be investigated andexposed; for in this way alone can we find thesecret of simplifying legislation. But in this ca-se, laws are nothing more than limitations ofour freedom upon conditions under which itsubsists in perfect harmony with itself; theyconsequently have for their object that which iscompletely our own work, and of which weourselves may be the cause by means of theseconceptions. But how objects as things in them-selves- how the nature of things is subordina-ted to principles and is to be determined, ac-cording to conceptions, is a question which itseems well nigh impossible to answer. Be this,however, as it may—for on this point our in-

vestigation is yet to be made—it is at least ma-nifest from what we have said that cognitionfrom principles is something very differentfrom cognition by means of the understanding,which may indeed precede other cognitions inthe form of a principle, but in itself—in so far asit is synthetical—is neither based upon merethought, nor contains a general propositiondrawn from conceptions alone.

The understanding may be a faculty for theproduction of unity of phenomena by virtue ofrules; the reason is a faculty for the productionof unity of rules (of the understanding) underprinciples. Reason, therefore, never appliesdirectly to experience, or to any sensuous ob-ject; its object is, on the contrary, the unders-tanding, to the manifold cognition of which itgives a unity a priori by means of concep-tions—a unity which may be called rationalunity, and which is of a nature very different

from that of the unity produced by the unders-tanding.

The above is the general conception of the fa-culty of reason, in so far as it has been possibleto make it comprehensible in the absence ofexamples. These will be given in the sequel.

B. OF THE LOGICAL USE OF REASON.

A distinction is commonly made between thatwhich is immediately cognized and that whichis inferred or concluded. That in a figure whichis bounded by three straight lines there arethree angles, is an immediate cognition; butthat these angles are together equal to two rightangles, is an inference or conclusion. Now, aswe are constantly employing this mode ofthought and have thus become quite accusto-med to it, we no longer remark the above dis-

tinction, and, as in the case of the so-called de-ceptions of sense, consider as immediately per-ceived, what has really been inferred. In everyreasoning or syllogism, there is a fundamentalproposition, afterwards a second drawn fromit, and finally the conclusion, which connectsthe truth in the first with the truth in the se-cond—and that infallibly. If the judgement con-cluded is so contained in the first propositionthat it can be deduced from it without the me-ditation of a third notion, the conclusion is ca-lled immediate (consequentia immediata); Iprefer the term conclusion of the understan-ding. But if, in addition to the fundamentalcognition, a second judgement is necessary forthe production of the conclusion, it is called aconclusion of the reason. In the proposition: Allmen are mortal, are contained the propositions:Some men are mortal, Nothing that is not mor-tal is a man, and these are therefore immediateconclusions from the first. On the other hand,the proposition: all the learned are mortal, is

not contained in the main proposition (for theconception of a learned man does not occur init), and it can be deduced from the main propo-sition only by means of a mediating judgement.

In every syllogism I first cogitate a rule (themajor) by means of the understanding. In thenext place I subsume a cognition under the con-dition of the rule (and this is the minor) bymeans of the judgement. And finally I determi-ne my cognition by means of the predicate ofthe rule (this is the conclusio), consequently, Idetermine it a priori by means of the reason.The relations, therefore, which the major pro-position, as the rule, represents between a cog-nition and its condition, constitute the differentkinds of syllogisms. These are just threefold—analogously with all judgements, in so far asthey differ in the mode of expressing the rela-tion of a cognition in the understanding—namely, categorical, hypothetical, and disjunc-tive.

When as often happens, the conclusion is a jud-gement which may follow from other givenjudgements, through which a perfectly diffe-rent object is cogitated, I endeavour to discoverin the understanding whether the assertion inthis conclusion does not stand under certainconditions according to a general rule. If I findsuch a condition, and if the object mentioned inthe conclusion can be subsumed under the gi-ven condition, then this conclusion followsfrom a rule which is also valid for other objectsof cognition. From this we see that reason en-deavours to subject the great variety of thecognitions of the understanding to the smallestpossible number of principles (general condi-tions), and thus to produce in it the highestunity.

C. OF THE PURE USE OF REASON.

Can we isolate reason, and, if so, is it in thiscase a peculiar source of conceptions and jud-gements which spring from it alone, andthrough which it can be applied to objects; or isit merely a subordinate faculty, whose duty it isto give a certain form to given cognitions—aform which is called logical, and through whichthe cognitions of the understanding are subor-dinated to each other, and lower rules to higher(those, to wit, whose condition comprises in itssphere the condition of the others), in so far asthis can be done by comparison? This is thequestion which we have at present to answer.Manifold variety of rules and unity of princi-ples is a requirement of reason, for the purposeof bringing the understanding into completeaccordance with itself, just as understandingsubjects the manifold content of intuition toconceptions, and thereby introduces connectioninto it. But this principle prescribes no law to

objects, and does not contain any ground of thepossibility of cognizing or of determining themas such, but is merely a subjective law for theproper arrangement of the content of the un-derstanding. The purpose of this law is, by acomparison of the conceptions of the unders-tanding, to reduce them to the smallest possiblenumber, although, at the same time, it does notjustify us in demanding from objects themsel-ves such a uniformity as might contribute tothe convenience and the enlargement of thesphere of the understanding, or in expectingthat it will itself thus receive from them objecti-ve validity. In one word, the question is: "doesreason in itself, that is, does pure reason con-tain a priori synthetical principles and rules,and what are those principles?"

The formal and logical procedure of reason insyllogisms gives us sufficient information inregard to the ground on which the transcen-

dental principle of reason in its pure syntheticalcognition will rest.

1. Reason, as observed in the syllogistic pro-cess, is not applicable to intuitions, for the pur-pose of subjecting them to rules—for this is theprovince of the understanding with its catego-ries—but to conceptions and judgements. Ifpure reason does apply to objects and the intui-tion of them, it does so not immediately, butmediately- through the understanding and itsjudgements, which have a direct relation to thesenses and their intuition, for the purpose ofdetermining their objects. The unity of reason istherefore not the unity of a possible experience,but is essentially different from this unity,which is that of the understanding. That eve-rything which happens has a cause, is not aprinciple cognized and prescribed by reason.This principle makes the unity of experiencepossible and borrows nothing from reason,which, without a reference to possible expe-

rience, could never have produced by means ofmere conceptions any such synthetical unity.

2. Reason, in its logical use, endeavours to dis-cover the general condition of its judgement(the conclusion), and a syllogism is itself not-hing but a judgement by means of the sub-sumption of its condition under a general rule(the major). Now as this rule may itself be sub-jected to the same process of reason, and thusthe condition of the condition be sought (bymeans of a prosyllogism) as long as the processcan be continued, it is very manifest that thepeculiar principle of reason in its logical use isto find for the conditioned cognition of the un-derstanding the unconditioned whereby theunity of the former is completed.

But this logical maxim cannot be a principle ofpure reason, unless we admit that, if the condi-tioned is given, the whole series of conditionssubordinated to one another—a series which isconsequently itself unconditioned—is also gi-

ven, that is, contained in the object and its con-nection.

But this principle of pure reason is evidentlysynthetical; for, analytically, the conditionedcertainly relates to some condition, but not tothe unconditioned. From this principle alsothere must originate different synthetical pro-positions, of which the pure understanding isperfectly ignorant, for it has to do only withobjects of a possible experience, the cognitionand synthesis of which is always conditioned.The unconditioned, if it does really exist, mustbe especially considered in regard to the de-terminations which distinguish it from whate-ver is conditioned, and will thus afford us ma-terial for many a priori synthetical proposi-tions.

The principles resulting from this highest prin-ciple of pure reason will, however, be trans-cendent in relation to phenomena, that is tosay, it will be impossible to make any adequate

empirical use of this principle. It is thereforecompletely different from all principles of theunderstanding, the use made of which is entire-ly immanent, their object and purpose beingmerely the possibility of experience. Now ourduty in the transcendental dialectic is as fo-llows. To discover whether the principle thatthe series of conditions (in the synthesis ofphenomena, or of thought in general) extendsto the unconditioned is objectively true, or not;what consequences result therefrom affectingthe empirical use of the understanding, or rat-her whether there exists any such objectivelyvalid proposition of reason, and whether it isnot, on the contrary, a merely logical preceptwhich directs us to ascend perpetually to stillhigher conditions, to approach completeness inthe series of them, and thus to introduce intoour cognition the highest possible unity of rea-son. We must ascertain, I say, whether this re-quirement of reason has not been regarded, bya misunderstanding, as a transcendental prin-

ciple of pure reason, which postulates a tho-rough completeness in the series of conditionsin objects themselves. We must show, moreo-ver, the misconceptions and illusions that in-trude into syllogisms, the major proposition ofwhich pure reason has supplied—a propositionwhich has perhaps more of the character of apetitio than of a postulatum—and that proceedfrom experience upwards to its conditions. Thesolution of these problems is our task in trans-cendental dialectic, which we are about to ex-pose even at its source, that lies deep in humanreason. We shall divide it into two parts, thefirst of which will treat of the transcendent con-ceptions of pure reason, the second of trans-cendent and dialectical syllogisms.

BOOK I.

OF THE CONCEPTIONS OF PURE REA-SON.

The conceptions of pure reason—we do nothere speak of the possibility of them—are notobtained by reflection, but by inference or con-clusion. The conceptions of understanding arealso cogitated a priori antecedently to experien-ce, and render it possible; but they contain not-hing but the unity of reflection upon phenome-na, in so far as these must necessarily belong toa possible empirical consciousness. Throughthem alone are cognition and the determinationof an object possible. It is from them, accordin-gly, that we receive material for reasoning, andantecedently to them we possess no a prioriconceptions of objects from which they mightbe deduced, On the other hand, the sole basis oftheir objective reality consists in the necessity

imposed on them, as containing the intellectualform of all experience, of restricting their appli-cation and influence to the sphere of experien-ce.

But the term, conception of reason, or rationalconception, itself indicates that it does not con-fine itself within the limits of experience, be-cause its object-matter is a cognition, of whichevery empirical cognition is but a part—nay,the whole of possible experience may be itselfbut a part of it—a cognition to which no actualexperience ever fully attains, although it doesalways pertain to it. The aim of rational concep-tions is the comprehension, as that of the con-ceptions of understanding is the understandingof perceptions. If they contain the unconditio-ned, they relate to that to which all experienceis subordinate, but which is never itself an ob-ject of experience—that towards which reasontends in all its conclusions from experience,and by the standard of which it estimates the

degree of their empirical use, but which is ne-ver itself an element in an empirical synthesis.If, notwithstanding, such conceptions possessobjective validity, they may be called conceptusratiocinati (conceptions legitimately conclu-ded); in cases where they do not, they havebeen admitted on account of having the appea-rance of being correctly concluded, and may becalled conceptus ratiocinantes (sophistical con-ceptions). But as this can only be sufficientlydemonstrated in that part of our treatise whichrelates to the dialectical conclusions of reason,we shall omit any consideration of it in thisplace. As we called the pure conceptions of theunderstanding categories, we shall also distin-guish those of pure reason by a new name andcall them transcendental ideas. These terms,however, we must in the first place explain andjustify.

SECTION I—Of Ideas in General.

Despite the great wealth of words which Euro-pean languages possess, the thinker finds him-self often at a loss for an expression exactlysuited to his conception, for want of which he isunable to make himself intelligible either toothers or to himself. To coin new words is apretension to legislation in language which isseldom successful; and, before recourse is takento so desperate an expedient, it is advisable toexamine the dead and learned languages, withthe hope and the probability that we may theremeet with some adequate expression of thenotion we have in our minds. In this case, evenif the original meaning of the word has becomesomewhat uncertain, from carelessness or wantof caution on the part of the authors of it, it isalways better to adhere to and confirm its pro-per meaning—even although it may be doubt-ful whether it was formerly used in exactly thissense—than to make our labour vain by want

of sufficient care to render ourselves intelligi-ble.

For this reason, when it happens that thereexists only a single word to express a certainconception, and this word, in its usual accepta-tion, is thoroughly adequate to the conception,the accurate distinction of which from relatedconceptions is of great importance, we oughtnot to employ the expression improvidently, or,for the sake of variety and elegance of style, useit as a synonym for other cognate words. It isour duty, on the contrary, carefully to preserveits peculiar signification, as otherwise it easilyhappens that when the attention of the reader isno longer particularly attracted to the expres-sion, and it is lost amid the multitude of otherwords of very different import, the thoughtwhich it conveyed, and which it alone conve-yed, is lost with it.

Plato employed the expression idea in a waythat plainly showed he meant by it something

which is never derived from the senses, butwhich far transcends even the conceptions ofthe understanding (with which Aristotle occu-pied himself), inasmuch as in experience not-hing perfectly corresponding to them could befound. Ideas are, according to him, archetypesof things themselves, and not merely keys topossible experiences, like the categories. In hisview they flow from the highest reason, bywhich they have been imparted to human rea-son, which, however, exists no longer in itsoriginal state, but is obliged with great labourto recall by reminiscence—which is called phi-losophy—the old but now sadly obscuredideas. I will not here enter upon any literaryinvestigation of the sense which this sublimephilosopher attached to this expression. I shallcontent myself with remarking that it is not-hing unusual, in common conversation as wellas in written works, by comparing the thoughtswhich an author has delivered upon a subject,to understand him better than he understood

himself inasmuch as he may not have sufficien-tly determined his conception, and thus havesometimes spoken, nay even thought, in oppo-sition to his own opinions.

Plato perceived very clearly that our faculty ofcognition has the feeling of a much higher vo-cation than that of merely spelling out pheno-mena according to synthetical unity, for thepurpose of being able to read them as experien-ce, and that our reason naturally raises itself tocognitions far too elevated to admit of the pos-sibility of an object given by experience corres-ponding to them- cognitions which are nevert-heless real, and are not mere phantoms of thebrain.

This philosopher found his ideas especially inall that is practical,* that is, which rests uponfreedom, which in its turn ranks under cogni-tions that are the peculiar product of reason. Hewho would derive from experience the concep-tions of virtue, who would make (as many have

really done) that, which at best can but serve asan imperfectly illustrative example, a model foror the formation of a perfectly adequate idea onthe subject, would in fact transform virtue intoa nonentity changeable according to time andcircumstance and utterly incapable of beingemployed as a rule. On the contrary, every oneis conscious that, when any one is held up tohim as a model of virtue, he compares this so-called model with the true original which hepossesses in his own mind and values him ac-cording to this standard. But this standard isthe idea of virtue, in relation to which all possi-ble objects of experience are indeed serviceableas examples—proofs of the practicability in acertain degree of that which the conception ofvirtue demands—but certainly not as archety-pes. That the actions of man will never be inperfect accordance with all the requirements ofthe pure ideas of reason, does not prove thethought to be chimerical. For only through thisidea are all judgements as to moral merit or

demerit possible; it consequently lies at thefoundation of every approach to moral perfec-tion, however far removed from it the obstaclesin human nature- indeterminable as to de-gree—may keep us.

[*Footnote: He certainly extended the applica-tion of his conception to speculative cognitionsalso, provided they were given pure and com-pletely a priori, nay, even to mathematics, alt-hough this science cannot possess an objectotherwhere than in Possible experience. I can-not follow him in this, and as little can I followhim in his mystical deduction of these ideas, orin his hypostatization of them; although, intruth, the elevated and exaggerated languagewhich he employed in describing them is quitecapable of an interpretation more subdued andmore in accordance with fact and the nature ofthings.]

The Platonic Republic has become proverbial asan example—and a striking one—of imaginary

perfection, such as can exist only in the brain ofthe idle thinker; and Brucker ridicules the phi-losopher for maintaining that a prince can ne-ver govern well, unless he is participant in theideas. But we should do better to follow up thisthought and, where this admirable thinker lea-ves us without assistance, employ new effortsto place it in clearer light, rather than carelesslyfling it aside as useless, under the very misera-ble and pernicious pretext of impracticability. Aconstitution of the greatest possible humanfreedom according to laws, by which the libertyof every individual can consist with the libertyof every other (not of the greatest possible hap-piness, for this follows necessarily from theformer), is, to say the least, a necessary idea,which must be placed at the foundation notonly of the first plan of the constitution of astate, but of all its laws. And, in this, it not ne-cessary at the outset to take account of the obs-tacles which lie in our way—obstacles whichperhaps do not necessarily arise from the cha-

racter of human nature, but rather from theprevious neglect of true ideas in legislation. Forthere is nothing more pernicious and more un-worthy of a philosopher, than the vulgar ap-peal to a so-called adverse experience, whichindeed would not have existed, if those institu-tions had been established at the proper timeand in accordance with ideas; while, instead ofthis, conceptions, crude for the very reason thatthey have been drawn from experience, havemarred and frustrated all our better views andintentions. The more legislation and govern-ment are in harmony with this idea, the morerare do punishments become and thus it is qui-te reasonable to maintain, as Plato did, that in aperfect state no punishments at all would benecessary. Now although a perfect state maynever exist, the idea is not on that account theless just, which holds up this maximum as thearchetype or standard of a constitution, in or-der to bring legislative government always nea-rer and nearer to the greatest possible perfec-

tion. For at what precise degree human naturemust stop in its progress, and how wide mustbe the chasm which must necessarily exist bet-ween the idea and its realization, are problemswhich no one can or ought to determine- andfor this reason, that it is the destination of free-dom to overstep all assigned limits betweenitself and the idea.

But not only in that wherein human reason is areal causal agent and where ideas are operativecauses (of actions and their objects), that is tosay, in the region of ethics, but also in regard tonature herself, Plato saw clear proofs of an ori-gin from ideas. A plant, and animal, the regularorder of nature—probably also the dispositionof the whole universe—give manifest evidencethat they are possible only by means of andaccording to ideas; that, indeed, no one creatu-re, under the individual conditions of its exis-tence, perfectly harmonizes with the idea of themost perfect of its kind- just as little as man

with the idea of humanity, which neverthelesshe bears in his soul as the archetypal standardof his actions; that, notwithstanding, theseideas are in the highest sense individually, un-changeably, and completely determined, andare the original causes of things; and that thetotality of connected objects in the universe isalone fully adequate to that idea. Setting asidethe exaggerations of expression in the writingsof this philosopher, the mental power exhibitedin this ascent from the ectypal mode of regar-ding the physical world to the architectonicconnection thereof according to ends, that is,ideas, is an effort which deserves imitation andclaims respect. But as regards the principles ofethics, of legislation, and of religion, spheres inwhich ideas alone render experience possible,although they never attain to full expressiontherein, he has vindicated for himself a positionof peculiar merit, which is not appreciated onlybecause it is judged by the very empirical rules,the validity of which as principles is destroyed

by ideas. For as regards nature, experience pre-sents us with rules and is the source of truth,but in relation to ethical laws experience is theparent of illusion, and it is in the highest degreereprehensible to limit or to deduce the lawswhich dictate what I ought to do, from what isdone.

We must, however, omit the consideration ofthese important subjects, the development ofwhich is in reality the peculiar duty and dignityof philosophy, and confine ourselves for thepresent to the more humble but not less usefultask of preparing a firm foundation for thosemajestic edifices of moral science. For thisfoundation has been hitherto insecure from themany subterranean passages which reason inits confident but vain search for treasures hasmade in all directions. Our present duty is tomake ourselves perfectly acquainted with thetranscendental use made of pure reason, itsprinciples and ideas, that we may be able pro-

perly to determine and value its influence andreal worth. But before bringing these introduc-tory remarks to a close, I beg those who reallyhave philosophy at heart—and their number isbut small—if they shall find themselves con-vinced by the considerations following as wellas by those above, to exert themselves to pre-serve to the expression idea its original signifi-cation, and to take care that it be not lost amongthose other expressions by which all sorts ofrepresentations are loosely designated—thatthe interests of science may not thereby suffer.We are in no want of words to denominateadequately every mode of representation, wit-hout the necessity of encroaching upon termswhich are proper to others. The following is agraduated list of them. The genus is representa-tion in general (representatio). Under it standsrepresentation with consciousness (perceptio).A perception which relates solely to the subjectas a modification of its state, is a sensation (sen-satio), an objective perception is a cognition

(cognitio). A cognition is either an intuition or aconception (intuitus vel conceptus). The formerhas an immediate relation to the object and issingular and individual; the latter has but amediate relation, by means of a characteristicmark which may be common to several things.A conception is either empirical or pure. A pureconception, in so far as it has its origin in theunderstanding alone, and is not the conceptionof a pure sensuous image, is called notio. Aconception formed from notions, which trans-cends the possibility of experience, is an idea,or a conception of reason. To one who has ac-customed himself to these distinctions, it mustbe quite intolerable to hear the representationof the colour red called an idea. It ought noteven to be called a notion or conception of un-derstanding.

SECTION II. Of Transcendental Ideas.

Transcendental analytic showed us how themere logical form of our cognition can containthe origin of pure conceptions a priori, concep-tions which represent objects antecedently to allexperience, or rather, indicate the syntheticalunity which alone renders possible an empiri-cal cognition of objects. The form of judge-ments—converted into a conception of thesynthesis of intuitions—produced the catego-ries which direct the employment of the un-derstanding in experience. This considerationwarrants us to expect that the form of syllo-gisms, when applied to synthetical unity ofintuitions, following the rule of the categories,will contain the origin of particular a prioriconceptions, which we may call pure concep-tions of reason or transcendental ideas, andwhich will determine the use of the understan-ding in the totality of experience according toprinciples.

The function of reason in arguments consists inthe universality of a cognition according to con-ceptions, and the syllogism itself is a judgementwhich is determined a priori in the whole ex-tent of its condition. The proposition: "Caius ismortal," is one which may be obtained fromexperience by the aid of the understanding alo-ne; but my wish is to find a conception whichcontains the condition under which the predi-cate of this judgement is given—in this case, theconception of man—and after subsuming un-der this condition, taken in its whole extent (allmen are mortal), I determine according to it thecognition of the object thought, and say: "Caiusis mortal."

Hence, in the conclusion of a syllogism we res-trict a predicate to a certain object, after havingthought it in the major in its whole extent un-der a certain condition. This complete quantityof the extent in relation to such a condition iscalled universality (universalitas). To this co-

rresponds totality (universitas) of conditions inthe synthesis of intuitions. The transcendentalconception of reason is therefore nothing elsethan the conception of the totality of the condi-tions of a given conditioned. Now as the un-conditioned alone renders possible totality ofconditions, and, conversely, the totality of con-ditions is itself always unconditioned; a purerational conception in general can be definedand explained by means of the conception ofthe unconditioned, in so far as it contains a ba-sis for the synthesis of the conditioned.

To the number of modes of relation which theunderstanding cogitates by means of the cate-gories, the number of pure rational conceptionswill correspond. We must therefore seek for,first, an unconditioned of the categorical synt-hesis in a subject; secondly, of the hypotheticalsynthesis of the members of a series; thirdly, ofthe disjunctive synthesis of parts in a system.

There are exactly the same number of modes ofsyllogisms, each of which proceeds throughprosyllogisms to the unconditioned—one to thesubject which cannot be employed as predicate,another to the presupposition which supposesnothing higher than itself, and the third to anaggregate of the members of the complete divi-sion of a conception. Hence the pure rationalconceptions of totality in the synthesis of condi-tions have a necessary foundation in the natureof human reason—at least as modes of eleva-ting the unity of the understanding to the un-conditioned. They may have no valid applica-tion, corresponding to their transcendental em-ployment, in concreto, and be thus of no grea-ter utility than to direct the understanding how,while extending them as widely as possible, tomaintain its exercise and application in perfectconsistence and harmony.

But, while speaking here of the totality of con-ditions and of the unconditioned as the com-

mon title of all conceptions of reason, we againlight upon an expression which we find it im-possible to dispense with, and which nevert-heless, owing to the ambiguity attaching to itfrom long abuse, we cannot employ with safe-ty. The word absolute is one of the few wordswhich, in its original signification, was perfec-tly adequate to the conception it was intendedto convey—a conception which no other wordin the same language exactly suits, and theloss—or, which is the same thing, the incau-tious and loose employment—of which must befollowed by the loss of the conception itself.And, as it is a conception which occupies muchof the attention of reason, its loss would begreatly to the detriment of all transcendentalphilosophy. The word absolute is at presentfrequently used to denote that something canbe predicated of a thing considered in itself andintrinsically. In this sense absolutely possiblewould signify that which is possible in itself(interne)- which is, in fact, the least that one can

predicate of an object. On the other hand, it issometimes employed to indicate that a thing isvalid in all respects—for example, absolutesovereignty. Absolutely possible would in thissense signify that which is possible in all rela-tions and in every respect; and this is the mostthat can be predicated of the possibility of athing. Now these significations do in truth fre-quently coincide. Thus, for example, that whichis intrinsically impossible, is also impossible inall relations, that is, absolutely impossible. Butin most cases they differ from each other totocaelo, and I can by no means conclude that,because a thing is in itself possible, it is alsopossible in all relations, and therefore absolute-ly. Nay, more, I shall in the sequel show thatabsolute necessity does not by any means de-pend on internal necessity, and that, therefore,it must not be considered as synonymous withit. Of an opposite which is intrinsically impos-sible, we may affirm that it is in all respectsimpossible, and that, consequently, the thing

itself, of which this is the opposite, is absolutelynecessary; but I cannot reason conversely andsay, the opposite of that which is absolutelynecessary is intrinsically impossible, that is,that the absolute necessity of things is an inter-nal necessity. For this internal necessity is incertain cases a mere empty word with whichthe least conception cannot be connected, whilethe conception of the necessity of a thing in allrelations possesses very peculiar determina-tions. Now as the loss of a conception of greatutility in speculative science cannot be a matterof indifference to the philosopher, I trust thatthe proper determination and careful preserva-tion of the expression on which the conceptiondepends will likewise be not indifferent to him.

In this enlarged signification, then, shall I em-ploy the word absolute, in opposition to thatwhich is valid only in some particular respect;for the latter is restricted by conditions, the for-mer is valid without any restriction whatever.

Now the transcendental conception of reasonhas for its object nothing else than absolute to-tality in the synthesis of conditions and doesnot rest satisfied till it has attained to the abso-lutely, that is, in all respects and relations, un-conditioned. For pure reason leaves to the un-derstanding everything that immediately rela-tes to the object of intuition or rather to theirsynthesis in imagination. The former restrictsitself to the absolute totality in the employmentof the conceptions of the understanding andaims at carrying out the synthetical unity whichis cogitated in the category, even to the uncon-ditioned. This unity may hence be called therational unity of phenomena, as the other,which the category expresses, may be termedthe unity of the understanding. Reason, there-fore, has an immediate relation to the use of theunderstanding, not indeed in so far as the lattercontains the ground of possible experience (forthe conception of the absolute totality of condi-tions is not a conception that can be employed

in experience, because no experience is uncon-ditioned), but solely for the purpose of direc-ting it to a certain unity, of which the unders-tanding has no conception, and the aim ofwhich is to collect into an absolute whole allacts of the understanding. Hence the objectiveemployment of the pure conceptions of reasonis always transcendent, while that of the pureconceptions of the understanding must, accor-ding to their nature, be always immanent,inasmuch as they are limited to possible expe-rience.

I understand by idea a necessary conception ofreason, to which no corresponding object canbe discovered in the world of sense. Accordin-gly, the pure conceptions of reason at presentunder consideration are transcendental ideas.They are conceptions of pure reason, for theyregard all empirical cognition as determined bymeans of an absolute totality of conditions.They are not mere fictions, but natural and ne-

cessary products of reason, and have hence anecessary relation to the whole sphere of theexercise of the understanding. And, finally,they are transcendent, and overstep the limitsof all experiences, in which, consequently, noobject can ever be presented that would be per-fectly adequate to a transcendental idea. Whenwe use the word idea, we say, as regards itsobject (an object of the pure understanding), agreat deal, but as regards its subject (that is, inrespect of its reality under conditions of expe-rience), exceedingly little, because the idea, asthe conception of a maximum, can never becompletely and adequately presented in con-creto. Now, as in the merely speculative em-ployment of reason the latter is properly thesole aim, and as in this case the approximationto a conception, which is never attained in prac-tice, is the same thing as if the conception werenon-existent—it is commonly said of the con-ception of this kind, "it is only an idea." So wemight very well say, "the absolute totality of all

phenomena is only an idea," for, as we nevercan present an adequate representation of it, itremains for us a problem incapable of solution.On the other hand, as in the practical use of theunderstanding we have only to do with actionand practice according to rules, an idea of purereason can always be given really in concreto,although only partially, nay, it is the indispen-sable condition of all practical employment ofreason. The practice or execution of the idea isalways limited and defective, but neverthelesswithin indeterminable boundaries, consequen-tly always under the influence of the concep-tion of an absolute perfection. And thus thepractical idea is always in the highest degreefruitful, and in relation to real actions indispen-sably necessary. In the idea, pure reason pos-sesses even causality and the power of produ-cing that which its conception contains. Hencewe cannot say of wisdom, in a disparagingway, "it is only an idea." For, for the very rea-son that it is the idea of the necessary unity of

all possible aims, it must be for all practicalexertions and endeavours the primitive condi-tion and rule—a rule which, if not constitutive,is at least limitative.

Now, although we must say of the transcen-dental conceptions of reason, "they are onlyideas," we must not, on this account, look uponthem as superfluous and nugatory. For, alt-hough no object can be determined by them,they can be of great utility, unobserved and atthe basis of the edifice of the understanding, asthe canon for its extended and self-consistentexercise—a canon which, indeed, does not ena-ble it to cognize more in an object than it wouldcognize by the help of its own conceptions, butwhich guides it more securely in its cognition.Not to mention that they perhaps render possi-ble a transition from our conceptions of natureand the non-ego to the practical conceptions,and thus produce for even ethical ideas kee-ping, so to speak, and connection with the spe-

culative cognitions of reason. The explication ofall this must be looked for in the sequel.

But setting aside, in conformity with our origi-nal purpose, the consideration of the practicalideas, we proceed to contemplate reason in itsspeculative use alone, nay, in a still more res-tricted sphere, to wit, in the transcendental use;and here must strike into the same path whichwe followed in our deduction of the categories.That is to say, we shall consider the logicalform of the cognition of reason, that we maysee whether reason may not be thereby a sourceof conceptions which enables us to regard ob-jects in themselves as determined syntheticallya priori, in relation to one or other of the func-tions of reason.

Reason, considered as the faculty of a certainlogical form of cognition, is the faculty of con-clusion, that is, of mediate judgement—bymeans of the subsumption of the condition of apossible judgement under the condition of a

given judgement. The given judgement is thegeneral rule (major). The subsumption of thecondition of another possible judgement underthe condition of the rule is the minor. The ac-tual judgement, which enounces the assertionof the rule in the subsumed case, is the conclu-sion (conclusio). The rule predicates somethinggenerally under a certain condition. The condi-tion of the rule is satisfied in some particularcase. It follows that what was valid in generalunder that condition must also be consideredas valid in the particular case which satisfiesthis condition. It is very plain that reason at-tains to a cognition, by means of acts of the un-derstanding which constitute a series of condi-tions. When I arrive at the proposition, "Allbodies are changeable," by beginning with themore remote cognition (in which the concep-tion of body does not appear, but which ne-vertheless contains the condition of that con-ception), "All compound is changeable," byproceeding from this to a less remote cognition,

which stands under the condition of the former,"Bodies are compound," and hence to a third,which at length connects for me the remotecognition (changeable) with the one before me,"Consequently, bodies are changeable"—I havearrived at a cognition (conclusion) through aseries of conditions (premisses). Now everyseries, whose exponent (of the categorical orhypothetical judgement) is given, can be conti-nued; consequently the same procedure of rea-son conducts us to the ratiocinatio polysyllogis-tica, which is a series of syllogisms, that can becontinued either on the side of the conditions(per prosyllogismos) or of the conditioned (perepisyllogismos) to an indefinite extent.

But we very soon perceive that the chain orseries of prosyllogisms, that is, of deduced cog-nitions on the side of the grounds or conditionsof a given cognition, in other words, the ascen-ding series of syllogisms must have a very dif-ferent relation to the faculty of reason from that

of the descending series, that is, the progressiveprocedure of reason on the side of the conditio-ned by means of episyllogisms. For, as in theformer case the cognition (conclusio) is givenonly as conditioned, reason can attain to thiscognition only under the presupposition thatall the members of the series on the side of theconditions are given (totality in the series ofpremisses), because only under this supposi-tion is the judgement we may be consideringpossible a priori; while on the side of the condi-tioned or the inferences, only an incompleteand becoming, and not a presupposed or givenseries, consequently only a potential progres-sion, is cogitated. Hence, when a cognition iscontemplated as conditioned, reason is compe-lled to consider the series of conditions in anascending line as completed and given in theirtotality. But if the very same condition is consi-dered at the same time as the condition of othercognitions, which together constitute a series ofinferences or consequences in a descending

line, reason may preserve a perfect indifference,as to how far this progression may extend aparte posteriori, and whether the totality of thisseries is possible, because it stands in no needof such a series for the purpose of arriving atthe conclusion before it, inasmuch as this con-clusion is sufficiently guaranteed and determi-ned on grounds a parte priori. It may be thecase, that upon the side of the conditions theseries of premisses has a first or highest condi-tion, or it may not possess this, and so be a par-te priori unlimited; but it must, nevertheless,contain totality of conditions, even admittingthat we never could succeed in completely ap-prehending it; and the whole series must beunconditionally true, if the conditioned, whichis considered as an inference resulting from it,is to be held as true. This is a requirement ofreason, which announces its cognition as de-termined a priori and as necessary, either initself—and in this case it needs no grounds torest upon—or, if it is deduced, as a member of a

series of grounds, which is itself unconditiona-lly true.

SECTION III. System of TranscendentalIdeas.

We are not at present engaged with a logicaldialectic, which makes complete abstraction ofthe content of cognition and aims only at unvei-ling the illusory appearance in the form of sy-llogisms. Our subject is transcendental dialec-tic, which must contain, completely a priori, theorigin of certain cognitions drawn from purereason, and the origin of certain deduced con-ceptions, the object of which cannot be givenempirically and which therefore lie beyond thesphere of the faculty of understanding. Wehave observed, from the natural relation whichthe transcendental use of our cognition, in sy-llogisms as well as in judgements, must have to

the logical, that there are three kinds of dialec-tical arguments, corresponding to the threemodes of conclusion, by which reason attains tocognitions on principles; and that in all it is thebusiness of reason to ascend from the conditio-ned synthesis, beyond which the understan-ding never proceeds, to the unconditionedwhich the understanding never can reach.

Now the most general relations which can existin our representations are: 1st, the relation tothe subject; 2nd, the relation to objects, either asphenomena, or as objects of thought in general.If we connect this subdivision with the maindivision, all the relations of our representations,of which we can form either a conception or anidea, are threefold: 1. The relation to the sub-ject; 2. The relation to the manifold of the objectas a phenomenon; 3. The relation to all thingsin general.

Now all pure conceptions have to do in generalwith the synthetical unity of representations;

conceptions of pure reason (transcendentalideas), on the other hand, with the unconditio-nal synthetical unity of all conditions. It followsthat all transcendental ideas arrange themsel-ves in three classes, the first of which containsthe absolute (unconditioned) unity of the thin-king subject, the second the absolute unity ofthe series of the conditions of a phenomenon,the third the absolute unity of the condition ofall objects of thought in general.

The thinking subject is the object-matter of Psy-chology; the sum total of all phenomena (theworld) is the object-matter of Cosmology; andthe thing which contains the highest conditionof the possibility of all that is cogitable (thebeing of all beings) is the object-matter of allTheology. Thus pure reason presents us withthe idea of a transcendental doctrine of the soul(psychologia rationalis), of a transcendentalscience of the world (cosmologia rationalis),and finally of a transcendental doctrine of God

(theologia transcendentalis). Understandingcannot originate even the outline of any of the-se sciences, even when connected with the hig-hest logical use of reason, that is, all cogitablesyllogisms- for the purpose of proceeding fromone object (phenomenon) to all others, even tothe utmost limits of the empirical synthesis.They are, on the contrary, pure and genuineproducts, or problems, of pure reason.

What modi of the pure conceptions of reasonthese transcendental ideas are will be fully ex-posed in the following chapter. They follow theguiding thread of the categories. For pure rea-son never relates immediately to objects, but tothe conceptions of these contained in the un-derstanding. In like manner, it will be mademanifest in the detailed explanation of theseideas—how reason, merely through the synt-hetical use of the same function which it em-ploys in a categorical syllogism, necessarilyattains to the conception of the absolute unity

of the thinking subject—how the logical proce-dure in hypothetical ideas necessarily producesthe idea of the absolutely unconditioned in aseries of given conditions, and finally—how themere form of the disjunctive syllogism involvesthe highest conception of a being of all beings: athought which at first sight seems in the hig-hest degree paradoxical.

An objective deduction, such as we were ableto present in the case of the categories, is im-possible as regards these transcendental ideas.For they have, in truth, no relation to any ob-ject, in experience, for the very reason that theyare only ideas. But a subjective deduction ofthem from the nature of our reason is possible,and has been given in the present chapter.

It is easy to perceive that the sole aim of purereason is the absolute totality of the synthesison the side of the conditions, and that it doesnot concern itself with the absolute complete-ness on the Part of the conditioned. For of the

former alone does she stand in need, in order topreposit the whole series of conditions, andthus present them to the understanding a prio-ri. But if we once have a completely (and un-conditionally) given condition, there is no furt-her necessity, in proceeding with the series, fora conception of reason; for the understandingtakes of itself every step downward, from thecondition to the conditioned. Thus the trans-cendental ideas are available only for ascendingin the series of conditions, till we reach the un-conditioned, that is, principles. As regards des-cending to the conditioned, on the other hand,we find that there is a widely extensive logicaluse which reason makes of the laws of the un-derstanding, but that a transcendental use the-reof is impossible; and that when we form anidea of the absolute totality of such a synthesis,for example, of the whole series of all futurechanges in the world, this idea is a mere ensrationis, an arbitrary fiction of thought, and nota necessary presupposition of reason. For the

possibility of the conditioned presupposes thetotality of its conditions, but not of its conse-quences. Consequently, this conception is not atranscendental idea—and it is with these alonethat we are at present occupied.

Finally, it is obvious that there exists among thetranscendental ideas a certain connection andunity, and that pure reason, by means of them,collects all its cognitions into one system. Fromthe cognition of self to the cognition of theworld, and through these to the supreme being,the progression is so natural, that it seems toresemble the logical march of reason from thepremisses to the conclusion.* Now whetherthere lies unobserved at the foundation of theseideas an analogy of the same kind as exists bet-ween the logical and transcendental procedureof reason, is another of those questions, theanswer to which we must not expect till wearrive at a more advanced stage in our inqui-ries. In this cursory and preliminary view, we

have, meanwhile, reached our aim. For wehave dispelled the ambiguity which attached tothe transcendental conceptions of reason, fromtheir being commonly mixed up with otherconceptions in the systems of philosophers, andnot properly distinguished from the concep-tions of the understanding; we have exposedtheir origin and, thereby, at the same time theirdeterminate number, and presented them in asystematic connection, and have thus markedout and enclosed a definite sphere for pure rea-son.

[*Footnote: The science of Metaphysics has forthe proper object of its inquiries only threegrand ideas: GOD, FREEDOM, and IMMOR-TALITY, and it aims at showing, that the se-cond conception, conjoined with the first, mustlead to the third, as a necessary conclusion. Allthe other subjects with which it occupies itself,are merely means for the attainment and reali-

zation of these ideas. It does not require theseideas for the construction of a science of nature,but, on the contrary, for the purpose of passingbeyond the sphere of nature. A complete in-sight into and comprehension of them wouldrender Theology, Ethics, and, through the con-junction of both, Religion, solely dependent onthe speculative faculty of reason. In a systema-tic representation of these ideas the above-mentioned arrangement—the synthetical one—would be the most suitable; but in the investi-gation which must necessarily precede it, theanalytical, which reverses this arrangement,would be better adapted to our purpose, as in itwe should proceed from that which experienceimmediately presents to us—psychology, tocosmology, and thence to theology.]

BOOK II.

OF THE DIALECTICAL PROCEDURE OFPURE REASON.

It may be said that the object of a merely trans-cendental idea is something of which we haveno conception, although the idea may be a ne-cessary product of reason according to its ori-ginal laws. For, in fact, a conception of an objectthat is adequate to the idea given by reason, isimpossible. For such an object must be capableof being presented and intuited in a Possibleexperience. But we should express our meaningbetter, and with less risk of being misunders-tood, if we said that we can have no knowledgeof an object, which perfectly corresponds to anidea, although we may possess a problematicalconception thereof.

Now the transcendental (subjective) reality atleast of the pure conceptions of reason rests

upon the fact that we are led to such ideas by anecessary procedure of reason. There must the-refore be syllogisms which contain no empiricalpremisses, and by means of which we concludefrom something that we do know, to somethingof which we do not even possess a conception,to which we, nevertheless, by an unavoidableillusion, ascribe objective reality. Such argu-ments are, as regards their result, rather to betermed sophisms than syllogisms, althoughindeed, as regards their origin, they are verywell entitled to the latter name, inasmuch asthey are not fictions or accidental products ofreason, but are necessitated by its very nature.They are sophisms, not of men, but of pure rea-son herself, from which the Wisest cannot freehimself. After long labour he may be able toguard against the error, but he can never bethoroughly rid of the illusion which continuallymocks and misleads him.

Of these dialectical arguments there are threekinds, corresponding to the number of theideas which their conclusions present. In theargument or syllogism of the first class, I con-clude, from the transcendental conception ofthe subject contains no manifold, the absoluteunity of the subject itself, of which I cannot inthis manner attain to a conception. This dialec-tical argument I shall call the transcendentalparalogism. The second class of sophistical ar-guments is occupied with the transcendentalconception of the absolute totality of the seriesof conditions for a given phenomenon, and Iconclude, from the fact that I have always aself-contradictory conception of the uncondi-tioned synthetical unity of the series upon oneside, the truth of the opposite unity, of which Ihave nevertheless no conception. The conditionof reason in these dialectical arguments, I shallterm the antinomy of pure reason. Finally, ac-cording to the third kind of sophistical argu-ment, I conclude, from the totality of the condi-

tions of thinking objects in general, in so far asthey can be given, the absolute synthetical uni-ty of all conditions of the possibility of things ingeneral; that is, from things which I do notknow in their mere transcendental conception, Iconclude a being of all beings which I knowstill less by means of a transcendental concep-tion, and of whose unconditioned necessity Ican form no conception whatever. This dialecti-cal argument I shall call the ideal of pure rea-son.

CHAPTER I. Of the Paralogisms of PureReason.

The logical paralogism consists in the falsity ofan argument in respect of its form, be the con-tent what it may. But a transcendental paralo-gism has a transcendental foundation, and con-cludes falsely, while the form is correct and

unexceptionable. In this manner the paralogismhas its foundation in the nature of human rea-son, and is the parent of an unavoidable,though not insoluble, mental illusion.

We now come to a conception which was notinserted in the general list of transcendentalconceptions, and yet must be reckoned withthem, but at the same time without in the leastaltering, or indicating a deficiency in that table.This is the conception, or, if the term is prefe-rred, the judgement, "I think." But it is readilyperceived that this thought is as it were thevehicle of all conceptions in general, and con-sequently of transcendental conceptions also,and that it is therefore regarded as a transcen-dental conception, although it can have no pe-culiar claim to be so ranked, inasmuch as itsonly use is to indicate that all thought is ac-companied by consciousness. At the same time,pure as this conception is from empirical con-tent (impressions of the senses), it enables us to

distinguish two different kinds of objects. "I," asthinking, am an object of the internal sense, andam called soul. That which is an object of theexternal senses is called body. Thus the expres-sion, "I," as a thinking being, designates theobject-matter of psychology, which may becalled "the rational doctrine of the soul," inas-much as in this science I desire to know not-hing of the soul but what, independently of allexperience (which determines me in concreto),may be concluded from this conception "I," inso far as it appears in all thought.

Now, the rational doctrine of the soul is reallyan undertaking of this kind. For if the smallestempirical element of thought, if any particularperception of my internal state, were to be in-troduced among the grounds of cognition ofthis science, it would not be a rational, but anempirical doctrine of the soul. We have thusbefore us a pretended science, raised upon thesingle proposition, "I think," whose foundation

or want of foundation we may very properly,and agreeably with the nature of a transcenden-tal philosophy, here examine. It ought not to beobjected that in this proposition, which expres-ses the perception of one's self, an internal ex-perience is asserted, and that consequently therational doctrine of the soul which is foundedupon it, is not pure, but partly founded uponan empirical principle. For this internal percep-tion is nothing more than the mere appercep-tion, "I think," which in fact renders all trans-cendental conceptions possible, in which wesay, "I think substance, cause, etc." For internalexperience in general and its possibility, or per-ception in general, and its relation to other per-ceptions, unless some particular distinction ordetermination thereof is empirically given,cannot be regarded as empirical cognition, butas cognition of the empirical, and belongs to theinvestigation of the possibility of every expe-rience, which is certainly transcendental. Thesmallest object of experience (for example, only

pleasure or pain), that should be included inthe general representation of self-consciousness, would immediately change therational into an empirical psychology.

"I think" is therefore the only text of rationalpsychology, from which it must develop itswhole system. It is manifest that this thought,when applied to an object (myself), can containnothing but transcendental predicates thereof;because the least empirical predicate woulddestroy the purity of the science and its inde-pendence of all experience.

But we shall have to follow here the guidanceof the categories- only, as in the present case athing, "I," as thinking being, is at first given, weshall—not indeed change the order of the cate-gories as it stands in the table—but begin at thecategory of substance, by which at the a thingin itself is represented and proceeds backwardsthrough the series. The topic of the rationaldoctrine of the soul, from which everything else

it may contain must be deduced, is accordinglyas follows:

1 2 The Soul is SUBSTANCE As regards its quali-ty it is SIMPLE

3

As

regards

the

different

times

in

which

it

exists,

it

i

s

numerically

identica

l,

that

is

UNITY,

not

Plurality.

4

It

is

in

relation

to

possible

o

bjects

in

space*

[*Footnote: The reader, who may not so easilyperceive the psychological sense of these ex-pressions, taken here in their transcendentalabstraction, and cannot guess why the latter

attribute of the soul belongs to the category ofexistence, will find the expressions sufficientlyexplained and justified in the sequel. I have,moreover, to apologize for the Latin termswhich have been employed, instead of theirGerman synonyms, contrary to the rules of co-rrect writing. But I judged it better to sacrificeelegance to perspicuity.]

From these elements originate all the concep-tions of pure psychology, by combination alo-ne, without the aid of any other principle. Thissubstance, merely as an object of the internalsense, gives the conception of Immateriality; assimple substance, that of Incorruptibility; itsidentity, as intellectual substance, gives theconception of Personality; all these three toget-her, Spirituality. Its relation to objects in spacegives us the conception of connection (commer-cium) with bodies. Thus it represents thinkingsubstance as the principle of life in matter, thatis, as a soul (anima), and as the ground of Ani-

mality; and this, limited and determined by theconception of spirituality, gives us that of Im-mortality.

Now to these conceptions relate four paralo-gisms of a transcendental psychology, which isfalsely held to be a science of pure reason, tou-ching the nature of our thinking being. We can,however, lay at the foundation of this sciencenothing but the simple and in itself perfectlycontentless representation "I" which cannoteven be called a conception, but merely a cons-ciousness which accompanies all conceptions.By this "I," or "He," or "It," who or which thinks,nothing more is represented than a transcen-dental subject of thought = x, which is cognizedonly by means of the thoughts that are its pre-dicates, and of which, apart from these, wecannot form the least conception. Hence in aperpetual circle, inasmuch as we must alwaysemploy it, in order to frame any judgementrespecting it. And this inconvenience we find it

impossible to rid ourselves of, because cons-ciousness in itself is not so much a representa-tion distinguishing a particular object, as a formof representation in general, in so far as it maybe termed cognition; for in and by cognitionalone do I think anything.

It must, however, appear extraordinary at firstsight that the condition under which I think,and which is consequently a property of mysubject, should be held to be likewise valid forevery existence which thinks, and that we canpresume to base upon a seemingly empiricalproposition a judgement which is apodeicticand universal, to wit, that everything whichthinks is constituted as the voice of my cons-ciousness declares it to be, that is, as a self-conscious being. The cause of this belief is to befound in the fact that we necessarily attribute tothings a priori all the properties which constitu-te conditions under which alone we can cogita-te them. Now I cannot obtain the least repre-

sentation of a thinking being by means of ex-ternal experience, but solely through self-consciousness. Such objects are consequentlynothing more than the transference of thisconsciousness of mine to other things whichcan only thus be represented as thinkingbeings. The proposition, "I think," is, in the pre-sent case, understood in a problematical sense,not in so far as it contains a perception of anexistence (like the Cartesian "Cogito, ergosum"),[Footnote: "I think, therefore I am."] butin regard to its mere possibility—for the purpo-se of discovering what properties may be infe-rred from so simple a proposition and predica-ted of the subject of it.

If at the foundation of our pure rational cogni-tion of thinking beings there lay more than themere Cogito—if we could likewise call in aidobservations on the play of our thoughts, andthe thence derived natural laws of the thinkingself, there would arise an empirical psychology

which would be a kind of physiology of theinternal sense and might possibly be capable ofexplaining the phenomena of that sense. But itcould never be available for discovering thoseproperties which do not belong to possible ex-perience (such as the quality of simplicity), norcould it make any apodeictic enunciation onthe nature of thinking beings: it would therefo-re not be a rational psychology.

Now, as the proposition "I think" (in the pro-blematical sense) contains the form of everyjudgement in general and is the constant ac-companiment of all the categories, it is manifestthat conclusions are drawn from it only by atranscendental employment of the understan-ding. This use of the understanding excludesall empirical elements; and we cannot, as hasbeen shown above, have any favourable con-ception beforehand of its procedure. We shalltherefore follow with a critical eye this proposi-tion through all the predicaments of pure psy-

chology; but we shall, for brevity's sake, allowthis examination to proceed in an uninterrup-ted connection.

Before entering on this task, however, the fo-llowing general remark may help to quickenour attention to this mode of argument. It is notmerely through my thinking that I cognize anobject, but only through my determining a gi-ven intuition in relation to the unity of cons-ciousness in which all thinking consists. It fo-llows that I cognize myself, not through mybeing conscious of myself as thinking, but onlywhen I am conscious of the intuition of myselfas determined in relation to the function ofthought. All the modi of self-consciousness inthought are hence not conceptions of objects(conceptions of the understanding—categories);they are mere logical functions, which do notpresent to thought an object to be cognized,and cannot therefore present my Self as an ob-ject. Not the consciousness of the determining,

but only that of the determinable self, that is, ofmy internal intuition (in so far as the manifoldcontained in it can be connected conformablywith the general condition of the unity of ap-perception in thought), is the object.

1. In all judgements I am the determining sub-ject of that relation which constitutes a judge-ment. But that the I which thinks, must be con-sidered as in thought always a subject, and as athing which cannot be a predicate to thought, isan apodeictic and identical proposition. But thisproposition does not signify that I, as an object,am, for myself, a self-subsistent being or subs-tance. This latter statement- an ambitious one—requires to be supported by data which are notto be discovered in thought; and are perhaps(in so far as I consider the thinking self merelyas such) not to be discovered in the thinkingself at all.

2. That the I or Ego of apperception, and conse-quently in all thought, is singular or simple,

and cannot be resolved into a plurality of sub-jects, and therefore indicates a logically simplesubject—this is self-evident from the very con-ception of an Ego, and is consequently an ana-lytical proposition. But this is not tantamountto declaring that the thinking Ego is a simplesubstance- for this would be a synthetical pro-position. The conception of substance alwaysrelates to intuitions, which with me cannot beother than sensuous, and which consequentlylie completely out of the sphere of the unders-tanding and its thought: but to this sphere be-longs the affirmation that the Ego is simple inthought. It would indeed be surprising, if theconception of "substance," which in other casesrequires so much labour to distinguish from theother elements presented by intuition—somuch trouble, too, to discover whether it can besimple (as in the case of the parts of matter)—should be presented immediately to me, as ifby revelation, in the poorest mental representa-tion of all.

3. The proposition of the identity of my Selfamidst all the manifold representations ofwhich I am conscious, is likewise a propositionlying in the conceptions themselves, and is con-sequently analytical. But this identity of thesubject, of which I am conscious in all its repre-sentations, does not relate to or concern theintuition of the subject, by which it is given asan object. This proposition cannot thereforeenounce the identity of the person, by which isunderstood the consciousness of the identity ofits own substance as a thinking being in allchange and variation of circumstances. To pro-ve this, we should require not a mere analysisof the proposition, but synthetical judgementsbased upon a given intuition.

4. I distinguish my own existence, as that of athinking being, from that of other things exter-nal to me—among which my body also is rec-koned. This is also an analytical proposition,for other things are exactly those which I think

as different or distinguished from myself. Butwhether this consciousness of myself is possiblewithout things external to me; and whethertherefore I can exist merely as a thinking being(without being man)—cannot be known or in-ferred from this proposition.

Thus we have gained nothing as regards thecognition of myself as object, by the analysis ofthe consciousness of my Self in thought. Thelogical exposition of thought in general is mis-taken for a metaphysical determination of theobject.

Our Critique would be an investigation utterlysuperfluous, if there existed a possibility ofproving a priori, that all thinking beings are inthemselves simple substances, as such, therefo-re, possess the inseparable attribute of persona-lity, and are conscious of their existence apartfrom and unconnected with matter. For weshould thus have taken a step beyond theworld of sense, and have penetrated into the

sphere of noumena; and in this case the rightcould not be denied us of extending our know-ledge in this sphere, of establishing ourselves,and, under a favouring star, appropriating toourselves possessions in it. For the proposition:"Every thinking being, as such, is simple subs-tance," is an a priori synthetical proposition;because in the first place it goes beyond theconception which is the subject of it, and addsto the mere notion of a thinking being the modeof its existence, and in the second place annexesa predicate (that of simplicity) to the latter con-ception—a predicate which it could not havediscovered in the sphere of experience. Itwould follow that a priori synthetical proposi-tions are possible and legitimate, not only, aswe have maintained, in relation to objects ofpossible experience, and as principles of thepossibility of this experience itself, but are ap-plicable to things in themselves—an inferencewhich makes an end of the whole of this Criti-que, and obliges us to fall back on the old mode

of metaphysical procedure. But indeed thedanger is not so great, if we look a little closerinto the question.

There lurks in the procedure of rational Psycho-logy a paralogism, which is represented in thefollowing syllogism:

That which cannot be cogitated otherwise thanas subject, does not exist otherwise than as sub-ject, and is therefore substance.

A thinking being, considered merely as such,cannot be cogitated otherwise than as subject.

Therefore it exists also as such, that is, as subs-tance.

In the major we speak of a being that can becogitated generally and in every relation, con-sequently as it may be given in intuition. But inthe minor we speak of the same being only inso far as it regards itself as subject, relatively to

thought and the unity of consciousness, but notin relation to intuition, by which it is presentedas an object to thought. Thus the conclusion ishere arrived at by a Sophisma figurae dictio-nis.*

[*Footnote: Thought is taken in the two premis-ses in two totally different senses. In the majorit is considered as relating and applying to ob-jects in general, consequently to objects of intui-tion also. In the minor, we understand it as re-lating merely to self-consciousness. In this sen-se, we do not cogitate an object, but merely therelation to the self-consciousness of the subject,as the form of thought. In the former premisswe speak of things which cannot be cogitatedotherwise than as subjects. In the second, we donot speak of things, but of thought (all objectsbeing abstracted), in which the Ego is alwaysthe subject of consciousness. Hence the conclu-sion cannot be, "I cannot exist otherwise than assubject"; but only "I can, in cogitating my exis-

tence, employ my Ego only as the subject of thejudgement." But this is an identical proposition,and throws no light on the mode of my existen-ce.]

That this famous argument is a mere paralo-gism, will be plain to any one who will consi-der the general remark which precedes ourexposition of the principles of the pure unders-tanding, and the section on noumena. For itwas there proved that the conception of a thing,which can exist per se—only as a subject andnever as a predicate, possesses no objectivereality; that is to say, we can never know whet-her there exists any object to correspond to theconception; consequently, the conception isnothing more than a conception, and from it wederive no proper knowledge. If this conceptionis to indicate by the term substance, an objectthat can be given, if it is to become a cognition,we must have at the foundation of the cogni-tion a permanent intuition, as the indispensable

condition of its objective reality. For throughintuition alone can an object be given. But ininternal intuition there is nothing permanent,for the Ego is but the consciousness of mythought. If then, we appeal merely to thought,we cannot discover the necessary condition ofthe application of the conception of substance—that is, of a subject existing per se—to the sub-ject as a thinking being. And thus the concep-tion of the simple nature of substance, which isconnected with the objective reality of this con-ception, is shown to be also invalid, and to be,in fact, nothing more than the logical qualitati-ve unity of self-consciousness in thought;whilst we remain perfectly ignorant whetherthe subject is composite or not.

Refutation of the Argument of Mendelssohn fortheSubstantiality or Permanence of the Soul.

This acute philosopher easily perceived theinsufficiency of the common argument whichattempts to prove that the soul—it being gran-ted that it is a simple being—cannot perish bydissolution or decomposition; he saw it is notimpossible for it to cease to be by extinction, ordisappearance. He endeavoured to prove in hisPhaedo, that the soul cannot be annihilated, byshowing that a simple being cannot cease toexist. Inasmuch as, he said, a simple existencecannot diminish, nor gradually lose portions ofits being, and thus be by degrees reduced tonothing (for it possesses no parts, and thereforeno multiplicity), between the moment in whichit is, and the moment in which it is not, no timecan be discovered—which is impossible. Butthis philosopher did not consider that, grantingthe soul to possess this simple nature, whichcontains no parts external to each other andconsequently no extensive quantity, we cannotrefuse to it any less than to any other being,intensive quantity, that is, a degree of reality in

regard to all its faculties, nay, to all that consti-tutes its existence. But this degree of reality canbecome less and less through an infinite seriesof smaller degrees. It follows, therefore, thatthis supposed substance—this thing, the per-manence of which is not assured in any otherway, may, if not by decomposition, by gradualloss (remissio) of its powers (consequently byelanguescence, if I may employ this expres-sion), be changed into nothing. For conscious-ness itself has always a degree, which may belessened.* Consequently the faculty of beingconscious may be diminished; and so with allother faculties. The permanence of the soul,therefore, as an object of the internal sense, re-mains undemonstrated, nay, even indemons-trable. Its permanence in life is evident, per se,inasmuch as the thinking being (as man) is toitself, at the same time, an object of the externalsenses. But this does not authorize the rationalpsychologist to affirm, from mere conceptions,its permanence beyond life.*[2]

[*Footnote: Clearness is not, as logicians main-tain, the consciousness of a representation. Fora certain degree of consciousness, which maynot, however, be sufficient for recollection, is tobe met with in many dim representations. Forwithout any consciousness at all, we should notbe able to recognize any difference in the obs-cure representations we connect; as we reallycan do with many conceptions, such as those ofright and justice, and those of the musician,who strikes at once several notes in improvi-sing a piece of music. But a representation isclear, in which our consciousness is sufficientfor the consciousness of the difference of thisrepresentation from others. If we are only cons-cious that there is a difference, but are notconscious of the difference—that is, what thedifference is- the representation must be ter-med obscure. There is, consequently, an infiniteseries of degrees of consciousness down to itsentire disappearance.]

[*[2]Footnote: There are some who think theyhave done enough to establish a new possibilityin the mode of the existence of souls, when theyhave shown that there is no contradiction intheir hypotheses on this subject. Such are thosewho affirm the possibility of thought—ofwhich they have no other knowledge than whatthey derive from its use in connecting empiricalintuitions presented in this our human life—after this life has ceased. But it is very easy toembarrass them by the introduction of counter-possibilities, which rest upon quite as good afoundation. Such, for example, is the possibilityof the division of a simple substance into seve-ral substances; and conversely, of the coalitionof several into one simple substance. For, alt-hough divisibility presupposes composition, itdoes not necessarily require a composition ofsubstances, but only of the degrees (of the seve-ral faculties) of one and the same substance.Now we can cogitate all the powers and facul-ties of the soul—even that of consciousness—as

diminished by one half, the substance still re-maining. In the same way we can represent toourselves without contradiction, this oblitera-ted half as preserved, not in the soul, but wit-hout it; and we can believe that, as in this caseevery thing that is real in the soul, and has adegree—consequently its entire existence—hasbeen halved, a particular substance would ariseout of the soul. For the multiplicity, which hasbeen divided, formerly existed, but not as amultiplicity of substances, but of every realityas the quantum of existence in it; and the unityof substance was merely a mode of existence,which by this division alone has been trans-formed into a plurality of subsistence. In thesame manner several simple substances mightcoalesce into one, without anything being lostexcept the plurality of subsistence, inasmuch asthe one substance would contain the degree ofreality of all the former substances. Perhaps,indeed, the simple substances, which appearunder the form of matter, might (not indeed by

a mechanical or chemical influence upon eachother, but by an unknown influence, of whichthe former would be but the phenomenal ap-pearance), by means of such a dynamical divi-sion of the parent-souls, as intensive quantities,produce other souls, while the former repairedthe loss thus sustained with new matter of thesame sort. I am far from allowing any value tosuch chimeras; and the principles of our analy-tic have clearly proved that no other than anempirical use of the categories—that of subs-tance, for example—is possible. But if the ratio-nalist is bold enough to construct, on the mereauthority of the faculty of thought—withoutany intuition, whereby an object is given—aself-subsistent being, merely because the unityof apperception in thought cannot allow him tobelieve it a composite being, instead of decla-ring, as he ought to do, that he is unable to ex-plain the possibility of a thinking nature; whatought to hinder the materialist, with as comple-te an independence of experience, to employ

the principle of the rationalist in a directly op-posite manner— still preserving the formalunity required by his opponent?]

If, now, we take the above propositions—asthey must be accepted as valid for all thinkingbeings in the system of rational psychology—insynthetical connection, and proceed, from thecategory of relation, with the proposition: "Allthinking beings are, as such, substances," back-wards through the series, till the circle is com-pleted; we come at last to their existence, ofwhich, in this system of rational psychology,substances are held to be conscious, indepen-dently of external things; nay, it is asserted that,in relation to the permanence which is a neces-sary characteristic of substance, they can ofthemselves determine external things. It fo-llows that idealism—at least problematicalidealism, is perfectly unavoidable in this ratio-nalistic system. And, if the existence of out-ward things is not held to be requisite to the

determination of the existence of a substance intime, the existence of these outward things atall, is a gratuitous assumption which remainswithout the possibility of a proof.

But if we proceed analytically—the "I think" asa proposition containing in itself an existence asgiven, consequently modality being the princi-ple—and dissect this proposition, in order toascertain its content, and discover whether andhow this Ego determines its existence in timeand space without the aid of anything external;the propositions of rationalistic psychologywould not begin with the conception of a thin-king being, but with a reality, and the proper-ties of a thinking being in general would bededuced from the mode in which this reality iscogitated, after everything empirical had beenabstracted; as is shown in the following table:

1 I think,

2 3 as Subject, as simple Subject,

4 as identical Subject, in every state of my thought.

Now, inasmuch as it is not determined in thissecond proposition, whether I can exist and becogitated only as subject, and not also as a pre-dicate of another being, the conception of asubject is here taken in a merely logical sense;and it remains undetermined, whether subs-tance is to be cogitated under the conception ornot. But in the third proposition, the absoluteunity of apperception- the simple Ego in therepresentation to which all connection and se-paration, which constitute thought, relate, is of

itself important; even although it presents uswith no information about the constitution orsubsistence of the subject. Apperception is so-mething real, and the simplicity of its nature isgiven in the very fact of its possibility. Now inspace there is nothing real that is at the sametime simple; for points, which are the only sim-ple things in space, are merely limits, but notconstituent parts of space. From this follows theimpossibility of a definition on the basis of ma-terialism of the constitution of my Ego as a me-rely thinking subject. But, because my existenceis considered in the first proposition as given,for it does not mean, "Every thinking beingexists" (for this would be predicating of themabsolute necessity), but only, "I exist thinking";the proposition is quite empirical, and containsthe determinability of my existence merely inrelation to my representations in time. But as Irequire for this purpose something that is per-manent, such as is not given in internal intui-tion; the mode of my existence, whether as

substance or as accident, cannot be determinedby means of this simple self-consciousness.Thus, if materialism is inadequate to explainthe mode in which I exist, spiritualism is like-wise as insufficient; and the conclusion is thatwe are utterly unable to attain to any knowled-ge of the constitution of the soul, in so far asrelates to the possibility of its existence apartfrom external objects.

And, indeed, how should it be possible, merelyby the aid of the unity of consciousness—whichwe cognize only for the reason that it is indis-pensable to the possibility of experience—topass the bounds of experience (our existence inthis life); and to extend our cognition to thenature of all thinking beings by means of theempirical—but in relation to every sort of intui-tion, perfectly undetermined—proposition, "Ithink"?

There does not then exist any rational psycho-logy as a doctrine furnishing any addition to

our knowledge of ourselves. It is nothing morethan a discipline, which sets impassable limitsto speculative reason in this region of thought,to prevent it, on the one hand, from throwingitself into the arms of a soulless materialism,and, on the other, from losing itself in the ma-zes of a baseless spiritualism. It teaches us toconsider this refusal of our reason to give anysatisfactory answer to questions which reachbeyond the limits of this our human life, as ahint to abandon fruitless speculation; and todirect, to a practical use, our knowledge of our-selves—which, although applicable only toobjects of experience, receives its principlesfrom a higher source, and regulates its proce-dure as if our destiny reached far beyond theboundaries of experience and life.

From all this it is evident that rational psycho-logy has its origin in a mere misunderstanding.The unity of consciousness, which lies at thebasis of the categories, is considered to be an

intuition of the subject as an object; and thecategory of substance is applied to the intui-tion. But this unity is nothing more than theunity in thought, by which no object is given; towhich therefore the category of substance—which always presupposes a given intuition-cannot be applied. Consequently, the subjectcannot be cognized. The subject of the catego-ries cannot, therefore, for the very reason that itcogitates these, frame any conception of itselfas an object of the categories; for, to cogitatethese, it must lay at the foundation its own pu-re self-consciousness—the very thing that itwishes to explain and describe. In like manner,the subject, in which the representation of timehas its basis, cannot determine, for this veryreason, its own existence in time. Now, if thelatter is impossible, the former, as an attempt todetermine itself by means of the categories as athinking being in general, is no less so.*

[*Footnote: The "I think" is, as has been alreadystated, an empirical proposition, and containsthe proposition, "I exist." But I cannot say, "Eve-rything, which thinks, exists"; for in this casethe property of thought would constitute allbeings possessing it, necessary beings. Hencemy existence cannot be considered as an infe-rence from the proposition, "I think," as Descar-tes maintained—because in this case the majorpremiss, "Everything, which thinks, exists,"must precede—but the two propositions areidentical. The proposition, "I think," expressesan undetermined empirical intuition, that per-ception (proving consequently that sensation,which must belong to sensibility, lies at thefoundation of this proposition); but it precedesexperience, whose province it is to determinean object of perception by means of the catego-ries in relation to time; and existence in thisproposition is not a category, as it does not ap-ply to an undetermined given object, but onlyto one of which we have a conception, and

about which we wish to know whether it doesor does not exist, out of, and apart from thisconception. An undetermined perception signi-fies here merely something real that has beengiven, only, however, to thought in general—but not as a phenomenon, nor as a thing in it-self (noumenon), but only as something thatreally exists, and is designated as such in theproposition, "I think." For it must be remarkedthat, when I call the proposition, "I think," anempirical proposition, I do not thereby meanthat the Ego in the proposition is an empiricalrepresentation; on the contrary, it is purely inte-llectual, because it belongs to thought in gene-ral. But without some empirical representation,which presents to the mind material forthought, the mental act, "I think," would nottake place; and the empirical is only the condi-tion of the application or employment of thepure intellectual faculty.]

Thus, then, appears the vanity of the hope ofestablishing a cognition which is to extend itsrule beyond the limits of experience—a cogni-tion which is one of the highest interests ofhumanity; and thus is proved the futility of theattempt of speculative philosophy in this re-gion of thought. But, in this interest of thought,the severity of criticism has rendered to reasona not unimportant service, by the demonstra-tion of the impossibility of making any dogma-tical affirmation concerning an object of expe-rience beyond the boundaries of experience.She has thus fortified reason against all affirma-tions of the contrary. Now, this can be accom-plished in only two ways. Either our proposi-tion must be proved apodeictically; or, if this isunsuccessful, the sources of this inability mustbe sought for, and, if these are discovered toexist in the natural and necessary limitation ofour reason, our opponents must submit to thesame law of renunciation and refrain from ad-vancing claims to dogmatic assertion.

But the right, say rather the necessity to admit afuture life, upon principles of the practical con-joined with the speculative use of reason, haslost nothing by this renunciation; for the merelyspeculative proof has never had any influenceupon the common reason of men. It standsupon the point of a hair, so that even theschools have been able to preserve it from fa-lling only by incessantly discussing it and spin-ning it like a top; and even in their eyes it hasnever been able to present any safe foundationfor the erection of a theory. The proofs whichhave been current among men, preserve theirvalue undiminished; nay, rather gain in clear-ness and unsophisticated power, by the rejec-tion of the dogmatical assumptions of specula-tive reason. For reason is thus confined withinher own peculiar province—the arrangement ofends or aims, which is at the same time thearrangement of nature; and, as a practical facul-ty, without limiting itself to the latter, it is justi-fied in extending the former, and with it our

own existence, beyond the boundaries of expe-rience and life. If we turn our attention to theanalogy of the nature of living beings in thisworld, in the consideration of which reason isobliged to accept as a principle that no organ,no faculty, no appetite is useless, and that not-hing is superfluous, nothing disproportionateto its use, nothing unsuited to its end; but that,on the contrary, everything is perfectly confor-med to its destination in life—we shall find thatman, who alone is the final end and aim of thisorder, is still the only animal that seems to beexcepted from it. For his natural gifts—not me-rely as regards the talents and motives thatmay incite him to employ them, but especiallythe moral law in him—stretch so far beyond allmere earthly utility and advantage, that he feelshimself bound to prize the mere consciousnessof probity, apart from all advantageous conse-quences— even the shadowy gift of post-humous fame—above everything; and he isconscious of an inward call to constitute him-

self, by his conduct in this world—without re-gard to mere sublunary interests—the citizen ofa better. This mighty, irresistible proof—accompanied by an ever-increasing knowledgeof the conformability to a purpose in everyt-hing we see around us, by the conviction of theboundless immensity of creation, by the cons-ciousness of a certain illimitableness in the pos-sible extension of our knowledge, and by a de-sire commensurate therewith—remains tohumanity, even after the theoretical cognitionof ourselves has failed to establish the necessityof an existence after death.

Conclusion of the Solution of the PsychologicalParalogism.

The dialectical illusion in rational psychologyarises from our confounding an idea of reason(of a pure intelligence) with the conception—in

every respect undetermined—of a thinkingbeing in general. I cogitate myself in behalf of apossible experience, at the same time makingabstraction of all actual experience; and infertherefrom that I can be conscious of myselfapart from experience and its empirical condi-tions. I consequently confound the possibleabstraction of my empirically determined exis-tence with the supposed consciousness of apossible separate existence of my thinking self;and I believe that I cognize what is substantialin myself as a transcendental subject, when Ihave nothing more in thought than the unity ofconsciousness, which lies at the basis of all de-termination of cognition.

The task of explaining the community of thesoul with the body does not properly belong tothe psychology of which we are here speaking;because it proposes to prove the personality ofthe soul apart from this communion (afterdeath), and is therefore transcendent in the pro-

per sense of the word, although occupying it-self with an object of experience—only in so far,however, as it ceases to be an object of expe-rience. But a sufficient answer may be found tothe question in our system. The difficultywhich lies in the execution of this task consists,as is well known, in the presupposed heteroge-neity of the object of the internal sense (thesoul) and the objects of the external senses;inasmuch as the formal condition of the intui-tion of the one is time, and of that of the otherspace also. But if we consider that both kinds ofobjects do not differ internally, but only in sofar as the one appears externally to the other—consequently, that what lies at the basis of phe-nomena, as a thing in itself, may not be hetero-geneous; this difficulty disappears. There thenremains no other difficulty than is to be foundin the question—how a community of substan-ces is possible; a question which lies out of theregion of psychology, and which the reader,after what in our analytic has been said of pri-

mitive forces and faculties, will easily judge tobe also beyond the region of human cognition.

GENERAL REMARK

On the Transition from Rational Psychology toCosmology.

The proposition, "I think," or, "I exist thinking,"is an empirical proposition. But such a proposi-tion must be based on empirical intuition, andthe object cogitated as a phenomenon; and thusour theory appears to maintain that the soul,even in thought, is merely a phenomenon; andin this way our consciousness itself, in fact,abuts upon nothing.

Thought, per se, is merely the purely sponta-neous logical function which operates to con-nect the manifold of a possible intuition; and it

does not represent the subject of consciousnessas a phenomenon—for this reason alone, that itpays no attention to the question whether themode of intuiting it is sensuous or intellectual. Itherefore do not represent myself in thoughteither as I am, or as I appear to myself; I merelycogitate myself as an object in general, of themode of intuiting which I make abstraction.When I represent myself as the subject ofthought, or as the ground of thought, thesemodes of representation are not related to thecategories of substance or of cause; for these arefunctions of thought applicable only to our sen-suous intuition. The application of these cate-gories to the Ego would, however, be necessa-ry, if I wished to make myself an object ofknowledge. But I wish to be conscious of my-self only as thinking; in what mode my Self isgiven in intuition, I do not consider, and it maybe that I, who think, am a phenomenon—although not in so far as I am a thinking being;but in the consciousness of myself in mere

thought I am a being, though this conscious-ness does not present to me any property ofthis being as material for thought.

But the proposition, "I think," in so far as it de-clares, "I exist thinking," is not the mere repre-sentation of a logical function. It determines thesubject (which is in this case an object also) inrelation to existence; and it cannot be givenwithout the aid of the internal sense, whoseintuition presents to us an object, not as a thingin itself, but always as a phenomenon. In thisproposition there is therefore something moreto be found than the mere spontaneity ofthought; there is also the receptivity of intui-tion, that is, my thought of myself applied tothe empirical intuition of myself. Now, in thisintuition the thinking self must seek the condi-tions of the employment of its logical functionsas categories of substance, cause, and so forth;not merely for the purpose of distinguishingitself as an object in itself by means of the re-

presentation "I," but also for the purpose ofdetermining the mode of its existence, that is, ofcognizing itself as noumenon. But this is im-possible, for the internal empirical intuition issensuous, and presents us with nothing butphenomenal data, which do not assist the objectof pure consciousness in its attempt to cognizeitself as a separate existence, but are useful onlyas contributions to experience.

But, let it be granted that we could discover,not in experience, but in certain firmly-established a priori laws of the use of pure rea-son— laws relating to our existence, authorityto consider ourselves as legislating a priori inrelation to our own existence and as determi-ning this existence; we should, on this supposi-tion, find ourselves possessed of a spontaneity,by which our actual existence would be deter-minable, without the aid of the conditions ofempirical intuition. We should also becomeaware that in the consciousness of our existence

there was an a priori content, which wouldserve to determine our own existence—an exis-tence only sensuously determinable—relatively, however, to a certain internal facultyin relation to an intelligible world.

But this would not give the least help to theattempts of rational psychology. For this won-derful faculty, which the consciousness of themoral law in me reveals, would present mewith a principle of the determination of myown existence which is purely intellectual—butby what predicates? By none other than thosewhich are given in sensuous intuition. Thus Ishould find myself in the same position in ra-tional psychology which I formerly occupied,that is to say, I should find myself still in needof sensuous intuitions, in order to give signifi-cance to my conceptions of substance and cau-se, by means of which alone I can possess aknowledge of myself: but these intuitions cannever raise me above the sphere of experience. I

should be justified, however, in applying theseconceptions, in regard to their practical use,which is always directed to objects of experien-ce—in conformity with their analogical signifi-cance when employed theoretically—to free-dom and its subject. At the same time, I shouldunderstand by them merely the logical func-tions of subject and predicate, of principle andconsequence, in conformity with which all ac-tions are so determined, that they are capableof being explained along with the laws of natu-re, conformably to the categories of substanceand cause, although they originate from a verydifferent principle. We have made these obser-vations for the purpose of guarding againstmisunderstanding, to which the doctrine of ourintuition of self as a phenomenon is exposed.We shall have occasion to perceive their utilityin the sequel.

CHAPTER II. The Antinomy of Pure Rea-son.

We showed in the introduction to this part ofour work, that all transcendental illusion ofpure reason arose from dialectical arguments,the schema of which logic gives us in its threeformal species of syllogisms—just as the cate-gories find their logical schema in the fourfunctions of all judgements. The first kind ofthese sophistical arguments related to the un-conditioned unity of the subjective conditionsof all representations in general (of the subjector soul), in correspondence with the categoricalsyllogisms, the major of which, as the principle,enounces the relation of a predicate to a subject.The second kind of dialectical argument willtherefore be concerned, following the analogywith hypothetical syllogisms, with the uncondi-tioned unity of the objective conditions in thephenomenon; and, in this way, the theme of thethird kind to be treated of in the following

chapter will be the unconditioned unity of theobjective conditions of the possibility of objectsin general.

But it is worthy of remark that the transcenden-tal paralogism produced in the mind only aone-third illusion, in regard to the idea of thesubject of our thought; and the conceptions ofreason gave no ground to maintain the contraryproposition. The advantage is completely onthe side of Pneumatism; although this theoryitself passes into naught, in the crucible of purereason.

Very different is the case when we apply rea-son to the objective synthesis of phenomena.Here, certainly, reason establishes, with muchplausibility, its principle of unconditioned uni-ty; but it very soon falls into such contradic-tions that it is compelled, in relation to cosmo-logy, to renounce its pretensions.

For here a new phenomenon of human reasonmeets us—a perfectly natural antithetic, whichdoes not require to be sought for by subtle sop-histry, but into which reason of itself unavoi-dably falls. It is thereby preserved, to be sure,from the slumber of a fancied conviction—which a merely one-sided illusion produces;but it is at the same time compelled, either, onthe one hand, to abandon itself to a despairingscepticism, or, on the other, to assume a dog-matical confidence and obstinate persistence incertain assertions, without granting a fair hea-ring to the other side of the question. Either isthe death of a sound philosophy, although theformer might perhaps deserve the title of theeuthanasia of pure reason.

Before entering this region of discord and con-fusion, which the conflict of the laws of purereason (antinomy) produces, we shall presentthe reader with some considerations, in expla-nation and justification of the method we in-

tend to follow in our treatment of this subject. Iterm all transcendental ideas, in so far as theyrelate to the absolute totality in the synthesis ofphenomena, cosmical conceptions; partly onaccount of this unconditioned totality, onwhich the conception of the world-whole isbased—a conception, which is itself an idea—partly because they relate solely to the synt-hesis of phenomena—the empirical synthesis;while, on the other hand, the absolute totalityin the synthesis of the conditions of all possiblethings gives rise to an ideal of pure reason,which is quite distinct from the cosmical con-ception, although it stands in relation with it.Hence, as the paralogisms of pure reason laidthe foundation for a dialectical psychology, theantinomy of pure reason will present us withthe transcendental principles of a pretendedpure (rational) cosmology—not, however, todeclare it valid and to appropriate it, but—asthe very term of a conflict of reason sufficiently

indicates, to present it as an idea which cannotbe reconciled with phenomena and experience.

SECTION I. System of Cosmological Ideas.

That We may be able to enumerate with syste-matic precision these ideas according to a prin-ciple, we must remark, in the first place, that itis from the understanding alone that pure andtranscendental conceptions take their origin;that the reason does not properly give birth toany conception, but only frees the conception ofthe understanding from the unavoidable limita-tion of a possible experience, and thus endea-vours to raise it above the empirical, though itmust still be in connection with it. This happensfrom the fact that, for a given conditioned, rea-son demands absolute totality on the side of theconditions (to which the understanding sub-mits all phenomena), and thus makes of the

category a transcendental idea. This it does thatit may be able to give absolute completeness tothe empirical synthesis, by continuing it to theunconditioned (which is not to be found in ex-perience, but only in the idea). Reason requiresthis according to the principle: If the conditio-ned is given the whole of the conditions, andconsequently the absolutely unconditioned, isalso given, whereby alone the former was pos-sible. First, then, the transcendental ideas areproperly nothing but categories elevated to theunconditioned; and they may be arranged in atable according to the titles of the latter. But,secondly, all the categories are not available forthis purpose, but only those in which the synt-hesis constitutes a series—of conditions subor-dinated to, not co-ordinated with, each other.Absolute totality is required of reason only inso far as concerns the ascending series of theconditions of a conditioned; not, consequently,when the question relates to the descendingseries of consequences, or to the aggregate of

the co-ordinated conditions of these conse-quences. For, in relation to a given conditioned,conditions are presupposed and considered tobe given along with it. On the other hand, asthe consequences do not render possible theirconditions, but rather presuppose them—in theconsideration of the procession of consequen-ces (or in the descent from the given conditionto the conditioned), we may be quite unconcer-ned whether the series ceases or not; and theirtotality is not a necessary demand of reason.

Thus we cogitate—and necessarily—a giventime completely elapsed up to a given moment,although that time is not determinable by us.But as regards time future, which is not thecondition of arriving at the present, in order toconceive it; it is quite indifferent whether weconsider future time as ceasing at some point,or as prolonging itself to infinity. Take, forexample, the series m, n, o, in which n is givenas conditioned in relation to m, but at the same

time as the condition of o, and let the seriesproceed upwards from the conditioned n to m(l, k, i, etc.), and also downwards from the con-dition n to the conditioned o (p, q, r, etc.)—Imust presuppose the former series, to be able toconsider n as given, and n is according to rea-son (the totality of conditions) possible only bymeans of that series. But its possibility does notrest on the following series o, p, q, r, which forthis reason cannot be regarded as given, butonly as capable of being given (dabilis).

I shall term the synthesis of the series on theside of the conditions—from that nearest to thegiven phenomenon up to the more remote—regressive; that which proceeds on the side ofthe conditioned, from the immediate conse-quence to the more remote, I shall call the pro-gressive synthesis. The former proceeds in an-tecedentia, the latter in consequentia. The cos-mological ideas are therefore occupied with thetotality of the regressive synthesis, and proceed

in antecedentia, not in consequentia. When thelatter takes place, it is an arbitrary and not anecessary problem of pure reason; for we re-quire, for the complete understanding of whatis given in a phenomenon, not the consequen-ces which succeed, but the grounds or princi-ples which precede.

In order to construct the table of ideas in co-rrespondence with the table of categories, wetake first the two primitive quanta of all ourintuitions, time and space. Time is in itself aseries (and the formal condition of all series),and hence, in relation to a given present, wemust distinguish a priori in it the antecedentiaas conditions (time past) from the consequentia(time future). Consequently, the transcendentalidea of the absolute totality of the series of theconditions of a given conditioned, relates mere-ly to all past time. According to the idea of rea-son, the whole past time, as the condition of thegiven moment, is necessarily cogitated as gi-

ven. But, as regards space, there exists in it nodistinction between progressus and regressus;for it is an aggregate and not a series—its partsexisting together at the same time. I can consi-der a given point of time in relation to past timeonly as conditioned, because this given mo-ment comes into existence only through thepast time rather through the passing of the pre-ceding time. But as the parts of space are notsubordinated, but co-ordinated to each other,one part cannot be the condition of the possibi-lity of the other; and space is not in itself, liketime, a series. But the synthesis of the manifoldparts of space—(the syntheses whereby weapprehend space)—is nevertheless successive;it takes place, therefore, in time, and contains aseries. And as in this series of aggregated spa-ces (for example, the feet in a rood), beginningwith a given portion of space, those which con-tinue to be annexed form the condition of thelimits of the former—the measurement of aspace must also be regarded as a synthesis of

the series of the conditions of a given conditio-ned. It differs, however, in this respect fromthat of time, that the side of the conditioned isnot in itself distinguishable from the side of thecondition; and, consequently, regressus andprogressus in space seem to be identical. But,inasmuch as one part of space is not given, butonly limited, by and through another, we mustalso consider every limited space as conditio-ned, in so far as it presupposes some other spa-ce as the condition of its limitation, and so on.As regards limitation, therefore, our procedurein space is also a regressus, and the transcen-dental idea of the absolute totality of the synt-hesis in a series of conditions applies to spacealso; and I am entitled to demand the absolutetotality of the phenomenal synthesis in space aswell as in time. Whether my demand can besatisfied is a question to be answered in thesequel.

Secondly, the real in space—that is, matter—isconditioned. Its internal conditions are its parts,and the parts of parts its remote conditions; sothat in this case we find a regressive synthesis,the absolute totality of which is a demand ofreason. But this cannot be obtained otherwisethan by a complete division of parts, wherebythe real in matter becomes either nothing orthat which is not matter, that is to say, the sim-ple. Consequently we find here also a series ofconditions and a progress to the unconditioned.

Thirdly, as regards the categories of a real rela-tion between phenomena, the category of subs-tance and its accidents is not suitable for theformation of a transcendental idea; that is tosay, reason has no ground, in regard to it, toproceed regressively with conditions. For acci-dents (in so far as they inhere in a substance)are co-ordinated with each other, and do notconstitute a series. And, in relation to substan-ce, they are not properly subordinated to it, but

are the mode of existence of the substance itself.The conception of the substantial might nevert-heless seem to be an idea of the transcendentalreason. But, as this signifies nothing more thanthe conception of an object in general, whichsubsists in so far as we cogitate in it merely atranscendental subject without any predicates;and as the question here is of an unconditionedin the series of phenomena—it is clear that thesubstantial can form no member thereof. Thesame holds good of substances in community,which are mere aggregates and do not form aseries. For they are not subordinated to eachother as conditions of the possibility of eachother; which, however, may be affirmed of spa-ces, the limits of which are never determined inthemselves, but always by some other space. Itis, therefore, only in the category of causalitythat we can find a series of causes to a giveneffect, and in which we ascend from the latter,as the conditioned, to the former as the condi-tions, and thus answer the question of reason.

Fourthly, the conceptions of the possible, theactual, and the necessary do not conduct us toany series—excepting only in so far as the con-tingent in existence must always be regarded asconditioned, and as indicating, according to alaw of the understanding, a condition, underwhich it is necessary to rise to a higher, till inthe totality of the series, reason arrives at un-conditioned necessity.

There are, accordingly, only four cosmologicalideas, corresponding with the four titles of thecategories. For we can select only such as ne-cessarily furnish us with a series in the synt-hesis of the manifold.

1 The absolute Completeness of the COMPOSITION of the given totality of all phenomena.

2 The absolute Completeness of the DIVISION of given totality in a phenomenon.

3 The absolute Completeness of the ORIGINATION of a phenomenon.

4 The absolute Completeness of the DEPENDENCE of the EXISTENCE of what is changeable in a phenomenon.

We must here remark, in the first place, that theidea of absolute totality relates to nothing butthe exposition of phenomena, and therefore notto the pure conception of a totality of things.

Phenomena are here, therefore, regarded asgiven, and reason requires the absolute comple-teness of the conditions of their possibility, inso far as these conditions constitute a series-consequently an absolutely (that is, in everyrespect) complete synthesis, whereby a pheno-menon can be explained according to the lawsof the understanding.

Secondly, it is properly the unconditioned alo-ne that reason seeks in this serially and regres-sively conducted synthesis of conditions. Itwishes, to speak in another way, to attain tocompleteness in the series of premisses, so as torender it unnecessary to presuppose others.This unconditioned is always contained in theabsolute totality of the series, when we endea-vour to form a representation of it in thought.But this absolutely complete synthesis is itselfbut an idea; for it is impossible, at least beforehand, to know whether any such synthesis ispossible in the case of phenomena. When we

represent all existence in thought by means ofpure conceptions of the understanding, withoutany conditions of sensuous intuition, we maysay with justice that for a given conditioned thewhole series of conditions subordinated to eachother is also given; for the former is only giventhrough the latter. But we find in the case ofphenomena a particular limitation of the modein which conditions are given, that is, throughthe successive synthesis of the manifold of in-tuition, which must be complete in the regress.Now whether this completeness is sensuouslypossible, is a problem. But the idea of it lies inthe reason—be it possible or impossible to con-nect with the idea adequate empirical concep-tions. Therefore, as in the absolute totality ofthe regressive synthesis of the manifold in aphenomenon (following the guidance of thecategories, which represent it as a series of con-ditions to a given conditioned) the unconditio-ned is necessarily contained—it being still leftunascertained whether and how this totality

exists; reason sets out from the idea of totality,although its proper and final aim is the uncon-ditioned—of the whole series, or of a part the-reof.

This unconditioned may be cogitated—either asexisting only in the entire series, all the mem-bers of which therefore would be without ex-ception conditioned and only the totality abso-lutely unconditioned—and in this case the re-gressus is called infinite; or the absolutely un-conditioned is only a part of the series, to whichthe other members are subordinated, but whichIs not itself submitted to any other condition.*In the former case the series is a parte prioriunlimited (without beginning), that is, infinite,and nevertheless completely given. But the re-gress in it is never completed, and can only becalled potentially infinite. In the second casethere exists a first in the series. This first is ca-lled, in relation to past time, the beginning ofthe world; in relation to space, the limit of the

world; in relation to the parts of a given limitedwhole, the simple; in relation to causes, absolu-te spontaneity (liberty); and in relation to theexistence of changeable things, absolute physi-cal necessity.

[*Footnote: The absolute totality of the series ofconditions to a given conditioned is alwaysunconditioned; because beyond it there exist noother conditions, on which it might depend.But the absolute totality of such a series is onlyan idea, or rather a problematical conception,the possibility of which must be investigated-particularly in relation to the mode in whichthe unconditioned, as the transcendental ideawhich is the real subject of inquiry, may be con-tained therein.]

We possess two expressions, world and nature,which are generally interchanged. The firstdenotes the mathematical total of all phenome-na and the totality of their synthesis—in itsprogress by means of composition, as well as

by division. And the world is termed nature,*when it is regarded as a dynamical whole—when our attention is not directed to the aggre-gation in space and time, for the purpose ofcogitating it as a quantity, but to the unity inthe existence of phenomena. In this case thecondition of that which happens is called a cau-se; the unconditioned causality of the cause in aphenomenon is termed liberty; the conditionedcause is called in a more limited sense a naturalcause. The conditioned in existence is termedcontingent, and the unconditioned necessary.The unconditioned necessity of phenomenamay be called natural necessity.

[*Footnote: Nature, understood adjective (for-maliter), signifies the complex of the determi-nations of a thing, connected according to aninternal principle of causality. On the otherhand, we understand by nature, substantive(materialiter), the sum total of phenomena, inso far as they, by virtue of an internal principle

of causality, are connected with each otherthroughout. In the former sense we speak of thenature of liquid matter, of fire, etc., and employthe word only adjective; while, if speaking ofthe objects of nature, we have in our minds theidea of a subsisting whole.]

The ideas which we are at present engaged indiscussing I have called cosmological ideas;partly because by the term world is understoodthe entire content of all phenomena, and ourideas are directed solely to the unconditionedamong phenomena; partly also, because world,in the transcendental sense, signifies the abso-lute totality of the content of existing things,and we are directing our attention only to thecompleteness of the synthesis—although, pro-perly, only in regression. In regard to the factthat these ideas are all transcendent, and, alt-hough they do not transcend phenomena asregards their mode, but are concerned solelywith the world of sense (and not with noume-

na), nevertheless carry their synthesis to a de-gree far above all possible experience—it stillseems to me that we can, with perfect proprie-ty, designate them cosmical conceptions. Asregards the distinction between the mathemati-cally and the dynamically unconditioned whichis the aim of the regression of the synthesis, Ishould call the two former, in a more limitedsignification, cosmical conceptions, the remai-ning two transcendent physical conceptions.This distinction does not at present seem to beof particular importance, but we shall after-wards find it to be of some value.

SECTION II. Antithetic of Pure Reason.

Thetic is the term applied to every collection ofdogmatical propositions. By antithetic I do notunderstand dogmatical assertions of the oppo-site, but the self-contradiction of seemingly

dogmatical cognitions (thesis cum antithesis),in none of which we can discover any decidedsuperiority. Antithetic is not, therefore, occu-pied with one-sided statements, but is engagedin considering the contradictory nature of thegeneral cognitions of reason and its causes.Transcendental antithetic is an investigationinto the antinomy of pure reason, its causes andresult. If we employ our reason not merely inthe application of the principles of the unders-tanding to objects of experience, but venturewith it beyond these boundaries, there arisecertain sophistical propositions or theorems.These assertions have the following peculiari-ties: They can find neither confirmation norconfutation in experience; and each is in itselfnot only self-consistent, but possesses condi-tions of its necessity in the very nature of rea-son—only that, unluckily, there exist just asvalid and necessary grounds for maintainingthe contrary proposition.

The questions which naturally arise in the con-sideration of this dialectic of pure reason, aretherefore: 1st. In what propositions is pure rea-son unavoidably subject to an antinomy? 2nd.What are the causes of this antinomy? 3rd.Whether and in what way can reason free itselffrom this self-contradiction?

A dialectical proposition or theorem of purereason must, according to what has been said,be distinguishable from all sophistical proposi-tions, by the fact that it is not an answer to anarbitrary question, which may be raised at themere pleasure of any person, but to one whichhuman reason must necessarily encounter in itsprogress. In the second place, a dialectical pro-position, with its opposite, does not carry theappearance of a merely artificial illusion, whichdisappears as soon as it is investigated, but anatural and unavoidable illusion, which, evenwhen we are no longer deceived by it, conti-

nues to mock us and, although rendered harm-less, can never be completely removed.

This dialectical doctrine will not relate to theunity of understanding in empirical concep-tions, but to the unity of reason in pure ideas.The conditions of this doctrine are—inasmuchas it must, as a synthesis according to rules, beconformable to the understanding, and at thesame time as the absolute unity of the synt-hesis, to the reason—that, if it is adequate to theunity of reason, it is too great for the unders-tanding, if according with the understanding, itis too small for the reason. Hence arises a mu-tual opposition, which cannot be avoided, dowhat we will.

These sophistical assertions of dialectic open, asit were, a battle-field, where that side obtainsthe victory which has been permitted to makethe attack, and he is compelled to yield whohas been unfortunately obliged to stand on thedefensive. And hence, champions of ability,

whether on the right or on the wrong side, arecertain to carry away the crown of victory, ifthey only take care to have the right to makethe last attack, and are not obliged to sustainanother onset from their opponent. We caneasily believe that this arena has been oftentrampled by the feet of combatants, that manyvictories have been obtained on both sides, butthat the last victory, decisive of the affair bet-ween the contending parties, was won by himwho fought for the right, only if his adversarywas forbidden to continue the tourney. As im-partial umpires, we must lay aside entirely theconsideration whether the combatants are figh-ting for the right or for the wrong side, for thetrue or for the false, and allow the combat to befirst decided. Perhaps, after they have weariedmore than injured each other, they will disco-ver the nothingness of their cause of quarreland part good friends.

This method of watching, or rather of origina-ting, a conflict of assertions, not for the purposeof finally deciding in favour of either side, butto discover whether the object of the struggle isnot a mere illusion, which each strives in vainto reach, but which would be no gain evenwhen reached—this procedure, I say, may betermed the sceptical method. It is thoroughlydistinct from scepticism—the principle of atechnical and scientific ignorance, which un-dermines the foundations of all knowledge, inorder, if possible, to destroy our belief and con-fidence therein. For the sceptical method aimsat certainty, by endeavouring to discover in aconflict of this kind, conducted honestly andintelligently on both sides, the point of misun-derstanding; just as wise legislators derive,from the embarrassment of judges in lawsuits,information in regard to the defective and ill-defined parts of their statutes. The antinomywhich reveals itself in the application of laws, isfor our limited wisdom the best criterion of

legislation. For the attention of reason, which inabstract speculation does not easily becomeconscious of its errors, is thus roused to themomenta in the determination of its principles.

But this sceptical method is essentially peculiarto transcendental philosophy, and can perhapsbe dispensed with in every other field of inves-tigation. In mathematics its use would be ab-surd; because in it no false assertions can longremain hidden, inasmuch as its demonstrationsmust always proceed under the guidance ofpure intuition, and by means of an always evi-dent synthesis. In experimental philosophy,doubt and delay may be very useful; but nomisunderstanding is possible, which cannot beeasily removed; and in experience means ofsolving the difficulty and putting an end to thedissension must at last be found, whether soo-ner or later. Moral philosophy can always ex-hibit its principles, with their practical conse-quences, in concreto—at least in possible expe-

riences, and thus escape the mistakes and am-biguities of abstraction. But transcendentalpropositions, which lay claim to insight beyondthe region of possible experience, cannot, onthe one hand, exhibit their abstract synthesis inany a priori intuition, nor, on the other, exposea lurking error by the help of experience.Transcendental reason, therefore, presents uswith no other criterion than that of an attemptto reconcile such assertions, and for this purpo-se to permit a free and unrestrained conflictbetween them. And this we now proceed toarrange.*

[*Footnote: The antinomies stand in the orderof the four transcendental ideas above detai-led.]

FIRST CONFLICT OF THE TRANSCEN-DENTAL IDEAS.

THESIS.

The world has a beginning in time, and is alsolimited in regard to space.

PROOF.Granted that the world has no beginning intime; up to every given moment of time, aneternity must have elapsed, and therewith pas-sed away an infinite series of successive condi-tions or states of things in the world. Now theinfinity of a series consists in the fact that itnever can be completed by means of a successi-ve synthesis. It follows that an infinite seriesalready elapsed is impossible and that, conse-quently, a beginning of the world is a necessary

condition of its existence. And this was the firstthing to be proved.

As regards the second, let us take the oppositefor granted. In this case, the world must be aninfinite given total of coexistent things. Nowwe cannot cogitate the dimensions of a quanti-ty, which is not given within certain limits of anintuition,* in any other way than by means ofthe synthesis of its parts, and the total of such aquantity only by means of a completed synt-hesis, or the repeated addition of unity to itself.Accordingly, to cogitate the world, which fillsall spaces, as a whole, the successive synthesisof the parts of an infinite world must be lookedupon as completed, that is to say, an infinitetime must be regarded as having elapsed in theenumeration of all co-existing things; which isimpossible. For this reason an infinite aggregateof actual things cannot be considered as a givenwhole, consequently, not as a contempora-neously given whole. The world is consequen-

tly, as regards extension in space, not infinite,but enclosed in limits. And this was the secondthing to be proved.

[*Footnote: We may consider an undeterminedquantity as a whole, when it is enclosed withinlimits, although we cannot construct or ascer-tain its totality by measurement, that is, by thesuccessive synthesis of its parts. For its limits ofthemselves determine its completeness as awhole.]

ANTITHESIS.The world has no beginning, and no limits inspace, but is, in relation both to time and space,infinite.

PROOF.For let it be granted that it has a beginning. Abeginning is an existence which is preceded bya time in which the thing does not exist. On theabove supposition, it follows that there musthave been a time in which the world did notexist, that is, a void time. But in a void time theorigination of a thing is impossible; because nopart of any such time contains a distinctive con-dition of being, in preference to that of non-being (whether the supposed thing originate ofitself, or by means of some other cause). Conse-quently, many series of things may have a be-ginning in the world, but the world itself can-not have a beginning, and is, therefore, in rela-tion to past time, infinite.

As regards the second statement, let us firsttake the opposite for granted—that the world isfinite and limited in space; it follows that itmust exist in a void space, which is not limited.We should therefore meet not only with a rela-

tion of things in space, but also a relation ofthings to space. Now, as the world is an absolu-te whole, out of and beyond which no object ofintuition, and consequently no correlate towhich can be discovered, this relation of theworld to a void space is merely a relation to noobject. But such a relation, and consequentlythe limitation of the world by void space, isnothing. Consequently, the world, as regardsspace, is not limited, that is, it is infinite in re-gard to extension.*

[*Footnote: Space is merely the form of externalintuition (formal intuition), and not a real objectwhich can be externally perceived. Space, priorto all things which determine it (fill or limit it),or, rather, which present an empirical intuitionconformable to it, is, under the title of absolutespace, nothing but the mere possibility of ex-ternal phenomena, in so far as they either existin themselves, or can annex themselves to gi-

ven intuitions. Empirical intuition is thereforenot a composition of phenomena and space (ofperception and empty intuition). The one is notthe correlate of the other in a synthesis, butthey are vitally connected in the same empiricalintuition, as matter and form. If we wish to setone of these two apart from the other—spacefrom phenomena—there arise all sorts of emptydeterminations of external intuition, which arevery far from being possible perceptions. Forexample, motion or rest of the world in an infi-nite empty space, or a determination of the mu-tual relation of both, cannot possibly be percei-ved, and is therefore merely the predicate of anotional entity.]

OBSERVATIONS ON THE FIRST ANTI-NOMY.

ON THE THESIS.

In bringing forward these conflicting argu-ments, I have not been on the search for sop-hisms, for the purpose of availing myself ofspecial pleading, which takes advantage of thecarelessness of the opposite party, appeals to amisunderstood statute, and erects its unrigh-teous claims upon an unfair interpretation.Both proofs originate fairly from the nature ofthe case, and the advantage presented by themistakes of the dogmatists of both parties hasbeen completely set aside.

The thesis might also have been unfairly de-monstrated, by the introduction of an erro-neous conception of the infinity of a givenquantity. A quantity is infinite, if a greater thanitself cannot possibly exist. The quantity is

measured by the number of given units- whichare taken as a standard—contained in it. Nowno number can be the greatest, because one ormore units can always be added. It follows thatan infinite given quantity, consequently an in-finite world (both as regards time and exten-sion) is impossible. It is, therefore, limited inboth respects. In this manner I might have con-ducted my proof; but the conception given in itdoes not agree with the true conception of aninfinite whole. In this there is no representationof its quantity, it is not said how large it is; con-sequently its conception is not the conceptionof a maximum. We cogitate in it merely its rela-tion to an arbitrarily assumed unit, in relationto which it is greater than any number. Now,just as the unit which is taken is greater or sma-ller, the infinite will be greater or smaller; butthe infinity, which consists merely in the rela-tion to this given unit, must remain always thesame, although the absolute quantity of thewhole is not thereby cognized.

The true (transcendental) conception of infinityis: that the successive synthesis of unity in themeasurement of a given quantum can never becompleted.* Hence it follows, without possibili-ty of mistake, that an eternity of actual succes-sive states up to a given (the present) momentcannot have elapsed, and that the world musttherefore have a beginning.

[*Footnote: The quantum in this sense containsa congeries of given units, which is greater thanany number—and this is the mathematical con-ception of the infinite.]

In regard to the second part of the thesis, thedifficulty as to an infinite and yet elapsed seriesdisappears; for the manifold of a world infinitein extension is contemporaneously given. But,in order to cogitate the total of this manifold, aswe cannot have the aid of limits constituting bythemselves this total in intuition, we are obli-ged to give some account of our conception,which in this case cannot proceed from the

whole to the determined quantity of the parts,but must demonstrate the possibility of a wholeby means of a successive synthesis of the parts.But as this synthesis must constitute a seriesthat cannot be completed, it is impossible for usto cogitate prior to it, and consequently not bymeans of it, a totality. For the conception oftotality itself is in the present case the represen-tation of a completed synthesis of the parts; andthis completion, and consequently its concep-tion, is impossible.

ON THE ANTITHESIS.

The proof in favour of the infinity of the cosmi-cal succession and the cosmical content is basedupon the consideration that, in the oppositecase, a void time and a void space must consti-tute the limits of the world. Now I am notunaware, that there are some ways of escaping

this conclusion. It may, for example, be alleged,that a limit to the world, as regards both spaceand time, is quite possible, without at the sametime holding the existence of an absolute timebefore the beginning of the world, or an absolu-te space extending beyond the actual world—which is impossible. I am quite well satisfiedwith the latter part of this opinion of the philo-sophers of the Leibnitzian school. Space is me-rely the form of external intuition, but not a realobject which can itself be externally intuited; itis not a correlate of phenomena, it is the form ofphenomena itself. Space, therefore, cannot beregarded as absolutely and in itself somethingdeterminative of the existence of things, becau-se it is not itself an object, but only the form ofpossible objects. Consequently, things, as phe-nomena, determine space; that is to say, theyrender it possible that, of all the possible predi-cates of space (size and relation), certain maybelong to reality. But we cannot affirm the con-verse, that space, as something self-subsistent,

can determine real things in regard to size orshape, for it is in itself not a real thing. Space(filled or void)* may therefore be limited byphenomena, but phenomena cannot be limitedby an empty space without them. This is true oftime also. All this being granted, it is nevert-heless indisputable, that we must assume thesetwo nonentities, void space without and voidtime before the world, if we assume the exis-tence of cosmical limits, relatively to space ortime.

[*Footnote: It is evident that what is meant hereis, that empty space, in so far as it is limited byphenomena—space, that is, within the world—does not at least contradict transcendental prin-ciples, and may therefore, as regards them, beadmitted, although its possibility cannot onthat account be affirmed.]

For, as regards the subterfuge adopted by thosewho endeavour to evade the consequence—that, if the world is limited as to space and ti-

me, the infinite void must determine the exis-tence of actual things in regard to their dimen-sions—it arises solely from the fact that insteadof a sensuous world, an intelligible world—ofwhich nothing is known—is cogitated; insteadof a real beginning (an existence, which is pre-ceded by a period in which nothing exists), anexistence which presupposes no other condi-tion than that of time; and, instead of limits ofextension, boundaries of the universe. But thequestion relates to the mundus phaenomenon,and its quantity; and in this case we cannotmake abstraction of the conditions of sensibili-ty, without doing away with the essential reali-ty of this world itself. The world of sense, if it islimited, must necessarily lie in the infinite void.If this, and with it space as the a priori condi-tion of the possibility of phenomena, is left outof view, the whole world of sense disappears.In our problem is this alone considered as gi-ven. The mundus intelligibilis is nothing butthe general conception of a world, in which

abstraction has been made of all conditions ofintuition, and in relation to which no syntheti-cal proposition—either affirmative or negati-ve—is possible.

SECOND CONFLICT OF TRANSCEN-DENTAL IDEAS.

THESIS.

Every composite substance in the world con-sists of simple parts; and there exists nothingthat is not either itself simple, or composed ofsimple parts.

PROOF.For, grant that composite substances do notconsist of simple parts; in this case, if all com-

bination or composition were annihilated inthought, no composite part, and (as, by the sup-position, there do not exist simple parts) nosimple part would exist. Consequently, nosubstance; consequently, nothing would exist.Either, then, it is impossible to annihilate com-position in thought; or, after such annihilation,there must remain something that subsists wit-hout composition, that is, something that issimple. But in the former case the compositecould not itself consist of substances, becausewith substances composition is merely a con-tingent relation, apart from which they muststill exist as self-subsistent beings. Now, as thiscase contradicts the supposition, the secondmust contain the truth- that the substantialcomposite in the world consists of simple parts.

It follows, as an immediate inference, that thethings in the world are all, without exception,simple beings—that composition is merely anexternal condition pertaining to them—and

that, although we never can separate and isola-te the elementary substances from the state ofcomposition, reason must cogitate these as theprimary subjects of all composition, and conse-quently, as prior thereto—and as simple subs-tances.

ANTITHESIS.No composite thing in the world consists ofsimple parts; and there does not exist in theworld any simple substance.

PROOF.Let it be supposed that a composite thing (assubstance) consists of simple parts. Inasmuchas all external relation, consequently all compo-sition of substances, is possible only in space;the space, occupied by that which is composite,must consist of the same number of parts as is

contained in the composite. But space does notconsist of simple parts, but of spaces. Therefore,every part of the composite must occupy a spa-ce. But the absolutely primary parts of what iscomposite are simple. It follows that what issimple occupies a space. Now, as everythingreal that occupies a space, contains a manifoldthe parts of which are external to each other,and is consequently composite—and a realcomposite, not of accidents (for these cannotexist external to each other apart from substan-ce), but of substances—it follows that the sim-ple must be a substantial composite, which isself-contradictory.

The second proposition of the antithesis—thatthere exists in the world nothing that is sim-ple—is here equivalent to the following: Theexistence of the absolutely simple cannot bedemonstrated from any experience or percep-tion either external or internal; and the absolu-tely simple is a mere idea, the objective reality

of which cannot be demonstrated in any possi-ble experience; it is consequently, in the exposi-tion of phenomena, without application andobject. For, let us take for granted that an objectmay be found in experience for this transcen-dental idea; the empirical intuition of such anobject must then be recognized to contain abso-lutely no manifold with its parts external toeach other, and connected into unity. Now, aswe cannot reason from the non-consciousnessof such a manifold to the impossibility of itsexistence in the intuition of an object, and as theproof of this impossibility is necessary for theestablishment and proof of absolute simplicity;it follows that this simplicity cannot be inferredfrom any perception whatever. As, therefore,an absolutely simple object cannot be given inany experience, and the world of sense must beconsidered as the sum total of all possible expe-riences: nothing simple exists in the world.

This second proposition in the antithesis has amore extended aim than the first. The first me-rely banishes the simple from the intuition ofthe composite; while the second drives it entire-ly out of nature. Hence we were unable to de-monstrate it from the conception of a givenobject of external intuition (of the composite),but we were obliged to prove it from the rela-tion of a given object to a possible experience ingeneral.

OBSERVATIONS ON THE SECOND AN-TINOMY.

THESIS.

When I speak of a whole, which necessarilyconsists of simple parts, I understand therebyonly a substantial whole, as the true composite;that is to say, I understand that contingent uni-

ty of the manifold which is given as perfectlyisolated (at least in thought), placed in recipro-cal connection, and thus constituted a unity.Space ought not to be called a compositum buta totum, for its parts are possible in the whole,and not the whole by means of the parts. Itmight perhaps be called a compositum ideale,but not a compositum reale. But this is of noimportance. As space is not a composite ofsubstances (and not even of real accidents), if Iabstract all composition therein—nothing, noteven a point, remains; for a point is possibleonly as the limit of a space—consequently of acomposite. Space and time, therefore, do notconsist of simple parts. That which belongsonly to the condition or state of a substance,even although it possesses a quantity (motionor change, for example), likewise does not con-sist of simple parts. That is to say, a certain de-gree of change does not originate from the ad-dition of many simple changes. Our inferenceof the simple from the composite is valid only

of self-subsisting things. But the accidents of astate are not self-subsistent. The proof, then, forthe necessity of the simple, as the componentpart of all that is substantial and composite,may prove a failure, and the whole case of thisthesis be lost, if we carry the proposition toofar, and wish to make it valid of everything thatis composite without distinction—as indeedhas really now and then happened. Besides, Iam here speaking only of the simple, in so faras it is necessarily given in the composite—thelatter being capable of solution into the formeras its component parts. The proper significationof the word monas (as employed by Leibnitz)ought to relate to the simple, given immediate-ly as simple substance (for example, in cons-ciousness), and not as an element of the compo-site. As an clement, the term atomus would bemore appropriate. And as I wish to prove theexistence of simple substances, only in relationto, and as the elements of, the composite, Imight term the antithesis of the second Anti-

nomy, transcendental Atomistic. But as thisword has long been employed to designate aparticular theory of corporeal phenomena (mo-leculae), and thus presupposes a basis of empi-rical conceptions, I prefer calling it the dialecti-cal principle of Monadology.

ANTITHESIS.Against the assertion of the infinite subdivisibi-lity of matter whose ground of proof is purelymathematical, objections have been alleged bythe Monadists. These objections lay themselvesopen, at first sight, to suspicion, from the factthat they do not recognize the clearest mat-hematical proofs as propositions relating to theconstitution of space, in so far as it is really theformal condition of the possibility of all matter,but regard them merely as inferences from abs-tract but arbitrary conceptions, which cannothave any application to real things. Just as if itwere possible to imagine another mode of in-

tuition than that given in the primitive intuitionof space; and just as if its a priori determina-tions did not apply to everything, the existenceof which is possible, from the fact alone of itsfilling space. If we listen to them, we shall findourselves required to cogitate, in addition tothe mathematical point, which is simple—not,however, a part, but a mere limit of space- phy-sical points, which are indeed likewise simple,but possess the peculiar property, as parts ofspace, of filling it merely by their aggregation. Ishall not repeat here the common and clearrefutations of this absurdity, which are to befound everywhere in numbers: every oneknows that it is impossible to undermine theevidence of mathematics by mere discursiveconceptions; I shall only remark that, if in thiscase philosophy endeavours to gain an advan-tage over mathematics by sophistical artifices, itis because it forgets that the discussion relatessolely to Phenomena and their conditions. It isnot sufficient to find the conception of the sim-

ple for the pure conception of the composite,but we must discover for the intuition of thecomposite (matter), the intuition of the simple.Now this, according to the laws of sensibility,and consequently in the case of objects of sense,is utterly impossible. In the case of a wholecomposed of substances, which is cogitatedsolely by the pure understanding, it may benecessary to be in possession of the simple be-fore composition is possible. But this does nothold good of the Totum substantiale phaeno-menon, which, as an empirical intuition in spa-ce, possesses the necessary property of contai-ning no simple part, for the very reason that nopart of space is simple. Meanwhile, the Mona-dists have been subtle enough to escape fromthis difficulty, by presupposing intuition andthe dynamical relation of substances as thecondition of the possibility of space, instead ofregarding space as the condition of the possibi-lity of the objects of external intuition, that is, ofbodies. Now we have a conception of bodies

only as phenomena, and, as such, they necessa-rily presuppose space as the condition of allexternal phenomena. The evasion is thereforein vain; as, indeed, we have sufficiently shownin our Aesthetic. If bodies were things in them-selves, the proof of the Monadists would beunexceptionable.

The second dialectical assertion possesses thepeculiarity of having opposed to it a dogmati-cal proposition, which, among all such sophis-tical statements, is the only one that undertakesto prove in the case of an object of experience,that which is properly a transcendental idea—the absolute simplicity of substance. The pro-position is that the object of the internal sense,the thinking Ego, is an absolute simple subs-tance. Without at present entering upon thissubject—as it has been considered at length in aformer chapter- I shall merely remark that, ifsomething is cogitated merely as an object, wit-hout the addition of any synthetical determina-

tion of its intuition—as happens in the case ofthe bare representation, I—it is certain that nomanifold and no composition can be perceivedin such a representation. As, moreover, thepredicates whereby I cogitate this object aremerely intuitions of the internal sense, therecannot be discovered in them anything to provethe existence of a manifold whose parts areexternal to each other, and, consequently, not-hing to prove the existence of real composition.Consciousness, therefore, is so constituted that,inasmuch as the thinking subject is at the sametime its own object, it cannot divide itself—although it can divide its inhering determina-tions. For every object in relation to itself is ab-solute unity. Nevertheless, if the subject is re-garded externally, as an object of intuition, itmust, in its character of phenomenon, possessthe property of composition. And it must al-ways be regarded in this manner, if we wish toknow whether there is or is not contained in it a

manifold whose parts are external to each ot-her.

THIRD CONFLICT OF THE TRANSCEN-DENTAL IDEAS.

THESIS.

Causality according to the laws of nature, is notthe only causality operating to originate thephenomena of the world. A causality of free-dom is also necessary to account fully for thesephenomena.

PROOF.Let it be supposed, that there is no other kindof causality than that according to the laws ofnature. Consequently, everything that happens

presupposes a previous condition, which itfollows with absolute certainty, in conformitywith a rule. But this previous condition mustitself be something that has happened (that hasarisen in time, as it did not exist before), for, ifit has always been in existence, its consequenceor effect would not thus originate for the firsttime, but would likewise have always existed.The causality, therefore, of a cause, wherebysomething happens, is itself a thing that hashappened. Now this again presupposes, in con-formity with the law of nature, a previous con-dition and its causality, and this another ante-rior to the former, and so on. If, then, everyt-hing happens solely in accordance with thelaws of nature, there cannot be any real firstbeginning of things, but only a subaltern orcomparative beginning. There cannot, therefo-re, be a completeness of series on the side of thecauses which originate the one from the other.But the law of nature is that nothing can hap-pen without a sufficient a priori determined

cause. The proposition therefore—if all causali-ty is possible only in accordance with the lawsof nature—is, when stated in this unlimited andgeneral manner, self-contradictory. It followsthat this cannot be the only kind of causality.

From what has been said, it follows that a cau-sality must be admitted, by means of whichsomething happens, without its cause beingdetermined according to necessary laws bysome other cause preceding. That is to say, the-re must exist an absolute spontaneity of cause,which of itself originates a series of phenomenawhich proceeds according to natural laws—consequently transcendental freedom, withoutwhich even in the course of nature the succes-sion of phenomena on the side of causes is ne-ver complete.

ANTITHESIS.There is no such thing as freedom, but everyt-hing in the world happens solely according tothe laws of nature.

PROOF.Granted, that there does exist freedom in thetranscendental sense, as a peculiar kind of cau-sality, operating to produce events in theworld—a faculty, that is to say, of originating astate, and consequently a series of consequen-ces from that state. In this case, not only theseries originated by this spontaneity, but thedetermination of this spontaneity itself to theproduction of the series, that is to say, the cau-sality itself must have an absolute commence-ment, such that nothing can precede to deter-mine this action according to unvarying laws.But every beginning of action presupposes inthe acting cause a state of inaction; and a dy-

namically primal beginning of action presup-poses a state, which has no connection—as re-gards causality—with the preceding state of thecause—which does not, that is, in any wise re-sult from it. Transcendental freedom is therefo-re opposed to the natural law of cause and ef-fect, and such a conjunction of successive statesin effective causes is destructive of the possibi-lity of unity in experience and for that reasonnot to be found in experience—is consequentlya mere fiction of thought.

We have, therefore, nothing but nature towhich we must look for connection and orderin cosmical events. Freedom—independence ofthe laws of nature—is certainly a deliverancefrom restraint, but it is also a relinquishing ofthe guidance of law and rule. For it cannot bealleged that, instead of the laws of nature, lawsof freedom may be introduced into the causali-ty of the course of nature. For, if freedom weredetermined according to laws, it would be no

longer freedom, but merely nature. Nature,therefore, and transcendental freedom are dis-tinguishable as conformity to law and lawless-ness. The former imposes upon understandingthe difficulty of seeking the origin of eventsever higher and higher in the series of causes,inasmuch as causality is always conditionedthereby; while it compensates this labour by theguarantee of a unity complete and in conformi-ty with law. The latter, on the contrary, holdsout to the understanding the promise of a pointof rest in the chain of causes, by conducting itto an unconditioned causality, which professesto have the power of spontaneous origination,but which, in its own utter blindness, deprivesit of the guidance of rules, by which alone acompletely connected experience is possible.

OBSERVATIONS ON THE THIRD AN-TINOMY.

ON THE THESIS.

The transcendental idea of freedom is far fromconstituting the entire content of the psycholo-gical conception so termed, which is for themost part empirical. It merely presents us withthe conception of spontaneity of action, as theproper ground for imputing freedom to thecause of a certain class of objects. It is, however,the true stumbling-stone to philosophy, whichmeets with unconquerable difficulties in theway of its admitting this kind of unconditionedcausality. That element in the question of thefreedom of the will, which has for so long atime placed speculative reason in such per-plexity, is properly only transcendental, andconcerns the question, whether there must beheld to exist a faculty of spontaneous origina-tion of a series of successive things or states.

How such a faculty is possible is not a necessa-ry inquiry; for in the case of natural causalityitself, we are obliged to content ourselves withthe a priori knowledge that such a causalitymust be presupposed, although we are quiteincapable of comprehending how the being ofone thing is possible through the being of anot-her, but must for this information look entirelyto experience. Now we have demonstrated thisnecessity of a free first beginning of a series ofphenomena, only in so far as it is required forthe comprehension of an origin of the world, allfollowing states being regarded as a successionaccording to laws of nature alone. But, as therehas thus been proved the existence of a facultywhich can of itself originate a series in time—although we are unable to explain how it canexist—we feel ourselves authorized to admit,even in the midst of the natural course ofevents, a beginning, as regards causality, ofdifferent successions of phenomena, and at thesame time to attribute to all substances a facul-

ty of free action. But we ought in this case notto allow ourselves to fall into a common mi-sunderstanding, and to suppose that, because asuccessive series in the world can only have acomparatively first beginning—another state orcondition of things always preceding—an abso-lutely first beginning of a series in the course ofnature is impossible. For we are not speakinghere of an absolutely first beginning in relationto time, but as regards causality alone. When,for example, I, completely of my own free will,and independently of the necessarily determi-native influence of natural causes, rise from mychair, there commences with this event, inclu-ding its material consequences in infinitum, anabsolutely new series; although, in relation totime, this event is merely the continuation of apreceding series. For this resolution and act ofmine do not form part of the succession of ef-fects in nature, and are not mere continuationsof it; on the contrary, the determining causes ofnature cease to operate in reference to this

event, which certainly succeeds the acts of na-ture, but does not proceed from them. For thesereasons, the action of a free agent must be ter-med, in regard to causality, if not in relation totime, an absolutely primal beginning of a seriesof phenomena.

The justification of this need of reason to restupon a free act as the first beginning of the se-ries of natural causes is evident from the fact,that all philosophers of antiquity (with the ex-ception of the Epicurean school) felt themselvesobliged, when constructing a theory of the mo-tions of the universe, to accept a prime mover,that is, a freely acting cause, which sponta-neously and prior to all other causes evolvedthis series of states. They always felt the need ofgoing beyond mere nature, for the purpose ofmaking a first beginning comprehensible.

ON THE ANTITHESIS.

The assertor of the all-sufficiency of nature inregard to causality (transcendental Physiocra-cy), in opposition to the doctrine of freedom,would defend his view of the question somew-hat in the following manner. He would say, inanswer to the sophistical arguments of the op-posite party: If you do not accept a mathemati-cal first, in relation to time, you have no need toseek a dynamical first, in regard to causality.Who compelled you to imagine an absolutelyprimal condition of the world, and therewithan absolute beginning of the gradually progres-sing successions of phenomena—and, as somefoundation for this fancy of yours, to setbounds to unlimited nature? Inasmuch as thesubstances in the world have always existed—at least the unity of experience renders such asupposition quite necessary—there is no diffi-culty in believing also, that the changes in theconditions of these substances have always

existed; and, consequently, that a first begin-ning, mathematical or dynamical, is by nomeans required. The possibility of such an infi-nite derivation, without any initial memberfrom which all the others result, is certainlyquite incomprehensible. But, if you are rashenough to deny the enigmatical secrets of natu-re for this reason, you will find yourselves obli-ged to deny also the existence of many funda-mental properties of natural objects (such asfundamental forces), which you can just as littlecomprehend; and even the possibility of sosimple a conception as that of change must pre-sent to you insuperable difficulties. For if expe-rience did not teach you that it was real, younever could conceive a priori the possibility ofthis ceaseless sequence of being and non-being.

But if the existence of a transcendental facultyof freedom is granted—a faculty of originatingchanges in the world—this faculty must at leastexist out of and apart from the world; although

it is certainly a bold assumption, that, over andabove the complete content of all possible intui-tions, there still exists an object which cannot bepresented in any possible perception. But, toattribute to substances in the world itself such afaculty, is quite inadmissible; for, in this case;the connection of phenomena reciprocally de-termining and determined according to generallaws, which is termed nature, and along with itthe criteria of empirical truth, which enable usto distinguish experience from mere visionarydreaming, would almost entirely disappear. Inproximity with such a lawless faculty of free-dom, a system of nature is hardly cogitable; forthe laws of the latter would be continually sub-ject to the intrusive influences of the former,and the course of phenomena, which wouldotherwise proceed regularly and uniformly,would become thereby confused and discon-nected.

FOURTH CONFLICT OF THE TRANS-CENDENTAL IDEAS.

THESIS.

There exists either in, or in connection with theworld—either as a part of it, or as the cause ofit—an absolutely necessary being.

PROOF.The world of sense, as the sum total of all phe-nomena, contains a series of changes. For, wit-hout such a series, the mental representation ofthe series of time itself, as the condition of thepossibility of the sensuous world, could not bepresented to us.* But every change stands un-der its condition, which precedes it in time andrenders it necessary. Now the existence of agiven condition presupposes a complete seriesof conditions up to the absolutely unconditio-

ned, which alone is absolutely necessary. Itfollows that something that is absolutely neces-sary must exist, if change exists as its conse-quence. But this necessary thing itself belongsto the sensuous world. For suppose it to existout of and apart from it, the series of cosmicalchanges would receive from it a beginning, andyet this necessary cause would not itself belongto the world of sense. But this is impossible.For, as the beginning of a series in time is de-termined only by that which precedes it in ti-me, the supreme condition of the beginning ofa series of changes must exist in the time inwhich this series itself did not exist; for a be-ginning supposes a time preceding, in whichthe thing that begins to be was not in existence.The causality of the necessary cause of changes,and consequently the cause itself, must for the-se reasons belong to time—and to phenomena,time being possible only as the form of pheno-mena. Consequently, it cannot be cogitated asseparated from the world of sense—the sum

total of all phenomena. There is, therefore, con-tained in the world, something that is absolute-ly necessary—whether it be the whole cosmicalseries itself, or only a part of it.

[*Footnote: Objectively, time, as the formalcondition of the possibility of change, precedesall changes; but subjectively, and in conscious-ness, the representation of time, like every ot-her, is given solely by occasion of perception.]

ANTITHESIS.An absolutely necessary being does not exist,either in the world, or out of it—as its cause.

PROOF.Grant that either the world itself is necessary,or that there is contained in it a necessary exis-tence. Two cases are possible. First, there must

either be in the series of cosmical changes abeginning, which is unconditionally necessary,and therefore uncaused- which is at variancewith the dynamical law of the determination ofall phenomena in time; or, secondly, the seriesitself is without beginning, and, although con-tingent and conditioned in all its parts, is ne-vertheless absolutely necessary and uncondi-tioned as a whole—which is self-contradictory.For the existence of an aggregate cannot be ne-cessary, if no single part of it possesses necessa-ry existence.

Grant, on the other band, that an absolutelynecessary cause exists out of and apart from theworld. This cause, as the highest member in theseries of the causes of cosmical changes, mustoriginate or begin* the existence of the latterand their series. In this case it must also beginto act, and its causality would therefore belongto time, and consequently to the sum total ofphenomena, that is, to the world. It follows that

the cause cannot be out of the world; which iscontradictory to the hypothesis. Therefore,neither in the world, nor out of it (but in causalconnection with it), does there exist any absolu-tely necessary being.

[*Footnote: The word begin is taken in two sen-ses. The first is active— the cause being regar-ded as beginning a series of conditions as itseffect (infit). The second is passive—the causali-ty in the cause itself beginning to operate (fit). Ireason here from the first to the second.]

OBSERVATIONS ON THE FOURTH AN-TINOMY.

ON THE THESIS.

To demonstrate the existence of a necessarybeing, I cannot be permitted in this place to

employ any other than the cosmological argu-ment, which ascends from the conditioned inphenomena to the unconditioned in concep-tion—the unconditioned being considered thenecessary condition of the absolute totality ofthe series. The proof, from the mere idea of asupreme being, belongs to another principle ofreason and requires separate discussion.

The pure cosmological proof demonstrates theexistence of a necessary being, but at the sametime leaves it quite unsettled, whether thisbeing is the world itself, or quite distinct fromit. To establish the truth of the latter view, prin-ciples are requisite, which are not cosmologicaland do not proceed in the series of phenomena.We should require to introduce into our proofconceptions of contingent beings—regardedmerely as objects of the understanding, andalso a principle which enables us to connectthese, by means of mere conceptions, with anecessary being. But the proper place for all

such arguments is a transcendent philosophy,which has unhappily not yet been established.

But, if we begin our proof cosmologically, bylaying at the foundation of it the series of phe-nomena, and the regress in it according to em-pirical laws of causality, we are not at liberty tobreak off from this mode of demonstration andto pass over to something which is not itself amember of the series. The condition must betaken in exactly the same signification as therelation of the conditioned to its condition inthe series has been taken, for the series mustconduct us in an unbroken regress to this su-preme condition. But if this relation is sen-suous, and belongs to the possible empiricalemployment of understanding, the supremecondition or cause must close the regressiveseries according to the laws of sensibility andconsequently, must belong to the series of time.It follows that this necessary existence must be

regarded as the highest member of the cosmicalseries.

Certain philosophers have, nevertheless, allo-wed themselves the liberty of making such asaltus (metabasis eis allo gonos). From thechanges in the world they have concluded theirempirical contingency, that is, their dependenceon empirically-determined causes, and theythus admitted an ascending series of empiricalconditions: and in this they are quite right. Butas they could not find in this series any primalbeginning or any highest member, they passedsuddenly from the empirical conception of con-tingency to the pure category, which presentsus with a series—not sensuous, but intellec-tual—whose completeness does certainly restupon the existence of an absolutely necessarycause. Nay, more, this intellectual series is nottied to any sensuous conditions; and is therefo-re free from the condition of time, which requi-res it spontaneously to begin its causality in

time. But such a procedure is perfectly inad-missible, as will be made plain from what fo-llows.

In the pure sense of the categories, that is con-tingent the contradictory opposite of which ispossible. Now we cannot reason from empiricalcontingency to intellectual. The opposite of thatwhich is changed—the opposite of its state—isactual at another time, and is therefore possible.Consequently, it is not the contradictory oppo-site of the former state. To be that, it is necessa-ry that, in the same time in which the precedingstate existed, its opposite could have existed inits place; but such a cognition is not given us inthe mere phenomenon of change. A body thatwas in motion = A, comes into a state of rest =non-A. Now it cannot be concluded from thefact that a state opposite to the state A followsit, that the contradictory opposite of A is possi-ble; and that A is therefore contingent. To pro-ve this, we should require to know that the sta-

te of rest could have existed in the very sametime in which the motion took place. Now weknow nothing more than that the state of restwas actual in the time that followed the state ofmotion; consequently, that it was also possible.But motion at one time, and rest at another ti-me, are not contradictorily opposed to eachother. It follows from what has been said thatthe succession of opposite determinations, thatis, change, does not demonstrate the fact ofcontingency as represented in the conceptionsof the pure understanding; and that it cannot,therefore, conduct us to the fact of the existenceof a necessary being. Change proves merelyempirical contingency, that is to say, that thenew state could not have existed without a cau-se, which belongs to the preceding time. Thiscause—even although it is regarded as absolu-tely necessary—must be presented to us in ti-me, and must belong to the series of phenome-na.

ON THE ANTITHESIS.The difficulties which meet us, in our attemptto rise through the series of phenomena to theexistence of an absolutely necessary supremecause, must not originate from our inability toestablish the truth of our mere conceptions ofthe necessary existence of a thing. That is tosay, our objections not be ontological, but mustbe directed against the causal connection with aseries of phenomena of a condition which isitself unconditioned. In one word, they must becosmological and relate to empirical laws. Wemust show that the regress in the series of cau-ses (in the world of sense) cannot concludewith an empirically unconditioned condition,and that the cosmological argument from thecontingency of the cosmical state—a contingen-cy alleged to arise from change—does not justi-fy us in accepting a first cause, that is, a primeoriginator of the cosmical series.

The reader will observe in this antinomy a veryremarkable contrast. The very same grounds ofproof which established in the thesis the exis-tence of a supreme being, demonstrated in theantithesis—and with equal strictness—the non-existence of such a being. We found, first, that anecessary being exists, because the whole timepast contains the series of all conditions, andwith it, therefore, the unconditioned (the neces-sary); secondly, that there does not exist anynecessary being, for the same reason, that thewhole time past contains the series of all condi-tions—which are themselves, therefore, in theaggregate, conditioned. The cause of this see-ming incongruity is as follows. We attend, inthe first argument, solely to the absolute totali-ty of the series of conditions, the one of whichdetermines the other in time, and thus arrive ata necessary unconditioned. In the second, weconsider, on the contrary, the contingency ofeverything that is determined in the series oftime- for every event is preceded by a time, in

which the condition itself must be determinedas conditioned—and thus everything that isunconditioned or absolutely necessary disap-pears. In both, the mode of proof is quite inaccordance with the common procedure ofhuman reason, which often falls into discordwith itself, from considering an object from twodifferent points of view. Herr von Mairan re-garded the controversy between two celebratedastronomers, which arose from a similar diffi-culty as to the choice of a proper standpoint, asa phenomenon of sufficient importance to wa-rrant a separate treatise on the subject. The oneconcluded: the moon revolves on its own axis,because it constantly presents the same side tothe earth; the other declared that the moon doesnot revolve on its own axis, for the same rea-son. Both conclusions were perfectly correct,according to the point of view from which themotions of the moon were considered.

SECTION III. Of the Interest of Reason inthese Self-contradictions.

We have thus completely before us the dialecti-cal procedure of the cosmological ideas. Nopossible experience can present us with an ob-ject adequate to them in extent. Nay, more, rea-son itself cannot cogitate them as accordingwith the general laws of experience. And yetthey are not arbitrary fictions of thought. Onthe contrary, reason, in its uninterrupted pro-gress in the empirical synthesis, is necessarilyconducted to them, when it endeavours to freefrom all conditions and to comprehend in itsunconditioned totality that which can only bedetermined conditionally in accordance withthe laws of experience. These dialectical propo-sitions are so many attempts to solve four natu-ral and unavoidable problems of reason. Thereare neither more, nor can there be less, than thisnumber, because there are no other series of

synthetical hypotheses, limiting a priori theempirical synthesis.

The brilliant claims of reason striving to extendits dominion beyond the limits of experience,have been represented above only in dry for-mulae, which contain merely the grounds of itspretensions. They have, besides, in conformitywith the character of a transcendental philo-sophy, been freed from every empirical ele-ment; although the full splendour of the promi-ses they hold out, and the anticipations theyexcite, manifests itself only when in connectionwith empirical cognitions. In the application ofthem, however, and in the advancing enlarge-ment of the employment of reason, whilestruggling to rise from the region of experienceand to soar to those sublime ideas, philosophydiscovers a value and a dignity, which, if itcould but make good its assertions, would raiseit far above all other departments of humanknowledge—professing, as it does, to present a

sure foundation for our highest hopes and theultimate aims of all the exertions of reason. Thequestions: whether the world has a beginningand a limit to its extension in space; whetherthere exists anywhere, or perhaps, in my ownthinking Self, an indivisible and indestructibleunity—or whether nothing but what is divisibleand transitory exists; whether I am a free agent,or, like other beings, am bound in the chains ofnature and fate; whether, finally, there is a su-preme cause of the world, or all our thoughtand speculation must end with nature and theorder of external things—are questions for thesolution of which the mathematician wouldwillingly exchange his whole science; for in itthere is no satisfaction for the highest aspira-tions and most ardent desires of humanity.Nay, it may even be said that the true value ofmathematics- that pride of human reason—consists in this: that she guides reason to theknowledge of nature—in her greater as well asin her less manifestations—in her beautiful or-

der and regularity—guides her, moreover, toan insight into the wonderful unity of the mo-ving forces in the operations of nature, far be-yond the expectations of a philosophy buildingonly on experience; and that she thus encoura-ges philosophy to extend the province of rea-son beyond all experience, and at the same timeprovides it with the most excellent materials forsupporting its investigations, in so far as theirnature admits, by adequate and accordant in-tuitions.

Unfortunately for speculation—but perhapsfortunately for the practical interests of huma-nity—reason, in the midst of her highest antici-pations, finds herself hemmed in by a press ofopposite and contradictory conclusions, fromwhich neither her honour nor her safety willpermit her to draw back. Nor can she regardthese conflicting trains of reasoning with indif-ference as mere passages at arms, still less canshe command peace; for in the subject of the

conflict she has a deep interest. There is no ot-her course left open to her than to reflect withherself upon the origin of this disunion in rea-son—whether it may not arise from a mere mi-sunderstanding. After such an inquiry, arro-gant claims would have to be given up on bothsides; but the sovereignty of reason over un-derstanding and sense would be based upon asure foundation.

We shall at present defer this radical inquiryand, in the meantime, consider for a little whatside in the controversy we should most willin-gly take, if we were obliged to become parti-sans at all. As, in this case, we leave out of sightaltogether the logical criterion of truth, andmerely consult our own interest in reference tothe question, these considerations, althoughinadequate to settle the question of right in eit-her party, will enable us to comprehend howthose who have taken part in the struggle,adopt the one view rather than the other—no

special insight into the subject, however,having influenced their choice. They will, at thesame time, explain to us many other things bythe way—for example, the fiery zeal on the oneside and the cold maintenance of their cause onthe other; why the one party has met with thewarmest approbations, and the other has al-ways been repulsed by irreconcilable prejudi-ces.

There is one thing, however, that determinesthe proper point of view, from which alone thispreliminary inquiry can be instituted and ca-rried on with the proper completeness—andthat is the comparison of the principles fromwhich both sides, thesis and antithesis, pro-ceed. My readers would remark in the proposi-tions of the antithesis a complete uniformity inthe mode of thought and a perfect unity ofprinciple. Its principle was that of pure empiri-cism, not only in the explication of the pheno-mena in the world, but also in the solution of

the transcendental ideas, even of that of theuniverse itself. The affirmations of the thesis,on the contrary, were based, in addition to theempirical mode of explanation employed in theseries of phenomena, on intellectual proposi-tions; and its principles were in so far not sim-ple. I shall term the thesis, in view of its essen-tial characteristic, the dogmatism of pure rea-son.

On the side of Dogmatism, or of the thesis, the-refore, in the determination of the cosmologicalideas, we find:

1. A practical interest, which must be very dearto every right-thinking man. That the word hasa beginning—that the nature of my thinkingself is simple, and therefore indestructible—that I am a free agent, and raised above thecompulsion of nature and her laws—and, fina-lly, that the entire order of things, which formthe world, is dependent upon a Supreme Being,from whom the whole receives unity and con-

nection—these are so many foundation-stonesof morality and religion. The antithesis depri-ves us of all these supports—or, at least, seemsso to deprive us.

2. A speculative interest of reason manifestsitself on this side. For, if we take the transcen-dental ideas and employ them in the mannerwhich the thesis directs, we can exhibit comple-tely a priori the entire chain of conditions, andunderstand the derivation of the conditioned—beginning from the unconditioned. This theantithesis does not do; and for this reason doesnot meet with so welcome a reception. For itcan give no answer to our question respectingthe conditions of its synthesis—except such asmust be supplemented by another question,and so on to infinity. According to it, we mustrise from a given beginning to one still higher;every part conducts us to a still smaller one;every event is preceded by another event whichis its cause; and the conditions of existence rest

always upon other and still higher conditions,and find neither end nor basis in some self-subsistent thing as the primal being.

3. This side has also the advantage of populari-ty; and this constitutes no small part of its claimto favour. The common understanding does notfind the least difficulty in the idea of the un-conditioned beginning of all synthesis—accustomed, as it is, rather to follow our conse-quences than to seek for a proper basis for cog-nition. In the conception of an absolute first,moreover—the possibility of which it does notinquire into—it is highly gratified to find afirmly-established point of departure for itsattempts at theory; while in the restless andcontinuous ascent from the conditioned to thecondition, always with one foot in the air, it canfind no satisfaction.

On the side of the antithesis, or Empiricism, inthe determination of the cosmological ideas:

1. We cannot discover any such practical inter-est arising from pure principles of reason asmorality and religion present. On the contrary,pure empiricism seems to empty them of alltheir power and influence. If there does notexist a Supreme Being distinct from theworld—if the world is without beginning, con-sequently without a Creator—if our wills arenot free, and the soul is divisible and subject tocorruption just like matter—the ideas and prin-ciples of morality lose all validity and fall withthe transcendental ideas which constitutedtheir theoretical support.

2. But empiricism, in compensation, holds outto reason, in its speculative interests, certainimportant advantages, far exceeding any thatthe dogmatist can promise us. For, when em-ployed by the empiricist, understanding is al-ways upon its proper ground of investigation—the field of possible experience, the laws ofwhich it can explore, and thus extend its cogni-

tion securely and with clear intelligence wit-hout being stopped by limits in any direction.Here can it and ought it to find and present tointuition its proper object—not only in itself,but in all its relations; or, if it employ concep-tions, upon this ground it can always presentthe corresponding images in clear and unmis-takable intuitions. It is quite unnecessary for itto renounce the guidance of nature, to attachitself to ideas, the objects of which it cannotknow; because, as mere intellectual entities,they cannot be presented in any intuition. Onthe contrary, it is not even permitted to aban-don its proper occupation, under the pretencethat it has been brought to a conclusion (for itnever can be), and to pass into the region ofidealizing reason and transcendent concep-tions, which it is not required to observe andexplore the laws of nature, but merely to thinkand to imagine—secure from being contradic-ted by facts, because they have not been calledas witnesses, but passed by, or perhaps subor-

dinated to the so-called higher interests andconsiderations of pure reason.

Hence the empiricist will never allow himselfto accept any epoch of nature for the first—theabsolutely primal state; he will not believe thatthere can be limits to his outlook into her widedomains, nor pass from the objects of nature,which he can satisfactorily explain by means ofobservation and mathematical thought—whichhe can determine synthetically in intuition, tothose which neither sense nor imagination canever present in concreto; he will not concedethe existence of a faculty in nature, operatingindependently of the laws of nature—a conces-sion which would introduce uncertainty intothe procedure of the understanding, which isguided by necessary laws to the observation ofphenomena; nor, finally, will he permit himselfto seek a cause beyond nature, inasmuch as weknow nothing but it, and from it alone receive

an objective basis for all our conceptions andinstruction in the unvarying laws of things.

In truth, if the empirical philosopher had noother purpose in the establishment of his antit-hesis than to check the presumption of a reasonwhich mistakes its true destination, whichboasts of its insight and its knowledge, justwhere all insight and knowledge cease to exist,and regards that which is valid only in relationto a practical interest, as an advancement of thespeculative interests of the mind (in order,when it is convenient for itself, to break thethread of our physical investigations, and, un-der pretence of extending our cognition, con-nect them with transcendental ideas, by meansof which we really know only that we knownothing)—if, I say, the empiricist rested satis-fied with this benefit, the principle advanced byhim would be a maxim recommending mode-ration in the pretensions of reason and modestyin its affirmations, and at the same time would

direct us to the right mode of extending theprovince of the understanding, by the help ofthe only true teacher, experience. In obedienceto this advice, intellectual hypotheses and faithwould not be called in aid of our practical inte-rests; nor should we introduce them under thepompous titles of science and insight. For spe-culative cognition cannot find an objective basisany other where than in experience; and, whenwe overstep its limits our synthesis, which re-quires ever new cognitions independent of ex-perience, has no substratum of intuition uponwhich to build.

But if—as often happens—empiricism, in rela-tion to ideas, becomes itself dogmatic and bold-ly denies that which is above the sphere of itsphenomenal cognition, it falls itself into theerror of intemperance—an error which is hereall the more reprehensible, as thereby the prac-tical interest of reason receives an irreparableinjury.

And this constitutes the opposition betweenEpicureanism* andPlatonism.

[*Footnote: It is, however, still a matter ofdoubt whether Epicurus ever propounded the-se principles as directions for the objective em-ployment of the understanding. If, indeed, theywere nothing more than maxims for the specu-lative exercise of reason, he gives evidence the-rein a more genuine philosophic spirit than anyof the philosophers of antiquity. That, in theexplanation of phenomena, we must proceed asif the field of inquiry had neither limits in spacenor commencement in time; that we must besatisfied with the teaching of experience in re-ference to the material of which the world isposed; that we must not look for any other mo-de of the origination of events than that whichis determined by the unalterable laws of nature;and finally, that we not employ the hypothesisof a cause distinct from the world to account for

a phenomenon or for the world itself—are prin-ciples for the extension of speculative philo-sophy, and the discovery of the true sources ofthe principles of morals, which, however littleconformed to in the present day, are undoub-tedly correct. At the same time, any one desi-rous of ignoring, in mere speculation, thesedogmatical propositions, need not for that rea-son be accused of denying them.]

Both Epicurus and Plato assert more in theirsystems than they know. The former encoura-ges and advances science—although to the pre-judice of the practical; the latter presents uswith excellent principles for the investigation ofthe practical, but, in relation to everything re-garding which we can attain to speculativecognition, permits reason to append idealisticexplanations of natural phenomena, to thegreat injury of physical investigation.

3. In regard to the third motive for the prelimi-nary choice of a party in this war of assertions,

it seems very extraordinary that empiricismshould be utterly unpopular. We should beinclined to believe that the common unders-tanding would receive it with pleasure—promising as it does to satisfy it without pas-sing the bounds of experience and its connectedorder; while transcendental dogmatism obligesit to rise to conceptions which far surpass theintelligence and ability of the most practisedthinkers. But in this, in truth, is to be found itsreal motive. For the common understandingthus finds itself in a situation where not eventhe most learned can have the advantage of it.If it understands little or nothing about thesetranscendental conceptions, no one can boast ofunderstanding any more; and although it maynot express itself in so scholastically correct amanner as others, it can busy itself with reaso-ning and arguments without end, wanderingamong mere ideas, about which one can alwaysbe very eloquent, because we know nothingabout them; while, in the observation and in-

vestigation of nature, it would be forced to re-main dumb and to confess its utter ignorance.Thus indolence and vanity form of themselvesstrong recommendations of these principles.Besides, although it is a hard thing for a philo-sopher to assume a principle, of which he cangive to himself no reasonable account, and stillmore to employ conceptions, the objective reali-ty of which cannot be established, nothing ismore usual with the common understanding. Itwants something which will allow it to go towork with confidence. The difficulty of evencomprehending a supposition does not disquietit, because—not knowing what comprehendingmeans—it never even thinks of the suppositionit may be adopting as a principle; and regardsas known that with which it has become fami-liar from constant use. And, at last, all specula-tive interests disappear before the practical in-terests which it holds dear; and it fancies that itunderstands and knows what its necessitiesand hopes incite it to assume or to believe.

Thus the empiricism of transcendentally ideali-zing reason is robbed of all popularity; and,however prejudicial it may be to the highestpractical principles, there is no fear that it willever pass the limits of the schools, or acquireany favour or influence in society or with themultitude.

Human reason is by nature architectonic. Thatis to say, it regards all cognitions as parts of apossible system, and hence accepts only suchprinciples as at least do not incapacitate a cog-nition to which we may have attained frombeing placed along with others in a generalsystem. But the propositions of the antithesisare of a character which renders the completionof an edifice of cognitions impossible. Accor-ding to these, beyond one state or epoch of theworld there is always to be found one moreancient; in every part always other parts them-selves divisible; preceding every event another,the origin of which must itself be sought still

higher; and everything in existence is conditio-ned, and still not dependent on an unconditio-ned and primal existence. As, therefore, theantithesis will not concede the existence of afirst beginning which might be available as afoundation, a complete edifice of cognition, inthe presence of such hypothesis, is utterly im-possible. Thus the architectonic interest of rea-son, which requires a unity—not empirical, buta priori and rational—forms a natural recom-mendation for the assertions of the thesis in ourantinomy.

But if any one could free himself entirely fromall considerations of interest, and weigh wit-hout partiality the assertions of reason, atten-ding only to their content, irrespective of theconsequences which follow from them; such aperson, on the supposition that he knew noother way out of the confusion than to settle thetruth of one or other of the conflicting doctri-nes, would live in a state of continual hesita-

tion. Today, he would feel convinced that thehuman will is free; to-morrow, considering theindissoluble chain of nature, he would look onfreedom as a mere illusion and declare natureto be all-in-all. But, if he were called to action,the play of the merely speculative reasonwould disappear like the shapes of a dream,and practical interest would dictate his choiceof principles. But, as it well befits a reflectiveand inquiring being to devote certain periods oftime to the examination of its own reason—todivest itself of all partiality, and frankly tocommunicate its observations for the judge-ment and opinion of others; so no one can beblamed for, much less prevented from, placingboth parties on their trial, with permission toend themselves, free from intimidation, beforeintimidation, before a sworn jury of equal con-dition with themselves—the condition of weakand fallible men.

SECTION IV. Of the necessity imposedupon Pure Reason of presenting a Solution ofits Transcendental Problems.

To avow an ability to solve all problems and toanswer all questions would be a professioncertain to convict any philosopher of extrava-gant boasting and self-conceit, and at once todestroy the confidence that might otherwisehave been reposed in him. There are, however,sciences so constituted that every question ari-sing within their sphere must necessarily becapable of receiving an answer from the know-ledge already possessed, for the answer mustbe received from the same sources whence thequestion arose. In such sciences it is not allo-wable to excuse ourselves on the plea of neces-sary and unavoidable ignorance; a solution isabsolutely requisite. The rule of right andwrong must help us to the knowledge of whatis right or wrong in all possible cases; otherwi-se, the idea of obligation or duty would be ut-

terly null, for we cannot have any obligation tothat which we cannot know. On the other hand,in our investigations of the phenomena of natu-re, much must remain uncertain, and manyquestions continue insoluble; because what weknow of nature is far from being sufficient toexplain all the phenomena that are presented toour observation. Now the question is: Whetherthere is in transcendental philosophy any ques-tion, relating to an object presented to purereason, which is unanswerable by this reason;and whether we must regard the subject of thequestion as quite uncertain, so far as our know-ledge extends, and must give it a place amongthose subjects, of which we have just so muchconception as is sufficient to enable us to raise aquestion—faculty or materials failing us,however, when we attempt an answer.

Now I maintain that, among all speculativecognition, the peculiarity of transcendental phi-losophy is that there is no question, relating to

an object presented to pure reason, which isinsoluble by this reason; and that the professionof unavoidable ignorance- the problem beingalleged to be beyond the reach of our faculties-cannot free us from the obligation to present acomplete and satisfactory answer. For the veryconception which enables us to raise the ques-tion must give us the power of answering it;inasmuch as the object, as in the case of rightand wrong, is not to be discovered out of theconception.

But, in transcendental philosophy, it is only thecosmological questions to which we can de-mand a satisfactory answer in relation to theconstitution of their object; and the philosopheris not permitted to avail himself of the pretextof necessary ignorance and impenetrable obs-curity. These questions relate solely to the cos-mological ideas. For the object must be given inexperience, and the question relates to the ade-quateness of the object to an idea. If the object is

transcendental and therefore itself unknown; ifthe question, for example, is whether the ob-ject—the something, the phenomenon of which(internal—in ourselves) is thought—that is tosay, the soul, is in itself a simple being; orwhether there is a cause of all things, which isabsolutely necessary—in such cases we are see-king for our idea an object, of which we mayconfess that it is unknown to us, though wemust not on that account assert that it is impos-sible.* The cosmological ideas alone posses thepeculiarity that we can presuppose the object ofthem and the empirical synthesis requisite forthe conception of that object to be given; andthe question, which arises from these ideas,relates merely to the progress of this synthesis,in so far as it must contain absolute totality—which, however, is not empirical, as it cannotbe given in any experience. Now, as the ques-tion here is solely in regard to a thing as theobject of a possible experience and not as athing in itself, the answer to the transcendental

cosmological question need not be sought outof the idea, for the question does not regard anobject in itself. The question in relation to apossible experience is not, "What can be givenin an experience in concreto" but "what is con-tained in the idea, to which the empirical synt-hesis must approximate." The question musttherefore be capable of solution from the ideaalone. For the idea is a creation of reason itself,which therefore cannot disclaim the obligationto answer or refer us to the unknown object.

[*Footnote: The question, "What is the constitu-tion of a transcendental object?" is unanswera-ble—we are unable to say what it is; but we canperceive that the question itself is nothing; be-cause it does not relate to any object that can bepresented to us. For this reason, we must con-sider all the questions raised in transcendentalpsychology as answerable and as really answe-red; for they relate to the transcendental subjectof all internal phenomena, which is not itself

phenomenon and consequently not given as anobject, in which, moreover, none of the catego-ries—and it is to them that the question is pro-perly directed—find any conditions of its ap-plication. Here, therefore, is a case where noanswer is the only proper answer. For a ques-tion regarding the constitution of a somethingwhich cannot be cogitated by any determinedpredicate, being completely beyond the sphereof objects and experience, is perfectly null andvoid.]

It is not so extraordinary, as it at first sight ap-pears, that a science should demand and expectsatisfactory answers to all the questions thatmay arise within its own sphere (questionesdomesticae), although, up to a certain time,these answers may not have been discovered.There are, in addition to transcendental philo-sophy, only two pure sciences of reason; theone with a speculative, the other with a practi-cal content—pure mathematics and pure ethics.

Has any one ever heard it alleged that, fromour complete and necessary ignorance of theconditions, it is uncertain what exact relationthe diameter of a circle bears to the circle inrational or irrational numbers? By the formerthe sum cannot be given exactly, by the latteronly approximately; and therefore we decidethat the impossibility of a solution of the ques-tion is evident. Lambert presented us with ademonstration of this. In the general principlesof morals there can be nothing uncertain, forthe propositions are either utterly withoutmeaning, or must originate solely in our ratio-nal conceptions. On the other hand, there mustbe in physical science an infinite number ofconjectures, which can never become certain-ties; because the phenomena of nature are notgiven as objects dependent on our conceptions.The key to the solution of such questions can-not, therefore, be found in our conceptions, orin pure thought, but must lie without us andfor that reason is in many cases not to be disco-

vered; and consequently a satisfactory explana-tion cannot be expected. The questions of trans-cendental analytic, which relate to the deduc-tion of our pure cognition, are not to be regar-ded as of the same kind as those mentionedabove; for we are not at present treating of thecertainty of judgements in relation to the originof our conceptions, but only of that certainty inrelation to objects.

We cannot, therefore, escape the responsibilityof at least a critical solution of the questions ofreason, by complaints of the limited nature ofour faculties, and the seemingly humble con-fession that it is beyond the power of our rea-son to decide, whether the world has existedfrom all eternity or had a beginning—whetherit is infinitely extended, or enclosed within cer-tain limits—whether anything in the world issimple, or whether everything must be capableof infinite divisibility—whether freedom canoriginate phenomena, or whether everything is

absolutely dependent on the laws and order ofnature—and, finally, whether there exists abeing that is completely unconditioned andnecessary, or whether the existence of everyt-hing is conditioned and consequently depen-dent on something external to itself, and there-fore in its own nature contingent. For all thesequestions relate to an object, which can be gi-ven nowhere else than in thought. This object isthe absolutely unconditioned totality of thesynthesis of phenomena. If the conceptions inour minds do not assist us to some certain re-sult in regard to these problems, we must notdefend ourselves on the plea that the objectitself remains hidden from and unknown to us.For no such thing or object can be given—it isnot to be found out of the idea in our minds.We must seek the cause of our failure in ouridea itself, which is an insoluble problem and inregard to which we obstinately assume thatthere exists a real object corresponding andadequate to it. A clear explanation of the dialec-

tic which lies in our conception, will very soonenable us to come to a satisfactory decision inregard to such a question.

The pretext that we are unable to arrive at cer-tainty in regard to these problems may be metwith this question, which requires at least aplain answer: "From what source do the ideasoriginate, the solution of which involves you insuch difficulties? Are you seeking for an expla-nation of certain phenomena; and do you ex-pect these ideas to give you the principles orthe rules of this explanation?" Let it be granted,that all nature was laid open before you; thatnothing was hid from your senses and yourconsciousness. Still, you could not cognize inconcreto the object of your ideas in any expe-rience. For what is demanded is not only thisfull and complete intuition, but also a completesynthesis and the consciousness of its absolutetotality; and this is not possible by means ofany empirical cognition. It follows that your

question—your idea—is by no means necessaryfor the explanation of any phenomenon; andthe idea cannot have been in any sense givenby the object itself. For such an object can neverbe presented to us, because it cannot be givenby any possible experience. Whatever percep-tions you may attain to, you are still surroun-ded by conditions—in space, or in time—andyou cannot discover anything unconditioned;nor can you decide whether this unconditionedis to be placed in an absolute beginning of thesynthesis, or in an absolute totality of the serieswithout beginning. A whole, in the empiricalsignification of the term, is always merely com-parative. The absolute whole of quantity (theuniverse), of division, of derivation, of the con-dition of existence, with the question—whetherit is to be produced by finite or infinite synt-hesis, no possible experience can instruct usconcerning. You will not, for example, be ableto explain the phenomena of a body in the leastdegree better, whether you believe it to consist

of simple, or of composite parts; for a simplephenomenon—and just as little an infinite se-ries of composition—can never be presented toyour perception. Phenomena require and admitof explanation, only in so far as the conditionsof that explanation are given in perception; butthe sum total of that which is given in pheno-mena, considered as an absolute whole, is itselfa perception—and we cannot therefore seek forexplanations of this whole beyond itself, in ot-her perceptions. The explanation of this wholeis the proper object of the transcendental pro-blems of pure reason.

Although, therefore, the solution of these pro-blems is unattainable through experience, wemust not permit ourselves to say that it is un-certain how the object of our inquiries is consti-tuted. For the object is in our own mind andcannot be discovered in experience; and wehave only to take care that our thoughts areconsistent with each other, and to avoid falling

into the amphiboly of regarding our idea as arepresentation of an object empirically given,and therefore to be cognized according to thelaws of experience. A dogmatical solution istherefore not only unsatisfactory but impossi-ble. The critical solution, which may be a per-fectly certain one, does not consider the ques-tion objectively, but proceeds by inquiring intothe basis of the cognition upon which the ques-tion rests.

SECTION V. Sceptical Exposition of theCosmological Problems presented in the fourTranscendental Ideas.

We should be quite willing to desist from thedemand of a dogmatical answer to our ques-tions, if we understood beforehand that, be theanswer what it may, it would only serve to in-crease our ignorance, to throw us from one in-

comprehensibility into another, from one obs-curity into another still greater, and perhapslead us into irreconcilable contradictions. If adogmatical affirmative or negative answer isdemanded, is it at all prudent to set aside theprobable grounds of a solution which lie beforeus and to take into consideration what advan-tage we shall gain, if the answer is to favour theone side or the other? If it happens that in bothcases the answer is mere nonsense, we have inthis an irresistible summons to institute a criti-cal investigation of the question, for the purpo-se of discovering whether it is based on agroundless presupposition and relates to anidea, the falsity of which would be more easilyexposed in its application and consequencesthan in the mere representation of its content.This is the great utility of the sceptical mode oftreating the questions addressed by pure rea-son to itself. By this method we easily rid our-selves of the confusions of dogmatism, andestablish in its place a temperate criticism,

which, as a genuine cathartic, will successfullyremove the presumptuous notions of philosop-hy and their consequence—the vain pretensionto universal science.

If, then, I could understand the nature of a cos-mological idea and perceive, before I enteredon the discussion of the subject at all, that, wha-tever side of the question regarding the uncon-ditioned of the regressive synthesis of pheno-mena it favoured—it must either be too great ortoo small for every conception of the unders-tanding—I would be able to comprehend howthe idea, which relates to an object of experien-ce—an experience which must be adequate toand in accordance with a possible conception ofthe understanding—must be completely voidand without significance, inasmuch as its objectis inadequate, consider it as we may. And thisis actually the case with all cosmological con-ceptions, which, for the reason above mentio-ned, involve reason, so long as it remains atta-

ched to them, in an unavoidable antinomy. Forsuppose:

First, that the world has no beginning—in thiscase it is too large for our conception; for thisconception, which consists in a successive re-gress, cannot overtake the whole eternity thathas elapsed. Grant that it has a beginning, it isthen too small for the conception of the unders-tanding. For, as a beginning presupposes a timepreceding, it cannot be unconditioned; and thelaw of the empirical employment of the unders-tanding imposes the necessity of looking for ahigher condition of time; and the world is, the-refore, evidently too small for this law.

The same is the case with the double answer tothe question regarding the extent, in space, ofthe world. For, if it is infinite and unlimited, itmust be too large for every possible empiricalconception. If it is finite and limited, we have aright to ask: "What determines these limits?"Void space is not a self-subsistent correlate of

things, and cannot be a final condition—andstill less an empirical condition, forming a partof a possible experience. For how can we haveany experience or perception of an absolutevoid? But the absolute totality of the empiricalsynthesis requires that the unconditioned be anempirical conception. Consequently, a finiteworld is too small for our conception.

Secondly, if every phenomenon (matter) in spa-ce consists of an infinite number of parts, theregress of the division is always too great forour conception; and if the division of spacemust cease with some member of the division(the simple), it is too small for the idea of theunconditioned. For the member at which wehave discontinued our division still admits aregress to many more parts contained in theobject.

Thirdly, suppose that every event in the worldhappens in accordance with the laws of nature;the causality of a cause must itself be an event

and necessitates a regress to a still higher cause,and consequently the unceasing prolongationof the series of conditions a parte priori. Opera-tive nature is therefore too large for every con-ception we can form in the synthesis of cosmi-cal events.

If we admit the existence of spontaneously pro-duced events, that is, of free agency, we aredriven, in our search for sufficient reasons, onan unavoidable law of nature and are compe-lled to appeal to the empirical law of causality,and we find that any such totality of connectionin our synthesis is too small for our necessaryempirical conception.

Fourthly, if we assume the existence of an abso-lutely necessary being—whether it be the worldor something in the world, or the cause of theworld—we must place it in a time at an infinitedistance from any given moment; for, otherwi-se, it must be dependent on some other andhigher existence. Such an existence is, in this

case, too large for our empirical conception,and unattainable by the continued regress ofany synthesis.

But if we believe that everything in the world—be it condition or conditioned—is contingent;every given existence is too small for our con-ception. For in this case we are compelled toseek for some other existence upon which theformer depends.

We have said that in all these cases the cosmo-logical idea is either too great or too small forthe empirical regress in a synthesis, and conse-quently for every possible conception of theunderstanding. Why did we not express our-selves in a manner exactly the reverse of thisand, instead of accusing the cosmological ideaof over stepping or of falling short of its trueaim, possible experience, say that, in the firstcase, the empirical conception is always toosmall for the idea, and in the second too great,and thus attach the blame of these contradic-

tions to the empirical regress? The reason isthis. Possible experience can alone give realityto our conceptions; without it a conception ismerely an idea, without truth or relation to anobject. Hence a possible empirical conceptionmust be the standard by which we are to judgewhether an idea is anything more than an ideaand fiction of thought, or whether it relates toan object in the world. If we say of a thing thatin relation to some other thing it is too large ortoo small, the former is considered as existingfor the sake of the latter, and requiring to beadapted to it. Among the trivial subjects of dis-cussion in the old schools of dialectics was thisquestion: "If a ball cannot pass through a hole,shall we say that the ball is too large or the holetoo small?" In this case it is indifferent whatexpression we employ; for we do not knowwhich exists for the sake of the other. On theother hand, we cannot say: "The man is toolong for his coat"; but: "The coat is too short forthe man."

We are thus led to the well-founded suspicionthat the cosmological ideas, and all the conflic-ting sophistical assertions connected with them,are based upon a false and fictitious conceptionof the mode in which the object of these ideas ispresented to us; and this suspicion will proba-bly direct us how to expose the illusion that hasso long led us astray from the truth.

SECTION VI. Transcendental Idealism asthe Key to the Solution of Pure CosmologicalDialectic.

In the transcendental aesthetic we proved thateverything intuited in space and time, all ob-jects of a possible experience, are nothing butphenomena, that is, mere representations; andthat these, as presented to us—as extended bo-dies, or as series of changes—have no self-subsistent existence apart from human thought.

This doctrine I call Transcendental Idealism.*The realist in the transcendental sense regardsthese modifications of our sensibility, thesemere representations, as things subsisting inthemselves.

[*Footnote: I have elsewhere termed this theoryformal idealism, to distinguish it from materialidealism, which doubts or denies the existenceof external things. To avoid ambiguity, it seemsadvisable in many cases to employ this terminstead of that mentioned in the text.]

It would be unjust to accuse us of holding thelong-decried theory of empirical idealism,which, while admitting the reality of space,denies, or at least doubts, the existence of bo-dies extended in it, and thus leaves us withouta sufficient criterion of reality and illusion. Thesupporters of this theory find no difficulty inadmitting the reality of the phenomena of theinternal sense in time; nay, they go the lengthof maintaining that this internal experience is of

itself a sufficient proof of the real existence ofits object as a thing in itself.

Transcendental idealism allows that the objectsof external intuition—as intuited in space, andall changes in time—as represented by the in-ternal sense, are real. For, as space is the formof that intuition which we call external, and,without objects in space, no empirical represen-tation could be given us, we can and ought toregard extended bodies in it as real. The case isthe same with representations in time. But timeand space, with all phenomena therein, are notin themselves things. They are nothing but re-presentations and cannot exist out of and apartfrom the mind. Nay, the sensuous internal in-tuition of the mind (as the object of conscious-ness), the determination of which is represen-ted by the succession of different states in time,is not the real, proper self, as it exists in itself—not the transcendental subject—but only a phe-nomenon, which is presented to the sensibility

of this, to us, unknown being. This internalphenomenon cannot be admitted to be a self-subsisting thing; for its condition is time, andtime cannot be the condition of a thing in itself.But the empirical truth of phenomena in spaceand time is guaranteed beyond the possibilityof doubt, and sufficiently distinguished fromthe illusion of dreams or fancy—although bothhave a proper and thorough connection in anexperience according to empirical laws. Theobjects of experience then are not things inthemselves, but are given only in experience,and have no existence apart from and indepen-dently of experience. That there may be inhabi-tants in the moon, although no one has everobserved them, must certainly be admitted; butthis assertion means only, that we may in thepossible progress of experience discover themat some future time. For that which stands inconnection with a perception according to thelaws of the progress of experience is real. Theyare therefore really existent, if they stand in

empirical connection with my actual or realconsciousness, although they are not in them-selves real, that is, apart from the progress ofexperience.

There is nothing actually given—we can beconscious of nothing as real, except a percep-tion and the empirical progression from it toother possible perceptions. For phenomena, asmere representations, are real only in percep-tion; and perception is, in fact, nothing but thereality of an empirical representation, that is, aphenomenon. To call a phenomenon a realthing prior to perception means either that wemust meet with this phenomenon in the pro-gress of experience, or it means nothing at all.For I can say only of a thing in itself that itexists without relation to the senses and expe-rience. But we are speaking here merely ofphenomena in space and time, both of whichare determinations of sensibility, and not ofthings in themselves. It follows that phenome-

na are not things in themselves, but are mererepresentations, which if not given in us—inperception—are non-existent.

The faculty of sensuous intuition is properly areceptivity—a capacity of being affected in acertain manner by representations, the relationof which to each other is a pure intuition ofspace and time—the pure forms of sensibility.These representations, in so far as they are con-nected and determinable in this relation (inspace and time) according to laws of the unityof experience, are called objects. The non-sensuous cause of these representations iscompletely unknown to us and hence cannot beintuited as an object. For such an object couldnot be represented either in space or in time;and without these conditions intuition or repre-sentation is impossible. We may, at the sametime, term the non-sensuous cause of pheno-mena the transcendental object—but merely asa mental correlate to sensibility, considered as a

receptivity. To this transcendental object wemay attribute the whole connection and extentof our possible perceptions, and say that it isgiven and exists in itself prior to all experience.But the phenomena, corresponding to it, are notgiven as things in themselves, but in experiencealone. For they are mere representations, recei-ving from perceptions alone significance andrelation to a real object, under the conditionthat this or that perception—indicating an ob-ject—is in complete connection with all othersin accordance with the rules of the unity of ex-perience. Thus we can say: "The things thatreally existed in past time are given in thetranscendental object of experience." But theseare to me real objects, only in so far as I canrepresent to my own mind, that a regressiveseries of possible perceptions- following theindications of history, or the footsteps of causeand effect—in accordance with empiricallaws—that, in one word, the course of theworld conducts us to an elapsed series of time

as the condition of the present time. This seriesin past time is represented as real, not in itself,but only in connection with a possible expe-rience. Thus, when I say that certain eventsoccurred in past time, I merely assert the possi-bility of prolonging the chain of experience,from the present perception, upwards to theconditions that determine it according to time.

If I represent to myself all objects existing in allspace and time, I do not thereby place these inspace and time prior to all experience; on thecontrary, such a representation is nothing morethan the notion of a possible experience, in itsabsolute completeness. In experience alone arethose objects, which are nothing but representa-tions, given. But, when I say they existed priorto my experience, this means only that I mustbegin with the perception present to me andfollow the track indicated until I discover themin some part or region of experience. The causeof the empirical condition of this progression—

and consequently at what member therein Imust stop, and at what point in the regress I amto find this member—is transcendental, andhence necessarily incognizable. But with thiswe have not to do; our concern is only with thelaw of progression in experience, in which ob-jects, that is, phenomena, are given. It is a mat-ter of indifference, whether I say, "I may in theprogress of experience discover stars, at a hun-dred times greater distance than the most dis-tant of those now visible," or, "Stars at this dis-tance may be met in space, although no onehas, or ever will discover them." For, if they aregiven as things in themselves, without any rela-tion to possible experience, they are for menon-existent, consequently, are not objects, forthey are not contained in the regressive seriesof experience. But, if these phenomena must beemployed in the construction or support of thecosmological idea of an absolute whole, andwhen we are discussing a question that overs-teps the limits of possible experience, the pro-

per distinction of the different theories of thereality of sensuous objects is of great importan-ce, in order to avoid the illusion which mustnecessarily arise from the misinterpretation ofour empirical conceptions.

SECTION VII. Critical Solution of theCosmological Problem.

The antinomy of pure reason is based upon thefollowing dialectical argument: "If that which isconditioned is given, the whole series of itsconditions is also given; but sensuous objectsare given as conditioned; consequently…" Thissyllogism, the major of which seems so naturaland evident, introduces as many cosmologicalideas as there are different kinds of conditionsin the synthesis of phenomena, in so far as the-se conditions constitute a series. These ideasrequire absolute totality in the series, and thus

place reason in inextricable embarrassment.Before proceeding to expose the fallacy in thisdialectical argument, it will be necessary tohave a correct understanding of certain concep-tions that appear in it.

In the first place, the following proposition isevident, and indubitably certain: "If the condi-tioned is given, a regress in the series of all itsconditions is thereby imperatively required."For the very conception of a conditioned is aconception of something related to a condition,and, if this condition is itself conditioned, toanother condition—and so on through all themembers of the series. This proposition is, the-refore, analytical and has nothing to fear fromtranscendental criticism. It is a logical postulateof reason: to pursue, as far as possible, the con-nection of a conception with its conditions.

If, in the second place, both the conditionedand the condition are things in themselves, andif the former is given, not only is the regress to

the latter requisite, but the latter is really givenwith the former. Now, as this is true of all themembers of the series, the entire series of con-ditions, and with them the unconditioned, is atthe same time given in the very fact of the con-ditioned, the existence of which is possible onlyin and through that series, being given. In thiscase, the synthesis of the conditioned with itscondition, is a synthesis of the understandingmerely, which represents things as they are,without regarding whether and how we cancognize them. But if I have to do with pheno-mena, which, in their character of mere repre-sentations, are not given, if I do not attain to acognition of them (in other words, to themsel-ves, for they are nothing more than empiricalcognitions), I am not entitled to say: "If theconditioned is given, all its conditions (as phe-nomena) are also given." I cannot, therefore,from the fact of a conditioned being given, inferthe absolute totality of the series of its condi-tions. For phenomena are nothing but an empi-

rical synthesis in apprehension or perception,and are therefore given only in it. Now, inspeaking of phenomena it does not follow that,if the conditioned is given, the synthesis whichconstitutes its empirical condition is also there-by given and presupposed; such a synthesiscan be established only by an actual regress inthe series of conditions. But we are entitled tosay in this case that a regress to the conditionsof a conditioned, in other words, that a conti-nuous empirical synthesis is enjoined; that, ifthe conditions are not given, they are at leastrequired; and that we are certain to discoverthe conditions in this regress.

We can now see that the major, in the abovecosmological syllogism, takes the conditionedin the transcendental signification which it hasin the pure category, while the minor speaks ofit in the empirical signification which it has inthe category as applied to phenomena. There is,therefore, a dialectical fallacy in the syllogism—

a sophisma figurae dictionis. But this fallacy isnot a consciously devised one, but a perfectlynatural illusion of the common reason of man.For, when a thing is given as conditioned, wepresuppose in the major its conditions and theirseries, unperceived, as it were, and unseen;because this is nothing more than the logicalrequirement of complete and satisfactory pre-misses for a given conclusion. In this case, timeis altogether left out in the connection of theconditioned with the condition; they are sup-posed to be given in themselves, and contem-poraneously. It is, moreover, just as natural toregard phenomena (in the minor) as things inthemselves and as objects presented to the pureunderstanding, as in the major, in which com-plete abstraction was made of all conditions ofintuition. But it is under these conditions alonethat objects are given. Now we overlooked aremarkable distinction between the concep-tions. The synthesis of the conditioned with itscondition, and the complete series of the latter

(in the major) are not limited by time, and donot contain the conception of succession. Onthe contrary, the empirical synthesis and theseries of conditions in the phenomenal world—subsumed in the minor—are necessarily suc-cessive and given in time alone. It follows that Icannot presuppose in the minor, as I did in themajor, the absolute totality of the synthesis andof the series therein represented; for in the ma-jor all the members of the series are given asthings in themselves—without any limitationsor conditions of time, while in the minor theyare possible only in and through a successiveregress, which cannot exist, except it be actuallycarried into execution in the world of pheno-mena.

After this proof of the viciousness of the argu-ment commonly employed in maintaining cos-mological assertions, both parties may now bejustly dismissed, as advancing claims withoutgrounds or title. But the process has not been

ended by convincing them that one or bothwere in the wrong and had maintained an as-sertion which was without valid grounds ofproof. Nothing seems to be clearer than that, ifone maintains: "The world has a beginning,"and another: "The world has no beginning," oneof the two must be right. But it is likewise clearthat, if the evidence on both sides is equal, it isimpossible to discover on what side the truthlies; and the controversy continues, althoughthe parties have been recommended to peacebefore the tribunal of reason. There remains,then, no other means of settling the questionthan to convince the parties, who refute eachother with such conclusiveness and ability, thatthey are disputing about nothing, and that atranscendental illusion has been mocking themwith visions of reality where there is none. Themode of adjusting a dispute which cannot bedecided upon its own merits, we shall nowproceed to lay before our readers.

Zeno of Elea, a subtle dialectician, was severelyreprimanded by Plato as a sophist, who, merelyfrom the base motive of exhibiting his skill indiscussion, maintained and subverted the sameproposition by arguments as powerful and con-vincing on the one side as on the other. Hemaintained, for example, that God (who wasprobably nothing more, in his view, than theworld) is neither finite nor infinite, neither inmotion nor in rest, neither similar nor dissimi-lar to any other thing. It seemed to those philo-sophers who criticized his mode of discussionthat his purpose was to deny completely bothof two self-contradictory propositions—whichis absurd. But I cannot believe that there is anyjustice in this accusation. The first of these pro-positions I shall presently consider in a moredetailed manner. With regard to the others, ifby the word of God he understood merely theUniverse, his meaning must have been—that itcannot be permanently present in one place—that is, at rest—nor be capable of changing its

place—that is, of moving- because all places arein the universe, and the universe itself is, there-fore, in no place. Again, if the universe containsin itself everything that exists, it cannot be simi-lar or dissimilar to any other thing, becausethere is, in fact, no other thing with which it canbe compared. If two opposite judgements pre-suppose a contingent impossible, or arbitrarycondition, both—in spite of their opposition(which is, however, not properly or really acontradiction)—fall away; because the condi-tion, which ensured the validity of both, hasitself disappeared.

If we say: "Everybody has either a good or abad smell," we have omitted a third possiblejudgement—it has no smell at all; and thus bothconflicting statements may be false. If we say:"It is either good-smelling or not good-smelling(vel suaveolens vel non-suaveolens)," both jud-gements are contradictorily opposed; and thecontradictory opposite of the former judge-

ment—some bodies are not good-smelling—embraces also those bodies which have nosmell at all. In the preceding pair of opposedjudgements (per disparata), the contingentcondition of the conception of body (smell) at-tached to both conflicting statements, instead ofhaving been omitted in the latter, which is con-sequently not the contradictory opposite of theformer.

If, accordingly, we say: "The world is eitherinfinite in extension, or it is not infinite (non estinfinitus)"; and if the former proposition is fal-se, its contradictory opposite—the world is notinfinite—must be true. And thus I should denythe existence of an infinite, without, howeveraffirming the existence of a finite world. But ifwe construct our proposition thus: "The worldis either infinite or finite (non-infinite)," bothstatements may be false. For, in this case, weconsider the world as per se determined in re-gard to quantity, and while, in the one judge-

ment, we deny its infinite and consequently,perhaps, its independent existence; in the other,we append to the world, regarded as a thing initself, a certain determination—that of finitude;and the latter may be false as well as the for-mer, if the world is not given as a thing in itself,and thus neither as finite nor as infinite inquantity. This kind of opposition I may beallowed to term dialectical; that of contradicto-ries may be called analytical opposition. Thusthen, of two dialectically opposed judgementsboth may be false, from the fact, that the one isnot a mere contradictory of the other, but actua-lly enounces more than is requisite for a fulland complete contradiction.

When we regard the two propositions—"Theworld is infinite in quantity," and, "The world isfinite in quantity," as contradictory opposites,we are assuming that the world—the completeseries of phenomena—is a thing in itself. For itremains as a permanent quantity, whether I

deny the infinite or the finite regress in the se-ries of its phenomena. But if we dismiss thisassumption—this transcendental illusion—anddeny that it is a thing in itself, the contradictoryopposition is metamorphosed into a merelydialectical one; and the world, as not existing initself—independently of the regressive series ofmy representations—exists in like manner neit-her as a whole which is infinite nor as a wholewhich is finite in itself. The universe exists forme only in the empirical regress of the series ofphenomena and not per se. If, then, it is alwaysconditioned, it is never completely or as a who-le; and it is, therefore, not an unconditionedwhole and does not exist as such, either with aninfinite, or with a finite quantity.

What we have here said of the first cosmologi-cal idea—that of the absolute totality of quanti-ty in phenomena—applies also to the others.The series of conditions is discoverable only inthe regressive synthesis itself, and not in the

phenomenon considered as a thing in itself—given prior to all regress. Hence I am compe-lled to say: "The aggregate of parts in a givenphenomenon is in itself neither finite nor infini-te; and these parts are given only in the regres-sive synthesis of decomposition—a synthesiswhich is never given in absolute completeness,either as finite, or as infinite." The same is thecase with the series of subordinated causes, orof the conditioned up to the unconditioned andnecessary existence, which can never be regar-ded as in itself, ind in its totality, either as finiteor as infinite; because, as a series of subordinaterepresentations, it subsists only in the dynami-cal regress and cannot be regarded as existingpreviously to this regress, or as a self-subsistentseries of things.

Thus the antinomy of pure reason in its cosmo-logical ideas disappears. For the above demons-tration has established the fact that it is merelythe product of a dialectical and illusory opposi-

tion, which arises from the application of theidea of absolute totality—admissible only as acondition of things in themselves—to pheno-mena, which exist only in our representations,and—when constituting a series—in a successi-ve regress. This antinomy of reason may, howe-ver, be really profitable to our speculative inte-rests, not in the way of contributing any dog-matical addition, but as presenting to us anot-her material support in our critical investiga-tions. For it furnishes us with an indirect proofof the transcendental ideality of phenomena, ifour minds were not completely satisfied withthe direct proof set forth in the TrancendentalAesthetic. The proof would proceed in the fo-llowing dilemma. If the world is a whole exis-ting in itself, it must be either finite or infinite.But it is neither finite nor infinite—as has beenshown, on the one side, by the thesis, on theother, by the antithesis. Therefore the world—the content of all phenomena—is not a wholeexisting in itself. It follows that phenomena are

nothing, apart from our representations. Andthis is what we mean by transcendental ideali-ty.

This remark is of some importance. It enablesus to see that the proofs of the fourfold antino-my are not mere sophistries—are not fallacious,but grounded on the nature of reason, and va-lid—under the supposition that phenomena arethings in themselves. The opposition of thejudgements which follow makes it evident thata fallacy lay in the initial supposition, and thushelps us to discover the true constitution ofobjects of sense. This transcendental dialecticdoes not favour scepticism, although it presentsus with a triumphant demonstration of the ad-vantages of the sceptical method, the great uti-lity of which is apparent in the antinomy, whe-re the arguments of reason were allowed toconfront each other in undiminished force. Andalthough the result of these conflicts of reasonis not what we expected—although we have

obtained no positive dogmatical addition tometaphysical science—we have still reaped agreat advantage in the correction of our judge-ments on these subjects of thought.

SECTION VIII. Regulative Principle of Pu-re Reason in relation to the CosmologicalIdeas.

The cosmological principle of totality could notgive us any certain knowledge in regard to themaximum in the series of conditions in theworld of sense, considered as a thing in itself.The actual regress in the series is the onlymeans of approaching this maximum. Thisprinciple of pure reason, therefore, may still beconsidered as valid—not as an axiom enablingus to cogitate totality in the object as actual, butas a problem for the understanding, which re-quires it to institute and to continue, in confor-

mity with the idea of totality in the mind, theregress in the series of the conditions of a givenconditioned. For in the world of sense, that is,in space and time, every condition which wediscover in our investigation of phenomena isitself conditioned; because sensuous objects arenot things in themselves (in which case an ab-solutely unconditioned might be reached in theprogress of cognition), but are merely empiricalrepresentations the conditions of which mustalways be found in intuition. The principle ofreason is therefore properly a mere rule—prescribing a regress in the series of conditionsfor given phenomena, and prohibiting anypause or rest on an absolutely unconditioned. Itis, therefore, not a principle of the possibility ofexperience or of the empirical cognition of sen-suous objects—consequently not a principle ofthe understanding; for every experience is con-fined within certain proper limits determinedby the given intuition. Still less is it a constitu-tive principle of reason authorizing us to ex-

tend our conception of the sensuous world be-yond all possible experience. It is merely aprinciple for the enlargement and extension ofexperience as far as is possible for human facul-ties. It forbids us to consider any empirical li-mits as absolute. It is, hence, a principle of rea-son, which, as a rule, dictates how we ought toproceed in our empirical regress, but is unableto anticipate or indicate prior to the empiricalregress what is given in the object itself. I havetermed it for this reason a regulative principleof reason; while the principle of the absolutetotality of the series of conditions, as existing initself and given in the object, is a constitutivecosmological principle. This distinction will atonce demonstrate the falsehood of the constitu-tive principle, and prevent us from attributing(by a transcendental subreptio) objective realityto an idea, which is valid only as a rule.

In order to understand the proper meaning ofthis rule of pure reason, we must notice first

that it cannot tell us what the object is, but onlyhow the empirical regress is to be proceededwith in order to attain to the complete concep-tion of the object. If it gave us any informationin respect to the former statement, it would be aconstitutive principle—a principle impossiblefrom the nature of pure reason. It will not the-refore enable us to establish any such conclu-sions as: "The series of conditions for a givenconditioned is in itself finite," or, "It is infinite."For, in this case, we should be cogitating in themere idea of absolute totality, an object whichis not and cannot be given in experience; inas-much as we should be attributing a reality ob-jective and independent of the empirical synt-hesis, to a series of phenomena. This idea ofreason cannot then be regarded as valid—except as a rule for the regressive synthesis inthe series of conditions, according to which wemust proceed from the conditioned, through allintermediate and subordinate conditions, up tothe unconditioned; although this goal is unat-

tained and unattainable. For the absolutely un-conditioned cannot be discovered in the sphereof experience.

We now proceed to determine clearly our no-tion of a synthesis which can never be comple-te. There are two terms commonly employedfor this purpose. These terms are regarded asexpressions of different and distinguishablenotions, although the ground of the distinctionhas never been clearly exposed. The term em-ployed by the mathematicians is progressus ininfinitum. The philosophers prefer the expres-sion progressus in indefinitum. Without detai-ning the reader with an examination of the rea-sons for such a distinction, or with remarks onthe right or wrong use of the terms, I shall en-deavour clearly to determine these conceptions,so far as is necessary for the purpose in thisCritique.

We may, with propriety, say of a straight line,that it may be produced to infinity. In this case

the distinction between a progressus in infini-tum and a progressus in indefinitum is a merepiece of subtlety. For, although when we say,"Produce a straight line," it is more correct tosay in indefinitum than in infinitum; becausethe former means, "Produce it as far as youplease," the second, "You must not cease toproduce it"; the expression in infinitum is,when we are speaking of the power to do it,perfectly correct, for we can always make itlonger if we please—on to infinity. And thisremark holds good in all cases, when we speakof a progressus, that is, an advancement fromthe condition to the conditioned; this possibleadvancement always proceeds to infinity. Wemay proceed from a given pair in the descen-ding line of generation from father to son, andcogitate a never-ending line of descendantsfrom it. For in such a case reason does not de-mand absolute totality in the series, because itdoes not presuppose it as a condition and as

given (datum), but merely as conditioned, andas capable of being given (dabile).

Very different is the case with the problem:"How far the regress, which ascends from thegiven conditioned to the conditions, must ex-tend"; whether I can say: "It is a regress in infi-nitum," or only "in indefinitum"; and whether,for example, setting out from the human beingsat present alive in the world, I may ascend inthe series of their ancestors, in infinitum—mrwhether all that can be said is, that so far as Ihave proceeded, I have discovered no empiricalground for considering the series limited, sothat I am justified, and indeed, compelled tosearch for ancestors still further back, althoughI am not obliged by the idea of reason to pre-suppose them.

My answer to this question is: "If the series isgiven in empirical intuition as a whole, the re-gress in the series of its internal conditions pro-ceeds in infinitum; but, if only one member of

the series is given, from which the regress is toproceed to absolute totality, the regress is pos-sible only in indefinitum." For example, thedivision of a portion of matter given withincertain limits—of a body, that is—proceeds ininfinitum. For, as the condition of this whole isits part, and the condition of the part a part ofthe part, and so on, and as in this regress ofdecomposition an unconditioned indivisiblemember of the series of conditions is not to befound; there are no reasons or grounds in expe-rience for stopping in the division, but, on thecontrary, the more remote members of the divi-sion are actually and empirically given prior tothis division. That is to say, the division pro-ceeds to infinity. On the other hand, the seriesof ancestors of any given human being is notgiven, in its absolute totality, in any experience,and yet the regress proceeds from every genea-logical member of this series to one still higher,and does not meet with any empirical limitpresenting an absolutely unconditioned mem-

ber of the series. But as the members of such aseries are not contained in the empirical intui-tion of the whole, prior to the regress, this re-gress does not proceed to infinity, but only inindefinitum, that is, we are called upon to dis-cover other and higher members, which arethemselves always conditioned.

In neither case—the regressus in infinitum, northe regressus in indefinitum, is the series ofconditions to be considered as actually infinitein the object itself. This might be true of thingsin themselves, but it cannot be asserted of phe-nomena, which, as conditions of each other, areonly given in the empirical regress itself. Hen-ce, the question no longer is, "What is the quan-tity of this series of conditions in itself—is itfinite or infinite?" for it is nothing in itself; but,"How is the empirical regress to be commen-ced, and how far ought we to proceed with it?"And here a signal distinction in the applicationof this rule becomes apparent. If the whole is

given empirically, it is possible to recede in theseries of its internal conditions to infinity. But ifthe whole is not given, and can only be givenby and through the empirical regress, I can on-ly say: "It is possible to infinity, to proceed tostill higher conditions in the series." In the firstcase, I am justified in asserting that more mem-bers are empirically given in the object than Iattain to in the regress (of decomposition). Inthe second case, I am justified only in saying,that I can always proceed further in the regress,because no member of the series is given asabsolutely conditioned, and thus a highermember is possible, and an inquiry with regardto it is necessary. In the one case it is necessaryto find other members of the series, in the otherit is necessary to inquire for others, inasmuchas experience presents no absolute limitation ofthe regress. For, either you do not possess aperception which absolutely limits your empi-rical regress, and in this case the regress cannotbe regarded as complete; or, you do possess

such a limitative perception, in which case it isnot a part of your series (for that which limitsmust be distinct from that which is limited byit), and it is incumbent you to continue yourregress up to this condition, and so on.

These remarks will be placed in their properlight by their application in the following sec-tion.

SECTION IX. Of the Empirical Use of theRegulative Principle of Reason with regard tothe Cosmological Ideas.

We have shown that no transcendental use canbe made either of the conceptions of reason orof understanding. We have shown, likewise,that the demand of absolute totality in the se-ries of conditions in the world of sense arisesfrom a transcendental employment of reason,

resting on the opinion that phenomena are tobe regarded as things in themselves. It followsthat we are not required to answer the questionrespecting the absolute quantity of a series—whether it is in itself limited or unlimited. Weare only called upon to determine how far wemust proceed in the empirical regress fromcondition to condition, in order to discover, inconformity with the rule of reason, a full andcorrect answer to the questions proposed byreason itself.

This principle of reason is hence valid only as arule for the extension of a possible experience—its invalidity as a principle constitutive of phe-nomena in themselves having been sufficientlydemonstrated. And thus, too, the antinomialconflict of reason with itself is completely putan end to; inasmuch as we have not only pre-sented a critical solution of the fallacy lurkingin the opposite statements of reason, but haveshown the true meaning of the ideas which

gave rise to these statements. The dialecticalprinciple of reason has, therefore, been changedinto a doctrinal principle. But in fact, if thisprinciple, in the subjective signification whichwe have shown to be its only true sense, maybe guaranteed as a principle of the unceasingextension of the employment of our understan-ding, its influence and value are just as great asif it were an axiom for the a priori determina-tion of objects. For such an axiom could notexert a stronger influence on the extension andrectification of our knowledge, otherwise thanby procuring for the principles of the unders-tanding the most widely expanded employ-ment in the field of experience.

I. Solution of the Cosmological Idea of the Tota-lity of the Composition of Phenomena in the Universe.

Here, as well as in the case of the other cosmo-logical problems, the ground of the regulativeprinciple of reason is the proposition that in ourempirical regress no experience of an absolutelimit, and consequently no experience of a con-dition, which is itself absolutely unconditioned,is discoverable. And the truth of this proposi-tion itself rests upon the consideration thatsuch an experience must represent to us phe-nomena as limited by nothing or the mere void,on which our continued regress by means ofperception must abut—which is impossible.

Now this proposition, which declares that eve-ry condition attained in the empirical regressmust itself be considered empirically conditio-ned, contains the rule in terminis, which requi-res me, to whatever extent I may have procee-ded in the ascending series, always to look forsome higher member in the series—whetherthis member is to become known to methrough experience, or not.

Nothing further is necessary, then, for the solu-tion of the first cosmological problem, than todecide, whether, in the regress to the uncondi-tioned quantity of the universe (as regards spa-ce and time), this never limited ascent ought tobe called a regressus in infinitum or indefini-tum.

The general representation which we form inour minds of the series of all past states or con-ditions of the world, or of all the things whichat present exist in it, is itself nothing more thana possible empirical regress, which is cogita-ted—although in an undetermined manner—inthe mind, and which gives rise to the concep-tion of a series of conditions for a given object.*Now I have a conception of the universe, butnot an intuition—that is, not an intuition of it asa whole. Thus I cannot infer the magnitude ofthe regress from the quantity or magnitude ofthe world, and determine the former by meansof the latter; on the contrary, I must first of all

form a conception of the quantity or magnitudeof the world from the magnitude of the empiri-cal regress. But of this regress I know nothingmore than that I ought to proceed from everygiven member of the series of conditions to onestill higher. But the quantity of the universe isnot thereby determined, and we cannot affirmthat this regress proceeds in infinitum. Such anaffirmation would anticipate the members ofthe series which have not yet been reached, andrepresent the number of them as beyond thegrasp of any empirical synthesis; it would con-sequently determine the cosmical quantityprior to the regress (although only in a negativemanner)—which is impossible. For the world isnot given in its totality in any intuition: conse-quently, its quantity cannot be given prior tothe regress. It follows that we are unable tomake any declaration respecting the cosmicalquantity in itself—not even that the regress in itis a regress in infinitum; we must only endea-vour to attain to a conception of the quantity of

the universe, in conformity with the rule whichdetermines the empirical regress in it. But thisrule merely requires us never to admit an abso-lute limit to our series—how far soever we mayhave proceeded in it, but always, on the contra-ry, to subordinate every phenomenon to someother as its condition, and consequently to pro-ceed to this higher phenomenon. Such a regressis, therefore, the regressus in indefinitum,which, as not determining a quantity in theobject, is clearly distinguishable from the re-gressus in infinitum.

[*Footnote: The cosmical series can neither begreater nor smaller than the possible empiricalregress, upon which its conception is based.And as this regress cannot be a determinateinfinite regress, still less a determinate finite(absolutely limited), it is evident that we cannotregard the world as either finite or infinite, be-cause the regress, which gives us the represen-

tation of the world, is neither finite nor infini-te.]

It follows from what we have said that we arenot justified in declaring the world to be infini-te in space, or as regards past time. For thisconception of an infinite given quantity is em-pirical; but we cannot apply the conception ofan infinite quantity to the world as an object ofthe senses. I cannot say, "The regress from agiven perception to everything limited either inspace or time, proceeds in infinitum," for thispresupposes an infinite cosmical quantity; neit-her can I say, "It is finite," for an absolute limitis likewise impossible in experience. It followsthat I am not entitled to make any assertion atall respecting the whole object of experience—the world of sense; I must limit my declarationsto the rule according to which experience orempirical knowledge is to be attained.

To the question, therefore, respecting the cos-mical quantity, the first and negative answer is:

"The world has no beginning in time, and noabsolute limit in space."

For, in the contrary case, it would be limited bya void time on the one hand, and by a void spa-ce on the other. Now, since the world, as a phe-nomenon, cannot be thus limited in itself for aphenomenon is not a thing in itself; it must bepossible for us to have a perception of this limi-tation by a void time and a void space. But sucha perception—such an experience is impossible;because it has no content. Consequently, anabsolute cosmical limit is empirically, and the-refore absolutely, impossible.*

[*Footnote: The reader will remark that theproof presented above is very different fromthe dogmatical demonstration given in the an-tithesis of the first antinomy. In that demonstra-tion, it was taken for granted that the world is athing in itself—given in its totality prior to allregress, and a determined position in space andtime was denied to it—if it was not considered

as occupying all time and all space. Hence ourconclusion differed from that given above; forwe inferred in the antithesis the actual infinityof the world.]

From this follows the affirmative answer: "Theregress in the series of phenomena—as a de-termination of the cosmical quantity, proceedsin indefinitum." This is equivalent to saying:"The world of sense has no absolute quantity,but the empirical regress (through which alonethe world of sense is presented to us on the sideof its conditions) rests upon a rule, which re-quires it to proceed from every member of theseries, as conditioned, to one still more remote(whether through personal experience, or bymeans of history, or the chain of cause and ef-fect), and not to cease at any point in this exten-sion of the possible empirical employment ofthe understanding." And this is the proper andonly use which reason can make of its princi-ples.

The above rule does not prescribe an unceasingregress in one kind of phenomena. It does not,for example, forbid us, in our ascent from anindividual human being through the line of hisancestors, to expect that we shall discover atsome point of the regress a primeval pair, or toadmit, in the series of heavenly bodies, a sun atthe farthest possible distance from some centre.All that it demands is a perpetual progressfrom phenomena to phenomena, even althoughan actual perception is not presented by them(as in the case of our perceptions being so weakas that we are unable to become conscious ofthem), since they, nevertheless, belong to pos-sible experience.

Every beginning is in time, and all limits toextension are in space. But space and time arein the world of sense. Consequently phenome-na in the world are conditionally limited, butthe world itself is not limited, either conditiona-lly or unconditionally.

For this reason, and because neither the worldnor the cosmical series of conditions to a givenconditioned can be completely given, our con-ception of the cosmical quantity is given only inand through the regress and not prior to it—ina collective intuition. But the regress itself isreally nothing more than the determining of thecosmical quantity, and cannot therefore give usany determined conception of it—still less aconception of a quantity which is, in relation toa certain standard, infinite. The regress doesnot, therefore, proceed to infinity (an infinitygiven), but only to an indefinite extent, for orthe of presenting to us a quantity—realizedonly in and through the regress itself.

II. Solution of the Cosmological Idea of the To-tality of the Division of a Whole given in Intui-tion.

When I divide a whole which is given in intui-tion, I proceed from a conditioned to its condi-tions. The division of the parts of the whole(subdivisio or decompositio) is a regress in theseries of these conditions. The absolute totalityof this series would be actually attained andgiven to the mind, if the regress could arrive atsimple parts. But if all the parts in a continuousdecomposition are themselves divisible, thedivision, that is to say, the regress, proceedsfrom the conditioned to its conditions in infini-tum; because the conditions (the parts) arethemselves contained in the conditioned, and,as the latter is given in a limited intuition, theformer are all given along with it. This regresscannot, therefore, be called a regressus in inde-finitum, as happened in the case of the prece-ding cosmological idea, the regress in whichproceeded from the conditioned to the condi-tions not given contemporaneously and alongwith it, but discoverable only through the em-pirical regress. We are not, however, entitled to

affirm of a whole of this kind, which is divisiblein infinitum, that it consists of an infinite num-ber of parts. For, although all the parts are con-tained in the intuition of the whole, the wholedivision is not contained therein. The divisionis contained only in the progressing decompo-sition—in the regress itself, which is the condi-tion of the possibility and actuality of the series.Now, as this regress is infinite, all the members(parts) to which it attains must be contained inthe given whole as an aggregate. But the com-plete series of division is not contained therein.For this series, being infinite in succession andalways incomplete, cannot represent an infinitenumber of members, and still less a composi-tion of these members into a whole.

To apply this remark to space. Every limitedpart of space presented to intuition is a whole,the parts of which are always spaces—to wha-tever extent subdivided. Every limited space ishence divisible to infinity.

Let us again apply the remark to an externalphenomenon enclosed in limits, that is, a body.The divisibility of a body rests upon the divisi-bility of space, which is the condition of thepossibility of the body as an extended whole. Abody is consequently divisible to infinity,though it does not, for that reason, consist of aninfinite number of parts.

It certainly seems that, as a body must be cogi-tated as substance in space, the law of divisibi-lity would not be applicable to it as substance.For we may and ought to grant, in the case ofspace, that division or decomposition, to anyextent, never can utterly annihilate composition(that is to say, the smallest part of space muststill consist of spaces); otherwise space wouldentirely cease to exist- which is impossible. But,the assertion on the other band that when allcomposition in matter is annihilated in thought,nothing remains, does not seem to harmonizewith the conception of substance, which must

be properly the subject of all composition andmust remain, even after the conjunction of itsattributes in space- which constituted a body—is annihilated in thought. But this is not thecase with substance in the phenomenal world,which is not a thing in itself cogitated by thepure category. Phenomenal substance is not anabsolute subject; it is merely a permanent sen-suous image, and nothing more than an intui-tion, in which the unconditioned is not to befound.

But, although this rule of progress to infinity islegitimate and applicable to the subdivision ofa phenomenon, as a mere occupation or fillingof space, it is not applicable to a whole consis-ting of a number of distinct parts and constitu-ting a quantum discretum—that is to say, anorganized body. It cannot be admitted that eve-ry part in an organized whole is itself organi-zed, and that, in analysing it to infinity, wemust always meet with organized parts; alt-

hough we may allow that the parts of the mat-ter which we decompose in infinitum, may beorganized. For the infinity of the division of aphenomenon in space rests altogether on thefact that the divisibility of a phenomenon isgiven only in and through this infinity, that is,an undetermined number of parts is given,while the parts themselves are given and de-termined only in and through the subdivision;in a word, the infinity of the division necessari-ly presupposes that the whole is not alreadydivided in se. Hence our division determines anumber of parts in the whole—a number whichextends just as far as the actual regress in thedivision; while, on the other hand, the verynotion of a body organized to infinity repre-sents the whole as already and in itself divided.We expect, therefore, to find in it a determinate,but at the same time, infinite, number ofparts—which is self-contradictory. For weshould thus have a whole containing a series ofmembers which could not be completed in any

regress—which is infinite, and at the same timecomplete in an organized composite. Infinitedivisibility is applicable only to a quantum con-tinuum, and is based entirely on the infinitedivisibility of space, But in a quantum discre-tum the multitude of parts or units is alwaysdetermined, and hence always equal to somenumber. To what extent a body may be organi-zed, experience alone can inform us; and alt-hough, so far as our experience of this or thatbody has extended, we may not have discove-red any inorganic part, such parts must exist inpossible experience. But how far the transcen-dental division of a phenomenon must extend,we cannot know from experience—it is a ques-tion which experience cannot answer; it is ans-wered only by the principle of reason whichforbids us to consider the empirical regress, inthe analysis of extended body, as ever absolute-ly complete.

Concluding Remark on the Solution of theTranscendentalMathematical Ideas—and Introductory to theSolution of the Dynamical Ideas.

We presented the antinomy of pure reason in atabular form, and we endeavoured to show theground of this self-contradiction on the part ofreason, and the only means of bringing it to aconclusion— namely, by declaring both con-tradictory statements to be false. We represen-ted in these antinomies the conditions of phe-nomena as belonging to the conditioned accor-ding to relations of space and time- which is theusual supposition of the common understan-ding. In this respect, all dialectical representa-tions of totality, in the series of conditions to agiven conditioned, were perfectly homoge-neous. The condition was always a member ofthe series along with the conditioned, and thusthe homogeneity of the whole series was assu-red. In this case the regress could never be cogi-

tated as complete; or, if this was the case, amember really conditioned was falsely regar-ded as a primal member, consequently as un-conditioned. In such an antinomy, therefore,we did not consider the object, that is, the con-ditioned, but the series of conditions belongingto the object, and the magnitude of that series.And thus arose the difficulty—a difficulty notto be settled by any decision regarding theclaims of the two parties, but simply by cuttingthe knot—by declaring the series proposed byreason to be either too long or too short for theunderstanding, which could in neither casemake its conceptions adequate with the ideas.

But we have overlooked, up to this point, anessential difference existing between the con-ceptions of the understanding which reasonendeavours to raise to the rank of ideas—twoof these indicating a mathematical, and two adynamical synthesis of phenomena. Hitherto, itwas necessary to signalize this distinction; for,

just as in our general representation of all trans-cendental ideas, we considered them underphenomenal conditions, so, in the two mat-hematical ideas, our discussion is concernedsolely with an object in the world of phenome-na. But as we are now about to proceed to theconsideration of the dynamical conceptions ofthe understanding, and their adequatenesswith ideas, we must not lose sight of this dis-tinction. We shall find that it opens up to us anentirely new view of the conflict in which rea-son is involved. For, while in the first two anti-nomies, both parties were dismissed, on theground of having advanced statements basedupon false hypothesis; in the present case thehope appears of discovering a hypothesiswhich may be consistent with the demands ofreason, and, the judge completing the state-ment of the grounds of claim, which both par-ties had left in an unsatisfactory state, the ques-tion may be settled on its own merits, not bydismissing the claimants, but by a comparison

of the arguments on both sides. If we considermerely their extension, and whether they areadequate with ideas, the series of conditionsmay be regarded as all homogeneous. But theconception of the understanding which lies atthe basis of these ideas, contains either a synt-hesis of the homogeneous (presupposed in eve-ry quantity—in its composition as well as in itsdivision) or of the heterogeneous, which is thecase in the dynamical synthesis of cause andeffect, as well as of the necessary and the con-tingent.

Thus it happens that in the mathematical seriesof phenomena no other than a sensuous condi-tion is admissible—a condition which is itself amember of the series; while the dynamical se-ries of sensuous conditions admits a heteroge-neous condition, which is not a member of theseries, but, as purely intelligible, lies out of andbeyond it. And thus reason is satisfied, and anunconditioned placed at the head of the series

of phenomena, without introducing confusioninto or discontinuing it, contrary to the princi-ples of the understanding.

Now, from the fact that the dynamical ideasadmit a condition of phenomena which doesnot form a part of the series of phenomena,arises a result which we should not have expec-ted from an antinomy. In former cases, the re-sult was that both contradictory dialectical sta-tements were declared to be false. In the pre-sent case, we find the conditioned in the dyna-mical series connected with an empirically un-conditioned, but non-sensuous condition; andthus satisfaction is done to the understandingon the one hand and to the reason on the ot-her.* While, moreover, the dialectical argu-ments for unconditioned totality in mere phe-nomena fall to the ground, both propositions ofreason may be shown to be true in their propersignification. This could not happen in the caseof the cosmological ideas which demanded a

mathematically unconditioned unity; for nocondition could be placed at the head of theseries of phenomena, except one which wasitself a phenomenon and consequently a mem-ber of the series.

[*Footnote: For the understanding cannot admitamong phenomena a condition which is itselfempirically unconditioned. But if it is possibleto cogitate an intelligible condition—one whichis not a member of the series of phenomena—for a conditioned phenomenon, without brea-king the series of empirical conditions, such acondition may be admissible as empiricallyunconditioned, and the empirical regress conti-nue regular, unceasing, and intact.]

III. Solution of the Cosmological Idea of theTotality of the Deduction of Cosmical Eventsfrom their Causes.

There are only two modes of causality cogita-ble—the causality of nature or of freedom. Thefirst is the conjunction of a particular state withanother preceding it in the world of sense, theformer following the latter by virtue of a law.Now, as the causality of phenomena is subjectto conditions of time, and the preceding state, ifit had always existed, could not have producedan effect which would make its first appearanceat a particular time, the causality of a causemust itself be an effect—must itself have begunto be, and therefore, according to the principleof the understanding, itself requires a cause.

We must understand, on the contrary, by theterm freedom, in the cosmological sense, a fa-culty of the spontaneous origination of a state;the causality of which, therefore, is not subor-dinated to another cause determining it in time.Freedom is in this sense a pure transcendentalidea, which, in the first place, contains no empi-rical element; the object of which, in the second

place, cannot be given or determined in anyexperience, because it is a universal law of thevery possibility of experience, that everythingwhich happens must have a cause, that conse-quently the causality of a cause, being itselfsomething that has happened, must also have acause. In this view of the case, the whole fieldof experience, how far soever it may extend,contains nothing that is not subject to the lawsof nature. But, as we cannot by this means at-tain to an absolute totality of conditions in refe-rence to the series of causes and effects, reasoncreates the idea of a spontaneity, which canbegin to act of itself, and without any externalcause determining it to action, according to thenatural law of causality.

It is especially remarkable that the practicalconception of freedom is based upon the trans-cendental idea, and that the question of thepossibility of the former is difficult only as itinvolves the consideration of the truth of the

latter. Freedom, in the practical sense, is theindependence of the will of coercion by sen-suous impulses. A will is sensuous, in so far asit is pathologically affected (by sensuous im-pulses); it is termed animal (arbitrium brutum),when it is pathologically necessitated. Thehuman will is certainly an arbitrium sensiti-vum, not brutum, but liberum; because sen-suousness does not necessitate its action, a fa-culty existing in man of self-determination,independently of all sensuous coercion.

It is plain that, if all causality in the world ofsense were natural—and natural only—everyevent would be determined by another accor-ding to necessary laws, and that, consequently,phenomena, in so far as they determine thewill, must necessitate every action as a naturaleffect from themselves; and thus all practicalfreedom would fall to the ground with thetranscendental idea. For the latter presupposesthat although a certain thing has not happened,

it ought to have happened, and that, conse-quently, its phenomenal cause was not so po-werful and determinative as to exclude the cau-sality of our will—a causality capable of produ-cing effects independently of and even in oppo-sition to the power of natural causes, and capa-ble, consequently, of spontaneously originatinga series of events.

Here, too, we find it to be the case, as we gene-rally found in the self-contradictions and per-plexities of a reason which strives to pass thebounds of possible experience, that the problemis properly not physiological, but transcenden-tal. The question of the possibility of freedomdoes indeed concern psychology; but, as it restsupon dialectical arguments of pure reason, itssolution must engage the attention of transcen-dental philosophy. Before attempting this solu-tion, a task which transcendental philosophycannot decline, it will be advisable to make a

remark with regard to its procedure in the set-tlement of the question.

If phenomena were things in themselves, andtime and space forms of the existence of things,condition and conditioned would always bemembers of the same series; and thus wouldarise in the present case the antinomy commonto all transcendental ideas—that their series iseither too great or too small for the understan-ding. The dynamical ideas, which we are aboutto discuss in this and the following section,possess the peculiarity of relating to an object,not considered as a quantity, but as an existen-ce; and thus, in the discussion of the presentquestion, we may make abstraction of thequantity of the series of conditions, and consi-der merely the dynamical relation of the condi-tion to the conditioned. The question, then,suggests itself, whether freedom is possible;and, if it is, whether it can consist with the uni-versality of the natural law of causality; and,

consequently, whether we enounce a properdisjunctive proposition when we say: "Everyeffect must have its origin either in nature or infreedom," or whether both cannot exist togetherin the same event in different relations. Theprinciple of an unbroken connection betweenall events in the phenomenal world, in accor-dance with the unchangeable laws of nature, isa well-established principle of transcendentalanalytic which admits of no exception. Thequestion, therefore, is: "Whether an effect, de-termined according to the laws of nature, can atthe same time be produced by a free agent, orwhether freedom and nature mutually excludeeach other?" And here, the common but falla-cious hypothesis of the absolute reality of phe-nomena manifests its injurious influence inembarrassing the procedure of reason. For ifphenomena are things in themselves, freedomis impossible. In this case, nature is the comple-te and all-sufficient cause of every event; andcondition and conditioned, cause and effect are

contained in the same series, and necessitatedby the same law. If, on the contrary, phenome-na are held to be, as they are in fact, nothingmore than mere representations, connectedwith each other in accordance with empiricallaws, they must have a ground which is notphenomenal. But the causality of such an inte-lligible cause is not determined or determinableby phenomena; although its effects, as pheno-mena, must be determined by other phenome-nal existences. This cause and its causality existtherefore out of and apart from the series ofphenomena; while its effects do exist and arediscoverable in the series of empirical condi-tions. Such an effect may therefore be conside-red to be free in relation to its intelligible cause,and necessary in relation to the phenomenafrom which it is a necessary consequence—adistinction which, stated in this perfectly gene-ral and abstract manner, must appear in thehighest degree subtle and obscure. The sequelwill explain. It is sufficient, at present, to re-

mark that, as the complete and unbroken con-nection of phenomena is an unalterable law ofnature, freedom is impossible—on the supposi-tion that phenomena are absolutely real. Hencethose philosophers who adhere to the commonopinion on this subject can never succeed inreconciling the ideas of nature and freedom.

Possibility of Freedom in Harmony with theUniversal Law of Natural Necessity.

That element in a sensuous object which is notitself sensuous, I may be allowed to term inte-lligible. If, accordingly, an object which must beregarded as a sensuous phenomenon possessesa faculty which is not an object of sensuousintuition, but by means of which it is capable ofbeing the cause of phenomena, the causality ofan object or existence of this kind may be re-garded from two different points of view. It

may be considered to be intelligible, as regardsits action—the action of a thing which is a thingin itself, and sensuous, as regards its effects—the effects of a phenomenon belonging to thesensuous world. We should accordingly, haveto form both an empirical and an intellectualconception of the causality of such a faculty orpower—both, however, having reference to thesame effect. This twofold manner of cogitatinga power residing in a sensuous object does notrun counter to any of the conceptions which weought to form of the world of phenomena or ofa possible experience. Phenomena—not beingthings in themselves—must have a transcen-dental object as a foundation, which determinesthem as mere representations; and there seemsto be no reason why we should not ascribe tothis transcendental object, in addition to theproperty of self-phenomenization, a causalitywhose effects are to be met with in the world ofphenomena, although it is not itself a pheno-menon. But every effective cause must possess

a character, that is to say, a law of its causality,without which it would cease to be a cause. Inthe above case, then, every sensuous objectwould possess an empirical character, whichguaranteed that its actions, as phenomena,stand in complete and harmonious connection,conformably to unvarying natural laws, withall other phenomena, and can be deduced fromthese, as conditions, and that they do thus, inconnection with these, constitute a series in theorder of nature. This sensuous object must, inthe second place, possess an intelligible charac-ter, which guarantees it to be the cause of thoseactions, as phenomena, although it is not itselfa phenomenon nor subordinate to the condi-tions of the world of sense. The former may betermed the character of the thing as a pheno-menon, the latter the character of the thing as athing in itself.

Now this active subject would, in its characterof intelligible subject, be subordinate to no con-

ditions of time, for time is only a condition ofphenomena, and not of things in themselves.No action would begin or cease to be in thissubject; it would consequently be free from thelaw of all determination of time—the law ofchange, namely, that everything which hap-pens must have a cause in the phenomena of apreceding state. In one word, the causality ofthe subject, in so far as it is intelligible, wouldnot form part of the series of empirical condi-tions which determine and necessitate an eventin the world of sense. Again, this intelligiblecharacter of a thing cannot be immediatelycognized, because we can perceive nothing butphenomena, but it must be capable of beingcogitated in harmony with the empirical cha-racter; for we always find ourselves compelledto place, in thought, a transcendental object atthe basis of phenomena although we can neverknow what this object is in itself.

In virtue of its empirical character, this subjectwould at the same time be subordinate to allthe empirical laws of causality, and, as a phe-nomenon and member of the sensuous world,its effects would have to be accounted for by areference to preceding phenomena. Eternalphenomena must be capable of influencing it;and its actions, in accordance with naturallaws, must explain to us how its empirical cha-racter, that is, the law of its causality, is to becognized in and by means of experience. In aword, all requisites for a complete and necessa-ry determination of these actions must be pre-sented to us by experience.

In virtue of its intelligible character, on the ot-her hand (although we possess only a generalconception of this character), the subject mustbe regarded as free from all sensuous influen-ces, and from all phenomenal determination.Moreover, as nothing happens in this subject—for it is a noumenon, and there does not conse-

quently exist in it any change, demanding thedynamical determination of time, and for thesame reason no connection with phenomena ascauses—this active existence must in its actionsbe free from and independent of natural neces-sity, for or necessity exists only in the world ofphenomena. It would be quite correct to saythat it originates or begins its effects in theworld of sense from itself, although the actionproductive of these effects does not begin initself. We should not be in this case affirmingthat these sensuous effects began to exist ofthemselves, because they are always determi-ned by prior empirical conditions—by virtue ofthe empirical character, which is the phenome-non of the intelligible character—and are possi-ble only as constituting a continuation of theseries of natural causes. And thus nature andfreedom, each in the complete and absolutesignification of these terms, can exist, withoutcontradiction or disagreement, in the same ac-tion.

Exposition of the Cosmological Idea of Free-dom in Harmony with the Universal Law ofNatural Necessity.

I have thought it advisable to lay before thereader at first merely a sketch of the solution ofthis transcendental problem, in order to enablehim to form with greater ease a clear concep-tion of the course which reason must adopt inthe solution. I shall now proceed to exhibit theseveral momenta of this solution, and to consi-der them in their order.

The natural law that everything which happensmust have a cause, that the causality of thiscause, that is, the action of the cause (whichcannot always have existed, but must be itselfan event, for it precedes in time some effectwhich it has originated), must have itself aphenomenal cause, by which it is determinedand, and, consequently, all events are empirica-lly determined in an order of nature—this law,I say, which lies at the foundation of the possi-

bility of experience, and of a connected systemof phenomena or nature is a law of the unders-tanding, from which no departure, and towhich no exception, can be admitted. For toexcept even a single phenomenon from its ope-ration is to exclude it from the sphere of possi-ble experience and thus to admit it to be a merefiction of thought or phantom of the brain.

Thus we are obliged to acknowledge the exis-tence of a chain of causes, in which, however,absolute totality cannot be found. But we neednot detain ourselves with this question, for ithas already been sufficiently answered in ourdiscussion of the antinomies into which reasonfalls, when it attempts to reach the unconditio-ned in the series of phenomena. If we permitourselves to be deceived by the illusion oftranscendental idealism, we shall find that neit-her nature nor freedom exists. Now the ques-tion is: "Whether, admitting the existence ofnatural necessity in the world of phenomena, it

is possible to consider an effect as at the sametime an effect of nature and an effect of free-dom—or, whether these two modes of causalityare contradictory and incompatible?"

No phenomenal cause can absolutely and ofitself begin a series. Every action, in so far as itis productive of an event, is itself an event oroccurrence, and presupposes another precedingstate, in which its cause existed. Thus everyt-hing that happens is but a continuation of aseries, and an absolute beginning is impossiblein the sensuous world. The actions of naturalcauses are, accordingly, themselves effects, andpresuppose causes preceding them in time. Aprimal action which forms an absolute begin-ning, is beyond the causal power of phenome-na.

Now, is it absolutely necessary that, grantingthat all effects are phenomena, the causality ofthe cause of these effects must also be a phe-nomenon and belong to the empirical world? Is

it not rather possible that, although every effectin the phenomenal world must be connectedwith an empirical cause, according to the uni-versal law of nature, this empirical causalitymay be itself the effect of a non-empirical andintelligible causality—its connection with natu-ral causes remaining nevertheless intact? Sucha causality would be considered, in reference tophenomena, as the primal action of a cause,which is in so far, therefore, not phenomenal,but, by reason of this faculty or power, intelli-gible; although it must, at the same time, as alink in the chain of nature, be regarded as be-longing to the sensuous world.

A belief in the reciprocal causality of phenome-na is necessary, if we are required to look forand to present the natural conditions of naturalevents, that is to say, their causes. This beingadmitted as unexceptionably valid, the requi-rements of the understanding, which recogni-zes nothing but nature in the region of pheno-

mena, are satisfied, and our physical explana-tions of physical phenomena may proceed intheir regular course, without hindrance andwithout opposition. But it is no stumbling-block in the way, even assuming the idea to bea pure fiction, to admit that there are some na-tural causes in the possession of a faculty whichis not empirical, but intelligible, inasmuch as itis not determined to action by empirical condi-tions, but purely and solely upon groundsbrought forward by the understanding—thisaction being still, when the cause is phenome-nized, in perfect accordance with the laws ofempirical causality. Thus the acting subject, as acausal phenomenon, would continue to preser-ve a complete connection with nature and natu-ral conditions; and the phenomenon only of thesubject (with all its phenomenal causality)would contain certain conditions, which, if weascend from the empirical to the transcendentalobject, must necessarily be regarded as intelli-gible. For, if we attend, in our inquiries with

regard to causes in the world of phenomena, tothe directions of nature alone, we need nottrouble ourselves about the relation in whichthe transcendental subject, which is completelyunknown to us, stands to these phenomena andtheir connection in nature. The intelligibleground of phenomena in this subject does notconcern empirical questions. It has to do onlywith pure thought; and, although the effects ofthis thought and action of the pure understan-ding are discoverable in phenomena, thesephenomena must nevertheless be capable of afull and complete explanation, upon purelyphysical grounds and in accordance with natu-ral laws. And in this case we attend solely totheir empirical and omit all consideration oftheir intelligible character (which is the trans-cendental cause of the former) as completelyunknown, except in so far as it is exhibited bythe latter as its empirical symbol. Now let usapply this to experience. Man is a phenomenonof the sensuous world and, at the same time,

therefore, a natural cause, the causality ofwhich must be regulated by empirical laws. Assuch, he must possess an empirical character,like all other natural phenomena. We remarkthis empirical character in his actions, whichreveal the presence of certain powers and facul-ties. If we consider inanimate or merely animalnature, we can discover no reason for ascribingto ourselves any other than a faculty which isdetermined in a purely sensuous manner. Butman, to whom nature reveals herself onlythrough sense, cognizes himself not only by hissenses, but also through pure apperception;and this in actions and internal determinations,which he cannot regard as sensuous impres-sions. He is thus to himself, on the one hand, aphenomenon, but on the other hand, in respectof certain faculties, a purely intelligible object—intelligible, because its action cannot be ascri-bed to sensuous receptivity. These faculties areunderstanding and reason. The latter, especia-lly, is in a peculiar manner distinct from all

empirically-conditioned faculties, for it em-ploys ideas alone in the consideration of itsobjects, and by means of these determines theunderstanding, which then proceeds to makean empirical use of its own conceptions, which,like the ideas of reason, are pure and non-empirical.

That reason possesses the faculty of causality,or that at least we are compelled so to representit, is evident from the imperatives, which in thesphere of the practical we impose on many ofour executive powers. The words I ought ex-press a species of necessity, and imply a con-nection with grounds which nature does notand cannot present to the mind of man. Un-derstanding knows nothing in nature but thatwhich is, or has been, or will be. It would beabsurd to say that anything in nature ought tobe other than it is in the relations of time inwhich it stands; indeed, the ought, when weconsider merely the course of nature, has neit-

her application nor meaning. The question,"What ought to happen in the sphere of natu-re?" is just as absurd as the question, "Whatought to be the properties of a circle?" All thatwe are entitled to ask is, "What takes place innature?" or, in the latter case, "What are theproperties of a circle?"

But the idea of an ought or of duty indicates apossible action, the ground of which is a pureconception; while the ground of a merely natu-ral action is, on the contrary, always a pheno-menon. This action must certainly be possibleunder physical conditions, if it is prescribed bythe moral imperative ought; but these physicalor natural conditions do not concern the deter-mination of the will itself, they relate to its ef-fects alone, and the consequences of the effectin the world of phenomena. Whatever numberof motives nature may present to my will, wha-tever sensuous impulses—the moral ought it isbeyond their power to produce. They may pro-

duce a volition, which, so far from being neces-sary, is always conditioned—a volition towhich the ought enunciated by reason, sets anaim and a standard, gives permission or prohi-bition. Be the object what it may, purely sen-suous—as pleasure, or presented by pure rea-son—as good, reason will not yield to groundswhich have an empirical origin. Reason will notfollow the order of things presented by expe-rience, but, with perfect spontaneity, rearrangesthem according to ideas, with which it compelsempirical conditions to agree. It declares, in thename of these ideas, certain actions to be neces-sary which nevertheless have not taken placeand which perhaps never will take place; andyet presupposes that it possesses the faculty ofcausality in relation to these actions. For, in theabsence of this supposition, it could not expectits ideas to produce certain effects in the worldof experience.

Now, let us stop here and admit it to be at leastpossible that reason does stand in a really cau-sal relation to phenomena. In this case itmust—pure reason as it is—exhibit an empiri-cal character. For every cause supposes a rule,according to which certain phenomena followas effects from the cause, and every rule requi-res uniformity in these effects; and this is theproper ground of the conception of a cause—asa faculty or power. Now this conception (of acause) may be termed the empirical character ofreason; and this character is a permanent one,while the effects produced appear, in conformi-ty with the various conditions which accompa-ny and partly limit them, in various forms.

Thus the volition of every man has an empiricalcharacter, which is nothing more than the cau-sality of his reason, in so far as its effects in thephenomenal world manifest the presence of arule, according to which we are enabled to exa-mine, in their several kinds and degrees, the

actions of this causality and the rationalgrounds for these actions, and in this way todecide upon the subjective principles of thevolition. Now we learn what this empirical cha-racter is only from phenomenal effects, andfrom the rule of these which is presented byexperience; and for this reason all the actions ofman in the world of phenomena are determi-ned by his empirical character, and the co-operative causes of nature. If, then, we couldinvestigate all the phenomena of human voli-tion to their lowest foundation in the mind,there would be no action which we could notanticipate with certainty, and recognize to beabsolutely necessary from its preceding condi-tions. So far as relates to this empirical charac-ter, therefore, there can be no freedom; and it isonly in the light of this character that we canconsider the human will, when we confine our-selves to simple observation and, as is the casein anthropology, institute a physiological inves-tigation of the motive causes of human actions.

But when we consider the same actions in rela-tion to reason—not for the purpose of explai-ning their origin, that is, in relation to specula-tive reason, but to practical reason, as the pro-ducing cause of these actions—we shall disco-ver a rule and an order very different from tho-se of nature and experience. For the declarationof this mental faculty may be that what has andcould not but take place in the course of nature,ought not to have taken place. Sometimes, too,we discover, or believe that we discover, thatthe ideas of reason did actually stand in a cau-sal relation to certain actions of man; and thatthese actions have taken place because theywere determined, not by empirical causes, butby the act of the will upon grounds of reason.

Now, granting that reason stands in a causalrelation to phenomena; can an action of reasonbe called free, when we know that, sensuously,in its empirical character, it is completely de-termined and absolutely necessary? But this

empirical character is itself determined by theintelligible character. The latter we cannot cog-nize; we can only indicate it by means of phe-nomena, which enable us to have an immediatecognition only of the empirical character.* Anaction, then, in so far as it is to be ascribed to anintelligible cause, does not result from it in ac-cordance with empirical laws. That is to say,not the conditions of pure reason, but only theireffects in the internal sense, precede the act.Pure reason, as a purely intelligible faculty, isnot subject to the conditions of time. The causa-lity of reason in its intelligible character doesnot begin to be; it does not make its appearanceat a certain time, for the purpose of producingan effect. If this were not the case, the causalityof reason would be subservient to the naturallaw of phenomena, which determines themaccording to time, and as a series of causes andeffects in time; it would consequently cease tobe freedom and become a part of nature. Weare therefore justified in saying: "If reason

stands in a causal relation to phenomena, it is afaculty which originates the sensuous conditionof an empirical series of effects." For the condi-tion, which resides in the reason, is non-sensuous, and therefore cannot be originated,or begin to be. And thus we find—what wecould not discover in any empirical series—acondition of a successive series of events itselfempirically unconditioned. For, in the presentcase, the condition stands out of and beyondthe series of phenomena—it is intelligible, andit consequently cannot be subjected to any sen-suous condition, or to any time-determinationby a preceding cause.

[*Footnote: The real morality of actions—theirmerit or demerit, and even that of our own con-duct, is completely unknown to us. Our estima-tes can relate only to their empirical character.How much is the result of the action of freewill, how much is to be ascribed to nature andto blameless error, or to a happy constitution of

temperament (merito fortunae), no one candiscover, nor, for this reason, determine withperfect justice.]

But, in another respect, the same cause belongsalso to the series of phenomena. Man is himselfa phenomenon. His will has an empirical cha-racter, which is the empirical cause of all hisactions. There is no condition—determiningman and his volition in conformity with thischaracter—which does not itself form part ofthe series of effects in nature, and is subject totheir law—the law according to which an empi-rically undetermined cause of an event in timecannot exist. For this reason no given action canhave an absolute and spontaneous origination,all actions being phenomena, and belonging tothe world of experience. But it cannot be said ofreason, that the state in which it determines thewill is always preceded by some other statedetermining it. For reason is not a phenome-non, and therefore not subject to sensuous con-

ditions; and, consequently, even in relation toits causality, the sequence or conditions of timedo not influence reason, nor can the dynamicallaw of nature, which determines the sequenceof time according to certain rules, be applied toit.

Reason is consequently the permanent condi-tion of all actions of the human will. Each ofthese is determined in the empirical characterof the man, even before it has taken place. Theintelligible character, of which the former is butthe sensuous schema, knows no before or after;and every action, irrespective of the time-relation in which it stands with other pheno-mena, is the immediate effect of the intelligiblecharacter of pure reason, which, consequently,enjoys freedom of action, and is not dynamica-lly determined either by internal or externalpreceding conditions. This freedom must notbe described, in a merely negative manner, asindependence of empirical conditions, for in

this case the faculty of reason would cease to bea cause of phenomena; but it must be regarded,positively, as a faculty which can spontaneous-ly originate a series of events. At the same time,it must not be supposed that any beginning cantake place in reason; on the contrary, reason, asthe unconditioned condition of all action of thewill, admits of no time-conditions, although itseffect does really begin in a series of phenome-na—a beginning which is not, however, absolu-tely primal.

I shall illustrate this regulative principle of rea-son by an example, from its employment in theworld of experience; proved it cannot be by anyamount of experience, or by any number offacts, for such arguments cannot establish thetruth of transcendental propositions. Let ustake a voluntary action—for example, a false-hood—by means of which a man has introdu-ced a certain degree of confusion into the sociallife of humanity, which is judged according to

the motives from which it originated, and theblame of which and of the evil consequencesarising from it, is imputed to the offender. Weat first proceed to examine the empirical cha-racter of the offence, and for this purpose weendeavour to penetrate to the sources of thatcharacter, such as a defective education, badcompany, a shameless and wicked disposition,frivolity, and want of reflection—not forgettingalso the occasioning causes which prevailed atthe moment of the transgression. In this theprocedure is exactly the same as that pursuedin the investigation of the series of causeswhich determine a given physical effect. Now,although we believe the action to have beendetermined by all these circumstances, we donot the less blame the offender. We do not bla-me him for his unhappy disposition, nor for thecircumstances which influenced him, nay, noteven for his former course of life; for we pre-suppose that all these considerations may be setaside, that the series of preceding conditions

may be regarded as having never existed, andthat the action may be considered as complete-ly unconditioned in relation to any state prece-ding, just as if the agent commenced with it anentirely new series of effects. Our blame of theoffender is grounded upon a law of reason,which requires us to regard this faculty as acause, which could have and ought to haveotherwise determined the behaviour of the cul-prit, independently of all empirical conditions.This causality of reason we do not regard as aco-operating agency, but as complete in itself. Itmatters not whether the sensuous impulsesfavoured or opposed the action of this causali-ty, the offence is estimated according to its inte-lligible character—the offender is decidedlyworthy of blame, the moment he utters a false-hood. It follows that we regard reason, in spiteof the empirical conditions of the act, as com-pletely free, and therefore, therefore, as in thepresent case, culpable.

The above judgement is complete evidence thatwe are accustomed to think that reason is notaffected by sensuous conditions, that in it nochange takes place—although its phenomena,in other words, the mode in which it appears inits effects, are subject to change—that in it nopreceding state determines the following, and,consequently, that it does not form a memberof the series of sensuous conditions which ne-cessitate phenomena according to natural laws.Reason is present and the same in all humanactions and at all times; but it does not itselfexist in time, and therefore does not enter uponany state in which it did not formerly exist. It is,relatively to new states or conditions, determi-ning, but not determinable. Hence we cannotask: "Why did not reason determine itself in adifferent manner?" The question ought to bethus stated: "Why did not reason employ itspower of causality to determine certain phe-nomena in a different manner?" But this is aquestion which admits of no answer. For a dif-

ferent intelligible character would have exhibi-ted a different empirical character; and, whenwe say that, in spite of the course which hiswhole former life has taken, the offender couldhave refrained from uttering the falsehood, thismeans merely that the act was subject to thepower and authority- permissive or prohibiti-ve—of reason. Now, reason is not subject in itscausality to any conditions of phenomena or oftime; and a difference in time may produce adifference in the relation of phenomena to eachother—for these are not things and thereforenot causes in themselves—but it cannot produ-ce any difference in the relation in which theaction stands to the faculty of reason.

Thus, then, in our investigation into free actionsand the causal power which produced them,we arrive at an intelligible cause, beyondwhich, however, we cannot go; although wecan recognize that it is free, that is, independentof all sensuous conditions, and that, in this

way, it may be the sensuously unconditionedcondition of phenomena. But for what reasonthe intelligible character generates such andsuch phenomena and exhibits such and such anempirical character under certain circumstan-ces, it is beyond the power of our reason to de-cide. The question is as much above the powerand the sphere of reason as the followingwould be: "Why does the transcendental objectof our external sensuous intuition allow of noother form than that of intuition in space?" Butthe problem, which we were called upon tosolve, does not require us to entertain any suchquestions. The problem was merely this—whether freedom and natural necessity canexist without opposition in the same action. Tothis question we have given a sufficient answer;for we have shown that, as the former stands ina relation to a different kind of condition fromthose of the latter, the law of the one does notaffect the law of the other and that, consequen-

tly, both can exist together in independence ofand without interference with each other.

The reader must be careful to remark that myintention in the above remarks has not been toprove the actual existence of freedom, as a fa-culty in which resides the cause of certain sen-suous phenomena. For, not to mention thatsuch an argument would not have a transcen-dental character, nor have been limited to thediscussion of pure conceptions—all attempts atinferring from experience what cannot be cogi-tated in accordance with its laws, must ever beunsuccessful. Nay, more, I have not even aimedat demonstrating the possibility of freedom; forthis too would have been a vain endeavour,inasmuch as it is beyond the power of the mindto cognize the possibility of a reality or of acausal power by the aid of mere a priori con-ceptions. Freedom has been considered in theforegoing remarks only as a transcendentalidea, by means of which reason aims at origina-

ting a series of conditions in the world of phe-nomena with the help of that which is sen-suously unconditioned, involving itself, howe-ver, in an antinomy with the laws which itselfprescribes for the conduct of the understan-ding. That this antinomy is based upon a mereillusion, and that nature and freedom are atleast not opposed—this was the only thing inour power to prove, and the question which itwas our task to solve.

IV. Solution of the Cosmological Idea of theTotality of the Dependence of PhenomenalExistences.

In the preceding remarks, we considered thechanges in the world of sense as constituting adynamical series, in which each member is su-bordinated to another—as its cause. Our pre-sent purpose is to avail ourselves of this series

of states or conditions as a guide to an existencewhich may be the highest condition of all chan-geable phenomena, that is, to a necessary being.Our endeavour to reach, not the unconditionedcausality, but the unconditioned existence, ofsubstance. The series before us is therefore aseries of conceptions, and not of intuitions (inwhich the one intuition is the condition of theother).

But it is evident that, as all phenomena are sub-ject to change and conditioned in their existen-ce, the series of dependent existences cannotembrace an unconditioned member, the exis-tence of which would be absolutely necessary.It follows that, if phenomena were things inthemselves, and—as an immediate consequen-ce from this supposition- condition and condi-tioned belonged to the same series of pheno-mena, the existence of a necessary being, as thecondition of the existence of sensuous pheno-mena, would be perfectly impossible.

An important distinction, however, exists bet-ween the dynamical and the mathematical re-gress. The latter is engaged solely with thecombination of parts into a whole, or with thedivision of a whole into its parts; and thereforeare the conditions of its series parts of the se-ries, and to be consequently regarded as homo-geneous, and for this reason, as consisting, wit-hout exception, of phenomena. If the formerregress, on the contrary, the aim of which is notto establish the possibility of an unconditionedwhole consisting of given parts, or of an un-conditioned part of a given whole, but todemonstrate the possibility of the deduction ofa certain state from its cause, or of thecontingent existence of substance from thatwhich exists necessarily, it is not requisite thatthe condition should form part of an empiricalseries along with the conditioned.

In the case of the apparent antinomy withwhich we are at present dealing, there exists a

way of escape from the difficulty; for it is notimpossible that both of the contradictory sta-tements may be true in different relations. Allsensuous phenomena may be contingent, andconsequently possess only an empirically con-ditioned existence, and yet there may also exista non-empirical condition of the whole series,or, in other words, a necessary being. For thisnecessary being, as an intelligible condition,would not form a member—not even the hig-hest member—of the series; the whole world ofsense would be left in its empirically determi-ned existence uninterfered with and uninfluen-ced. This would also form a ground of distinc-tion between the modes of solution employedfor the third and fourth antinomies. For, whilein the consideration of freedom in the formerantinomy, the thing itself—the cause (substan-tia phaenomenon)—was regarded as belongingto the series of conditions, and only its causalityto the intelligible world—we are obliged in thepresent case to cogitate this necessary being as

purely intelligible and as existing entirely apartfrom the world of sense (as an ens extramun-danum); for otherwise it would be subject tothe phenomenal law of contingency and de-pendence.

In relation to the present problem, therefore,the regulative principle of reason is that eve-rything in the sensuous world possesses anempirically conditioned existence—that noproperty of the sensuous world possesses un-conditioned necessity—that we are bound toexpect, and, so far as is possible, to seek for theempirical condition of every member in theseries of conditions—and that there is no suffi-cient reason to justify us in deducing any exis-tence from a condition which lies out of andbeyond the empirical series, or in regardingany existence as independent and self-subsistent; although this should not prevent usfrom recognizing the possibility of the wholeseries being based upon a being which is inte-

lligible, and for this reason free from all empiri-cal conditions.

But it has been far from my intention, in theseremarks, to prove the existence of this uncondi-tioned and necessary being, or even to evidencethe possibility of a purely intelligible conditionof the existence or all sensuous phenomena. Asbounds were set to reason, to prevent it fromleaving the guiding thread of empirical condi-tions and losing itself in transcendent theorieswhich are incapable of concrete presentation; soit was my purpose, on the other band, to setbounds to the law of the purely empirical un-derstanding, and to protest against any at-tempts on its part at deciding on the possibilityof things, or declaring the existence of the inte-lligible to be impossible, merely on the groundthat it is not available for the explanation andexposition of phenomena. It has been shown, atthe same time, that the contingency of all thephenomena of nature and their empirical con-

ditions is quite consistent with the arbitraryhypothesis of a necessary, although purely inte-lligible condition, that no real contradictionexists between them and that, consequently,both may be true. The existence of such an ab-solutely necessary being may be impossible;but this can never be demonstrated from theuniversal contingency and dependence of sen-suous phenomena, nor from the principlewhich forbids us to discontinue the series atsome member of it, or to seek for its cause insome sphere of existence beyond the world ofnature. Reason goes its way in the empiricalworld, and follows, too, its peculiar path in thesphere of the transcendental.

The sensuous world contains nothing but phe-nomena, which are mere representations, andalways sensuously conditioned; things in them-selves are not, and cannot be, objects to us. It isnot to be wondered at, therefore, that we arenot justified in leaping from some member of

an empirical series beyond the world of sense,as if empirical representations were things inthemselves, existing apart from their transcen-dental ground in the human mind, and the cau-se of whose existence may be sought out of theempirical series. This would certainly be thecase with contingent things; but it cannot bewith mere representations of things, the con-tingency of which is itself merely a phenome-non and can relate to no other regress than thatwhich determines phenomena, that is, the em-pirical. But to cogitate an intelligible ground ofphenomena, as free, moreover, from the con-tingency of the latter, conflicts neither with theunlimited nature of the empirical regress, norwith the complete contingency of phenomena.And the demonstration of this was the onlything necessary for the solution of this apparentantinomy. For if the condition of every condi-tioned—as regards its existence—is sensuous,and for this reason a part of the same series, itmust be itself conditioned, as was shown in the

antithesis of the fourth antinomy. The emba-rrassments into which a reason, which postula-tes the unconditioned, necessarily falls, must,therefore, continue to exist; or the unconditio-ned must be placed in the sphere of the intelli-gible. In this way, its necessity does not require,nor does it even permit, the presence of an em-pirical condition: and it is, consequently, un-conditionally necessary.

The empirical employment of reason is not af-fected by the assumption of a purely intelligiblebeing; it continues its operations on the princi-ple of the contingency of all phenomena, pro-ceeding from empirical conditions to still hig-her and higher conditions, themselves empiri-cal. Just as little does this regulative principleexclude the assumption of an intelligible cause,when the question regards merely the pureemployment of reason—in relation to ends oraims. For, in this case, an intelligible cause sig-nifies merely the transcendental and to us unk-

nown ground of the possibility of sensuousphenomena, and its existence, necessary andindependent of all sensuous conditions, is notinconsistent with the contingency of phenome-na, or with the unlimited possibility of regresswhich exists in the series of empirical condi-tions.

Concluding Remarks on the Antinomy of PureReason.

So long as the object of our rational conceptionsis the totality of conditions in the world of phe-nomena, and the satisfaction, from this source,of the requirements of reason, so long are ourideas transcendental and cosmological. Butwhen we set the unconditioned- which is theaim of all our inquiries—in a sphere which liesout of the world of sense and possible expe-rience, our ideas become transcendent. They

are then not merely serviceable towards thecompletion of the exercise of reason (whichremains an idea, never executed, but always tobe pursued); they detach themselves complete-ly from experience and construct for themsel-ves objects, the material of which has not beenpresented by experience, and the objective rea-lity of which is not based upon the completionof the empirical series, but upon pure a prioriconceptions. The intelligible object of thesetranscendent ideas may be conceded, as atranscendental object. But we cannot cogitate itas a thing determinable by certain distinct pre-dicates relating to its internal nature, for it hasno connection with empirical conceptions; norare we justified in affirming the existence ofany such object. It is, consequently, a mere pro-duct of the mind alone. Of all the cosmologicalideas, however, it is that occasioning the fourthantinomy which compels us to venture uponthis step. For the existence of phenomena, al-ways conditioned and never self-subsistent,

requires us to look for an object different fromphenomena—an intelligible object, with whichall contingency must cease. But, as we haveallowed ourselves to assume the existence of aself-subsistent reality out of the field of expe-rience, and are therefore obliged to regard phe-nomena as merely a contingent mode of repre-senting intelligible objects employed by beingswhich are themselves intelligences—no othercourse remains for us than to follow analogyand employ the same mode in forming someconception of intelligible things, of which wehave not the least knowledge, which naturetaught us to use in the formation of empiricalconceptions. Experience made us acquaintedwith the contingent. But we are at present en-gaged in the discussion of things which are notobjects of experience; and must, therefore, de-duce our knowledge of them from that which isnecessary absolutely and in itself, that is, frompure conceptions. Hence the first step whichwe take out of the world of sense obliges us to

begin our system of new cognition with theinvestigation of a necessary being, and to de-duce from our conceptions of it all our concep-tions of intelligible things. This we propose toattempt in the following chapter.

CHAPTER III. The Ideal of Pure Reason.

SECTION I. Of the Ideal in General.

We have seen that pure conceptions do notpresent objects to the mind, except under sen-suous conditions; because the conditions ofobjective reality do not exist in these concep-tions, which contain, in fact, nothing but themere form of thought. They may, however,when applied to phenomena, be presented inconcreto; for it is phenomena that present to

them the materials for the formation of empiri-cal conceptions, which are nothing more thanconcrete forms of the conceptions of the un-derstanding. But ideas are still further removedfrom objective reality than categories; for nophenomenon can ever present them to thehuman mind in concreto. They contain a certainperfection, attainable by no possible empiricalcognition; and they give to reason a systematicunity, to which the unity of experience at-tempts to approximate, but can never comple-tely attain.

But still further removed than the idea fromobjective reality is the Ideal, by which term Iunderstand the idea, not in concreto, but inindividuo—as an individual thing, determina-ble or determined by the idea alone. The idea ofhumanity in its complete perfection supposesnot only the advancement of all the powers andfaculties, which constitute our conception ofhuman nature, to a complete attainment of

their final aims, but also everything which isrequisite for the complete determination of theidea; for of all contradictory predicates, onlyone can conform with the idea of the perfectman. What I have termed an ideal was in Pla-to's philosophy an idea of the divine mind—anindividual object present to its pure intuition,the most perfect of every kind of possiblebeings, and the archetype of all phenomenalexistences.

Without rising to these speculative heights, weare bound to confess that human reason con-tains not only ideas, but ideals, which possess,not, like those of Plato, creative, but certainlypractical power—as regulative principles, andform the basis of the perfectibility of certainactions. Moral conceptions are not perfectlypure conceptions of reason, because an empiri-cal element—of pleasure or pain—lies at thefoundation of them. In relation, however, to theprinciple, whereby reason sets bounds to a

freedom which is in itself without law, andconsequently when we attend merely to theirform, they may be considered as pure concep-tions of reason. Virtue and wisdom in theirperfect purity are ideas. But the wise man ofthe Stoics is an ideal, that is to say, a humanbeing existing only in thought and in completeconformity with the idea of wisdom. As theidea provides a rule, so the ideal serves as anarchetype for the perfect and complete deter-mination of the copy. Thus the conduct of thiswise and divine man serves us as a standard ofaction, with which we may compare and judgeourselves, which may help us to reform oursel-ves, although the perfection it demands cannever be attained by us. Although we cannotconcede objective reality to these ideals, theyare not to be considered as chimeras; on thecontrary, they provide reason with a standard,which enables it to estimate, by comparison,the degree of incompleteness in the objects pre-sented to it. But to aim at realizing the ideal in

an example in the world of experience—to des-cribe, for instance, the character of the perfectlywise man in a romance—is impracticable. Naymore, there is something absurd in the attempt;and the result must be little edifying, as thenatural limitations, which are continually brea-king in upon the perfection and completenessof the idea, destroy the illusion in the story andthrow an air of suspicion even on what is goodin the idea, which hence appears fictitious andunreal.

Such is the constitution of the ideal of reason,which is always based upon determinate con-ceptions, and serves as a rule and a model forlimitation or of criticism. Very different is thenature of the ideals of the imagination. Of theseit is impossible to present an intelligible con-ception; they are a kind of monogram, drawnaccording to no determinate rule, and formingrather a vague picture—the production of ma-ny diverse experiences—than a determinate

image. Such are the ideals which painters andphysiognomists profess to have in their minds,and which can serve neither as a model forproduction nor as a standard for appreciation.They may be termed, though improperly, sen-suous ideals, as they are declared to be modelsof certain possible empirical intuitions. Theycannot, however, furnish rules or standards forexplanation or examination.

In its ideals, reason aims at complete and per-fect determination according to a priori rules;and hence it cogitates an object, which must becompletely determinable in conformity withprinciples, although all empirical conditions areabsent, and the conception of the object is onthis account transcendent.

SECTION II. Of the Transcendental Ideal(Prototypon Trancendentale).

Every conception is, in relation to that which isnot contained in it, undetermined and subjectto the principle of determinability. This princi-ple is that, of every two contradictorily oppo-sed predicates, only one can belong to a con-ception. It is a purely logical principle, itselfbased upon the principle of contradiction;inasmuch as it makes complete abstraction ofthe content and attends merely to the logicalform of the cognition.

But again, everything, as regards its possibility,is also subject to the principle of complete de-termination, according to which one of all thepossible contradictory predicates of thingsmust belong to it. This principle is not basedmerely upon that of contradiction; for, in addi-tion to the relation between two contradictorypredicates, it regards everything as standing in

a relation to the sum of possibilities, as the sumtotal of all predicates of things, and, while pre-supposing this sum as an a priori condition,presents to the mind everything as receivingthe possibility of its individual existence fromthe relation it bears to, and the share it posses-ses in, the aforesaid sum of possibilities.* Theprinciple of complete determination relates thecontent and not to the logical form. It is theprinciple of the synthesis of all the predicateswhich are required to constitute the completeconception of a thing, and not a mere principleanalytical representation, which enounces thatone of two contradictory predicates must be-long to a conception. It contains, moreover, atranscendental presupposition— that, namely,of the material for all possibility, which mustcontain a priori the data for this or that particu-lar possibility.

[*Footnote: Thus this principle declares everyt-hing to possess a relation to a common correla-

te—the sum-total of possibility, which, if disco-vered to exist in the idea of one individualthing, would establish the affinity of all possi-ble things, from the identity of the ground oftheir complete determination. The determinabi-lity of every conception is subordinate to theuniversality (Allgemeinheit, universalitas) ofthe principle of excluded middle; the determi-nation of a thing to the totality (Allheit, univer-sitas) of all possible predicates.]

The proposition, everything which exists iscompletely determined, means not only thatone of every pair of given contradictory attribu-tes, but that one of all possible attributes, isalways predicable of the thing; in it the predica-tes are not merely compared logically with eachother, but the thing itself is transcendentallycompared with the sum-total of all possiblepredicates. The proposition is equivalent tosaying: "To attain to a complete knowledge of athing, it is necessary to possess a knowledge of

everything that is possible, and to determine itthereby in a positive or negative manner." Theconception of complete determination is conse-quently a conception which cannot be presen-ted in its totality in concreto, and is thereforebased upon an idea, which has its seat in thereason—the faculty which prescribes to theunderstanding the laws of its harmonious andperfect exercise.

Now, although this idea of the sum-total of allpossibility, in so far as it forms the condition ofthe complete determination of everything, isitself undetermined in relation to the predicateswhich may constitute this sum-total, and wecogitate in it merely the sum-total of all possi-ble predicates—we nevertheless find, uponcloser examination, that this idea, as a primitiveconception of the mind, excludes a large num-ber of predicates—those deduced and thoseirreconcilable with others, and that it is evolvedas a conception completely determined a priori.

Thus it becomes the conception of an indivi-dual object, which is completely determined byand through the mere idea, and must conse-quently be termed an ideal of pure reason.

When we consider all possible predicates, notmerely logically, but transcendentally, that is tosay, with reference to the content which may becogitated as existing in them a priori, we shallfind that some indicate a being, others merely anon-being. The logical negation expressed inthe word not does not properly belong to a con-ception, but only to the relation of one concep-tion to another in a judgement, and is conse-quently quite insufficient to present to the mindthe content of a conception. The expression notmortal does not indicate that a non-being iscogitated in the object; it does not concern thecontent at all. A transcendental negation, on thecontrary, indicates non-being in itself, and isopposed to transcendental affirmation, the con-ception of which of itself expresses a being.

Hence this affirmation indicates a reality, be-cause in and through it objects are consideredto be something—to be things; while the oppo-site negation, on the other band, indicates amere want, or privation, or absence, and, wheresuch negations alone are attached to a represen-tation, the non-existence of anything corres-ponding to the representation.

Now a negation cannot be cogitated as deter-mined, without cogitating at the same time theopposite affirmation. The man born blind hasnot the least notion of darkness, because he hasnone of light; the vagabond knows nothing ofpoverty, because he has never known what it isto be in comfort;* the ignorant man has no con-ception of his ignorance, because he has noconception of knowledge. All conceptions ofnegatives are accordingly derived or deducedconceptions; and realities contain the data, and,so to speak, the material or transcendental con-

tent of the possibility and complete determina-tion of all things.

[*Footnote: The investigations and calculationsof astronomers have taught us much that iswonderful; but the most important lesson wehave received from them is the discovery of theabyss of our ignorance in relation to the univer-se—an ignorance the magnitude of which rea-son, without the information thus derived,could never have conceived. This discovery ofour deficiencies must produce a great change inthe determination of the aims of human rea-son.]

If, therefore, a transcendental substratum lies atthe foundation of the complete determinationof things—a substratum which is to form thefund from which all possible predicates ofthings are to be supplied, this substratum can-not be anything else than the idea of a sum-total of reality (omnitudo realitatis). In thisview, negations are nothing but limitations—a

term which could not, with propriety, be ap-plied to them, if the unlimited (the all) did notform the true basis of our conception.

This conception of a sum-total of reality is theconception of a thing in itself, regarded as com-pletely determined; and the conception of anens realissimum is the conception of an indivi-dual being, inasmuch as it is determined bythat predicate of all possible contradictory pre-dicates, which indicates and belongs to being. Itis, therefore, a transcendental ideal whichforms the basis of the complete determinationof everything that exists, and is the highest ma-terial condition of its possibility—a conditionon which must rest the cogitation of all objectswith respect to their content. Nay, more, thisideal is the only proper ideal of which thehuman mind is capable; because in this casealone a general conception of a thing is comple-tely determined by and through itself, and cog-nized as the representation of an individuum.

The logical determination of a conception isbased upon a disjunctive syllogism, the majorof which contains the logical division of theextent of a general conception, the minor limitsthis extent to a certain part, while the conclu-sion determines the conception by this part.The general conception of a reality cannot bedivided a priori, because, without the aid ofexperience, we cannot know any determinatekinds of reality, standing under the former asthe genus. The transcendental principle of thecomplete determination of all things is therefo-re merely the representation of the sum-total ofall reality; it is not a conception which is thegenus of all predicates under itself, but onewhich comprehends them all within itself. Thecomplete determination of a thing is conse-quently based upon the limitation of this totalof reality, so much being predicated of thething, while all that remains over is excluded—a procedure which is in exact agreement withthat of the disjunctive syllogism and the deter-

mination of the objects in the conclusion by oneof the members of the division. It follows thatreason, in laying the transcendental ideal at thefoundation of its determination of all possiblethings, takes a course in exact analogy with thatwhich it pursues in disjunctive syllogisms—aproposition which formed the basis of the sys-tematic division of all transcendental ideas,according to which they are produced in com-plete parallelism with the three modes of syllo-gistic reasoning employed by the human mind.

It is self-evident that reason, in cogitating thenecessary complete determination of things,does not presuppose the existence of a beingcorresponding to its ideal, but merely the ideaof the ideal- for the purpose of deducing fromthe unconditional totality of complete determi-nation, The ideal is therefore the prototype ofall things, which, as defective copies (ectypa),receive from it the material of their possibility,and approximate to it more or less, though it is

impossible that they can ever attain to its per-fection.

The possibility of things must therefore be re-garded as derived- except that of the thingwhich contains in itself all reality, which mustbe considered to be primitive and original. Forall negations- and they are the only predicatesby means of which all other things can be dis-tinguished from the ens realissimum—are merelimitations of a greater and a higher—nay, thehighest reality; and they consequently presup-pose this reality, and are, as regards their con-tent, derived from it. The manifold nature ofthings is only an infinitely various mode oflimiting the conception of the highest reality,which is their common substratum; just as allfigures are possible only as different modes oflimiting infinite space. The object of the ideal ofreason—an object existing only in reason it-self—is also termed the primal being (ens ori-ginarium); as having no existence superior to

him, the supreme being (ens summum); and asbeing the condition of all other beings, whichrank under it, the being of all beings (ens en-tium). But none of these terms indicate the ob-jective relation of an actually existing object toother things, but merely that of an idea to con-ceptions; and all our investigations into thissubject still leave us in perfect uncertainty withregard to the existence of this being.

A primal being cannot be said to consist of ma-ny other beings with an existence which is de-rivative, for the latter presuppose the former,and therefore cannot be constitutive parts of it.It follows that the ideal of the primal beingmust be cogitated as simple.

The deduction of the possibility of all otherthings from this primal being cannot, strictlyspeaking, be considered as a limitation, or as akind of division of its reality; for this would beregarding the primal being as a mere aggrega-te—which has been shown to be impossible,

although it was so represented in our firstrough sketch. The highest reality must be re-garded rather as the ground than as the sum-total of the possibility of all things, and the ma-nifold nature of things be based, not upon thelimitation of the primal being itself, but uponthe complete series of effects which flow fromit. And thus all our powers of sense, as well asall phenomenal reality, phenomenal reality,may be with propriety regarded as belongingto this series of effects, while they could nothave formed parts of the idea, considered as anaggregate. Pursuing this track, and hypostati-zing this idea, we shall find ourselves authori-zed to determine our notion of the SupremeBeing by means of the mere conception of ahighest reality, as one, simple, all-sufficient,eternal, and so on—in one word, to determineit in its unconditioned completeness by the aidof every possible predicate. The conception ofsuch a being is the conception of God in itstranscendental sense, and thus the ideal of pure

reason is the object-matter of a transcendentaltheology.

But, by such an employment of the transcen-dental idea, we should be over stepping thelimits of its validity and purpose. For reasonplaced it, as the conception of all reality, at thebasis of the complete determination of things,without requiring that this conception be re-garded as the conception of an objective exis-tence. Such an existence would be purely ficti-tious, and the hypostatizing of the content ofthe idea into an ideal, as an individual being, isa step perfectly unauthorized. Nay, more, weare not even called upon to assume the possibi-lity of such an hypothesis, as none of the de-ductions drawn from such an ideal would af-fect the complete determination of things ingeneral—for the sake of which alone is the ideanecessary.

It is not sufficient to circumscribe the procedureand the dialectic of reason; we must also en-

deavour to discover the sources of this dialec-tic, that we may have it in our power to give arational explanation of this illusion, as a phe-nomenon of the human mind. For the ideal, ofwhich we are at present speaking, is based, notupon an arbitrary, but upon a natural, idea. Thequestion hence arises: How happens it that rea-son regards the possibility of all things as de-duced from a single possibility, that, to wit, ofthe highest reality, and presupposes this asexisting in an individual and primal being?

The answer is ready; it is at once presented bythe procedure of transcendental analytic. Thepossibility of sensuous objects is a relation ofthese objects to thought, in which something(the empirical form) may be cogitated a priori;while that which constitutes the matter—thereality of the phenomenon (that element whichcorresponds to sensation)—must be given fromwithout, as otherwise it could not even be cogi-tated by, nor could its possibility be presentable

to the mind. Now, a sensuous object is comple-tely determined, when it has been comparedwith all phenomenal predicates, and represen-ted by means of these either positively or nega-tively. But, as that which constitutes the thingitself—the real in a phenomenon, must be gi-ven, and that, in which the real of all phenome-na is given, is experience, one, sole, and all-embracing- the material of the possibility of allsensuous objects must be presupposed as givenin a whole, and it is upon the limitation of thiswhole that the possibility of all empirical ob-jects, their distinction from each other and theircomplete determination, are based. Now, noother objects are presented to us besides sen-suous objects, and these can be given only inconnection with a possible experience; it fo-llows that a thing is not an object to us, unless itpresupposes the whole or sum-total of empiri-cal reality as the condition of its possibility.Now, a natural illusion leads us to consider thisprinciple, which is valid only of sensuous ob-

jects, as valid with regard to things in general.And thus we are induced to hold the empiricalprinciple of our conceptions of the possibility ofthings, as phenomena, by leaving out this limi-tative condition, to be a transcendental princi-ple of the possibility of things in general.

We proceed afterwards to hypostatize this ideaof the sum-total of all reality, by changing thedistributive unity of the empirical exercise ofthe understanding into the collective unity ofan empirical whole—a dialectical illusion, andby cogitating this whole or sum of experienceas an individual thing, containing in itself allempirical reality. This individual thing or beingis then, by means of the above-mentionedtranscendental subreption, substituted for ournotion of a thing which stands at the head ofthe possibility of all things, the real conditionsof whose complete determination it presents.*

[*Footnote: This ideal of the ens realissimum—although merely a mental representation—is

first objectivized, that is, has an objective exis-tence attributed to it, then hypostatized, andfinally, by the natural progress of reason to thecompletion of unity, personified, as we shallshow presently. For the regulative unity of ex-perience is not based upon phenomena them-selves, but upon the connection of the varietyof phenomena by the understanding in a cons-ciousness, and thus the unity of the supremereality and the complete determinability of allthings, seem to reside in a supreme understan-ding, and, consequently, in a conscious intelli-gence.]

SECTION III. Of the Arguments employedby Speculative Reason in Proof of the Exis-tence of a Supreme Being.

Notwithstanding the pressing necessity whichreason feels, to form some presupposition that

shall serve the understanding as a proper basisfor the complete determination of its concep-tions, the idealistic and factitious nature of sucha presupposition is too evident to allow reasonfor a moment to persuade itself into a belief ofthe objective existence of a mere creation of itsown thought. But there are other considerationswhich compel reason to seek out some restingplace in the regress from the conditioned to theunconditioned, which is not given as an actualexistence from the mere conception of it, alt-hough it alone can give completeness to theseries of conditions. And this is the naturalcourse of every human reason, even of the mostuneducated, although the path at first entered itdoes not always continue to follow. It does notbegin from conceptions, but from common ex-perience, and requires a basis in actual existen-ce. But this basis is insecure, unless it restsupon the immovable rock of the absolutely ne-cessary. And this foundation is itself unworthyof trust, if it leave under and above it empty

space, if it do not fill all, and leave no room fora why or a wherefore, if it be not, in one word,infinite in its reality.

If we admit the existence of some one thing,whatever it may be, we must also admit thatthere is something which exists necessarily. Forwhat is contingent exists only under the condi-tion of some other thing, which is its cause; andfrom this we must go on to conclude the exis-tence of a cause which is not contingent, andwhich consequently exists necessarily and un-conditionally. Such is the argument by whichreason justifies its advances towards a primalbeing.

Now reason looks round for the conception of abeing that may be admitted, without inconsis-tency, to be worthy of the attribute of absolutenecessity, not for the purpose of inferring apriori, from the conception of such a being, itsobjective existence (for if reason allowed itselfto take this course, it would not require a basis

in given and actual existence, but merely thesupport of pure conceptions), but for the pur-pose of discovering, among all our conceptionsof possible things, that conception which pos-sesses no element inconsistent with the idea ofabsolute necessity. For that there must be someabsolutely necessary existence, it regards as atruth already established. Now, if it can removeevery existence incapable of supporting theattribute of absolute necessity, excepting one—this must be the absolutely necessary being,whether its necessity is comprehensible by us,that is, deducible from the conception of it alo-ne, or not.

Now that, the conception of which contains atherefore to every wherefore, which is not de-fective in any respect whatever, which is all-sufficient as a condition, seems to be the beingof which we can justly predicate absolute ne-cessity—for this reason, that, possessing theconditions of all that is possible, it does not and

cannot itself require any condition. And thus itsatisfies, in one respect at least, the require-ments of the conception of absolute necessity.In this view, it is superior to all other concep-tions, which, as deficient and incomplete, donot possess the characteristic of independenceof all higher conditions. It is true that we can-not infer from this that what does not containin itself the supreme and complete condition—the condition of all other things—must possessonly a conditioned existence; but as little canwe assert the contrary, for this supposed beingdoes not possess the only characteristic whichcan enable reason to cognize by means of an apriori conception the unconditioned and neces-sary nature of its existence.

The conception of an ens realissimum is thatwhich best agrees with the conception of anunconditioned and necessary being. The formerconception does not satisfy all the requirementsof the latter; but we have no choice, we are

obliged to adhere to it, for we find that we can-not do without the existence of a necessarybeing; and even although we admit it, we findit out of our power to discover in the wholesphere of possibility any being that can advan-ce well-grounded claims to such a distinction.

The following is, therefore, the natural courseof human reason. It begins by persuading itselfof the existence of some necessary being. In thisbeing it recognizes the characteristics of uncon-ditioned existence. It then seeks the conceptionof that which is independent of all conditions,and finds it in that which is itself the sufficientcondition of all other things—in other words, inthat which contains all reality. But the unlimi-ted all is an absolute unity, and is conceived bythe mind as a being one and supreme; and thusreason concludes that the Supreme Being, asthe primal basis of all things, possesses an exis-tence which is absolutely necessary.

This conception must be regarded as in somedegree satisfactory, if we admit the existence ofa necessary being, and consider that there existsa necessity for a definite and final answer tothese questions. In such a case, we cannot makea better choice, or rather we have no choice atall, but feel ourselves obliged to declare in fa-vour of the absolute unity of complete reality,as the highest source of the possibility ofthings. But if there exists no motive for comingto a definite conclusion, and we may leave thequestion unanswered till we have fully weig-hed both sides—in other words, when we aremerely called upon to decide how much wehappen to know about the question, and howmuch we merely flatter ourselves that weknow- the above conclusion does not appear tobe so great advantage, but, on the contrary,seems defective in the grounds upon which it issupported.

For, admitting the truth of all that has beensaid, that, namely, the inference from a givenexistence (my own, for example) to the existen-ce of an unconditioned and necessary being isvalid and unassailable; that, in the second pla-ce, we must consider a being which contains allreality, and consequently all the conditions ofother things, to be absolutely unconditioned;and admitting too, that we have thus discove-red the conception of a thing to which may beattributed, without inconsistency, absolute ne-cessity—it does not follow from all this that theconception of a limited being, in which the su-preme reality does not reside, is therefore in-compatible with the idea of absolute necessity.For, although I do not discover the element ofthe unconditioned in the conception of such abeing—an element which is manifestly existentin the sum-total of all conditions—I am notentitled to conclude that its existence is therefo-re conditioned; just as I am not entitled to af-firm, in a hypothetical syllogism, that where a

certain condition does not exist (in the present,completeness, as far as pure conceptions areconcerned), the conditioned does not exist eit-her. On the contrary, we are free to consider alllimited beings as likewise unconditionally ne-cessary, although we are unable to infer thisfrom the general conception which we have ofthem. Thus conducted, this argument is inca-pable of giving us the least notion of the pro-perties of a necessary being, and must be inevery respect without result.

This argument continues, however, to possess aweight and an authority, which, in spite of itsobjective insufficiency, it has never been dives-ted of. For, granting that certain responsibilitieslie upon us, which, as based on the ideas ofreason, deserve to be respected and submittedto, although they are incapable of a real or prac-tical application to our nature, or, in otherwords, would be responsibilities without moti-ves, except upon the supposition of a Supreme

Being to give effect and influence to the practi-cal laws: in such a case we should be bound toobey our conceptions, which, although objecti-vely insufficient, do, according to the standardof reason, preponderate over and are superiorto any claims that may be advanced from anyother quarter. The equilibrium of doubt wouldin this case be destroyed by a practical addi-tion; indeed, Reason would be compelled tocondemn herself, if she refused to comply withthe demands of the judgement, no superior towhich we know—however defective her un-derstanding of the grounds of these demandsmight be.

This argument, although in fact transcendental,inasmuch as it rests upon the intrinsic insuffi-ciency of the contingent, is so simple and natu-ral, that the commonest understanding can ap-preciate its value. We see things around uschange, arise, and pass away; they, or their con-dition, must therefore have a cause. The same

demand must again be made of the cause it-self—as a datum of experience. Now it is natu-ral that we should place the highest causalityjust where we place supreme causality, in thatbeing, which contains the conditions of all pos-sible effects, and the conception of which is sosimple as that of an all-embracing reality. Thishighest cause, then, we regard as absolutelynecessary, because we find it absolutely neces-sary to rise to it, and do not discover any rea-son for proceeding beyond it. Thus, among allnations, through the darkest polytheism glim-mer some faint sparks of monotheism, to whichthese idolaters have been led, not from reflec-tion and profound thought, but by the studyand natural progress of the common unders-tanding.

There are only three modes of proving the exis-tence of a Deity, on the grounds of speculativereason.

All the paths conducting to this end begin eit-her from determinate experience and the pecu-liar constitution of the world of sense, and rise,according to the laws of causality, from it to thehighest cause existing apart from the world—orfrom a purely indeterminate experience, that is,some empirical existence—or abstraction ismade of all experience, and the existence of asupreme cause is concluded from a priori con-ceptions alone. The first is the physico-theological argument, the second the cosmolo-gical, the third the ontological. More there arenot, and more there cannot be.

I shall show it is as unsuccessful on the onepath—the empirical- as on the other—the trans-cendental, and that it stretches its wings invain, to soar beyond the world of sense by themere might of speculative thought. As regardsthe order in which we must discuss those ar-guments, it will be exactly the reverse of that inwhich reason, in the progress of its develop-

ment, attains to them—the order in which theyare placed above. For it will be made manifestto the reader that, although experience presentsthe occasion and the starting-point, it is thetranscendental idea of reason which guides it inits pilgrimage and is the goal of all its struggles.I shall therefore begin with an examination ofthe transcendental argument, and afterwardsinquire what additional strength has accrued tothis mode of proof from the addition of theempirical element.

SECTION IV. Of the Impossibility of anOntological Proof of the Existence of God.

It is evident from what has been said that theconception of an absolutely necessary being is amere idea, the objective reality of which is farfrom being established by the mere fact that itis a need of reason. On the contrary, this idea

serves merely to indicate a certain unattainableperfection, and rather limits the operationsthan, by the presentation of new objects, ex-tends the sphere of the understanding. But astrange anomaly meets us at the very threshold;for the inference from a given existence in ge-neral to an absolutely necessary existenceseems to be correct and unavoidable, while theconditions of the understanding refuse to aidus in forming any conception of such a being.

Philosophers have always talked of an absolu-tely necessary being, and have neverthelessdeclined to take the trouble of conceivingwhether—and how—a being of this nature iseven cogitable, not to mention that its existenceis actually demonstrable. A verbal definition ofthe conception is certainly easy enough: it issomething the non-existence of which is impos-sible. But does this definition throw any lightupon the conditions which render it impossibleto cogitate the non-existence of a thing—

conditions which we wish to ascertain, that wemay discover whether we think anything in theconception of such a being or not? For the merefact that I throw away, by means of the wordunconditioned, all the conditions which theunderstanding habitually requires in order toregard anything as necessary, is very far frommaking clear whether by means of the concep-tion of the unconditionally necessary I think ofsomething, or really of nothing at all.

Nay, more, this chance-conception, now beco-me so current, many have endeavoured to ex-plain by examples which seemed to render anyinquiries regarding its intelligibility quite need-less. Every geometrical proposition—a trianglehas three angles—it was said, is absolutely ne-cessary; and thus people talked of an objectwhich lay out of the sphere of our understan-ding as if it were perfectly plain what the con-ception of such a being meant.

All the examples adduced have been drawn,without exception, from judgements, and notfrom things. But the unconditioned necessity ofa judgement does not form the absolute neces-sity of a thing. On the contrary, the absolutenecessity of a judgement is only a conditionednecessity of a thing, or of the predicate in a jud-gement. The proposition above-mentioneddoes not enounce that three angles necessarilyexist, but, upon condition that a triangle exists,three angles must necessarily exist—in it. Andthus this logical necessity has been the sourceof the greatest delusions. Having formed an apriori conception of a thing, the content ofwhich was made to embrace existence, we be-lieved ourselves safe in concluding that, becau-se existence belongs necessarily to the object ofthe conception (that is, under the condition ofmy positing this thing as given), the existenceof the thing is also posited necessarily, and thatit is therefore absolutely necessary—merely

because its existence has been cogitated in theconception.

If, in an identical judgement, I annihilate thepredicate in thought, and retain the subject, acontradiction is the result; and hence I say, theformer belongs necessarily to the latter. But if Isuppress both subject and predicate in thought,no contradiction arises; for there is nothing atall, and therefore no means of forming a con-tradiction. To suppose the existence of a trian-gle and not that of its three angles, is self-contradictory; but to suppose the non-existenceof both triangle and angles is perfectly admissi-ble. And so is it with the conception of an abso-lutely necessary being. Annihilate its existencein thought, and you annihilate the thing itselfwith all its predicates; how then can there beany room for contradiction? Externally, there isnothing to give rise to a contradiction, for athing cannot be necessary externally; nor inter-nally, for, by the annihilation or suppression of

the thing itself, its internal properties are alsoannihilated. God is omnipotent—that is a ne-cessary judgement. His omnipotence cannot bedenied, if the existence of a Deity is posited—the existence, that is, of an infinite being, thetwo conceptions being identical. But when yousay, God does not exist, neither omnipotencenor any other predicate is affirmed; they mustall disappear with the subject, and in this jud-gement there cannot exist the least self-contradiction.

You have thus seen that when the predicate ofa judgement is annihilated in thought alongwith the subject, no internal contradiction canarise, be the predicate what it may. There is nopossibility of evading the conclusion—you findyourselves compelled to declare: There are cer-tain subjects which cannot be annihilated inthought. But this is nothing more than saying:There exist subjects which are absolutely neces-sary—the very hypothesis which you are called

upon to establish. For I find myself unable toform the slightest conception of a thing whichwhen annihilated in thought with all its predi-cates, leaves behind a contradiction; and con-tradiction is the only criterion of impossibilityin the sphere of pure a priori conceptions.

Against these general considerations, the justiceof which no one can dispute, one argument isadduced, which is regarded as furnishing asatisfactory demonstration from the fact. It isaffirmed that there is one and only one concep-tion, in which the non-being or annihilation ofthe object is self-contradictory, and this is theconception of an ens realissimum. It possesses,you say, all reality, and you feel yourselvesjustified in admitting the possibility of such abeing. (This I am willing to grant for the pre-sent, although the existence of a conceptionwhich is not self-contradictory is far from beingsufficient to prove the possibility of an object.)*Now the notion of all reality embraces in it that

of existence; the notion of existence lies, there-fore, in the conception of this possible thing. Ifthis thing is annihilated in thought, the internalpossibility of the thing is also annihilated,which is self-contradictory.

[*Footnote: A conception is always possible, if itis not self-contradictory. This is the logical cri-terion of possibility, distinguishing the object ofsuch a conception from the nihil negativum.But it may be, notwithstanding, an empty con-ception, unless the objective reality of this synt-hesis, but which it is generated, is demonstra-ted; and a proof of this kind must be basedupon principles of possible experience, and notupon the principle of analysis or contradiction.This remark may be serviceable as a warningagainst concluding, from the possibility of aconception—which is logical—the possibility ofa thing—which is real.]

I answer: It is absurd to introduce—under wha-tever term disguised—into the conception of a

thing, which is to be cogitated solely in referen-ce to its possibility, the conception of its exis-tence. If this is admitted, you will have appa-rently gained the day, but in reality haveenounced nothing but a mere tautology. I ask,is the proposition, this or that thing (which I amadmitting to be possible) exists, an analytical ora synthetical proposition? If the former, there isno addition made to the subject of your thoughtby the affirmation of its existence; but then theconception in your minds is identical with thething itself, or you have supposed the existenceof a thing to be possible, and then inferred itsexistence from its internal possibility—which isbut a miserable tautology. The word reality inthe conception of the thing, and the word exis-tence in the conception of the predicate, willnot help you out of the difficulty. For, suppo-sing you were to term all positing of a thingreality, you have thereby posited the thing withall its predicates in the conception of the subjectand assumed its actual existence, and this you

merely repeat in the predicate. But if you con-fess, as every reasonable person must, that eve-ry existential proposition is synthetical, howcan it be maintained that the predicate of exis-tence cannot be denied without contradic-tion?—a property which is the characteristic ofanalytical propositions, alone.

I should have a reasonable hope of putting anend for ever to this sophistical mode of argu-mentation, by a strict definition of the concep-tion of existence, did not my own experienceteach me that the illusion arising from our con-founding a logical with a real predicate (a pre-dicate which aids in the determination of athing) resists almost all the endeavours of ex-planation and illustration. A logical predicatemay be what you please, even the subject maybe predicated of itself; for logic pays no regardto the content of a judgement. But the determi-nation of a conception is a predicate, which

adds to and enlarges the conception. It mustnot, therefore, be contained in the conception.

Being is evidently not a real predicate, that is, aconception of something which is added to theconception of some other thing. It is merely thepositing of a thing, or of certain determinationsin it. Logically, it is merely the copula of a jud-gement. The proposition, God is omnipotent,contains two conceptions, which have a certainobject or content; the word is, is no additionalpredicate—it merely indicates the relation ofthe predicate to the subject. Now, if I take thesubject (God) with all its predicates (omnipo-tence being one), and say: God is, or, There is aGod, I add no new predicate to the conceptionof God, I merely posit or affirm the existence ofthe subject with all its predicates—I posit theobject in relation to my conception. The contentof both is the same; and there is no additionmade to the conception, which expresses mere-ly the possibility of the object, by my cogitating

the object—in the expression, it is—as absolute-ly given or existing. Thus the real contains nomore than the possible. A hundred real dollarscontain no more than a hundred possible do-llars. For, as the latter indicate the conception,and the former the object, on the suppositionthat the content of the former was greater thanthat of the latter, my conception would not bean expression of the whole object, and wouldconsequently be an inadequate conception of it.But in reckoning my wealth there may be saidto be more in a hundred real dollars than in ahundred possible dollars—that is, in the mereconception of them. For the real object—thedollars—is not analytically contained in myconception, but forms a synthetical addition tomy conception (which is merely a determina-tion of my mental state), although this objectivereality—this existence—apart from my concep-tions, does not in the least degree increase theaforesaid hundred dollars.

By whatever and by whatever number of pre-dicates—even to the complete determination ofit—I may cogitate a thing, I do not in the leastaugment the object of my conception by theaddition of the statement: This thing exists.Otherwise, not exactly the same, but somethingmore than what was cogitated in my concep-tion, would exist, and I could not affirm thatthe exact object of my conception had real exis-tence. If I cogitate a thing as containing all mo-des of reality except one, the mode of realitywhich is absent is not added to the conceptionof the thing by the affirmation that the thingexists; on the contrary, the thing exists—if itexist at all—with the same defect as that cogita-ted in its conception; otherwise not that whichwas cogitated, but something different, exists.Now, if I cogitate a being as the highest reality,without defect or imperfection, the questionstill remains—whether this being exists or not?For, although no element is wanting in the pos-sible real content of my conception, there is a

defect in its relation to my mental state, that is,I am ignorant whether the cognition of the ob-ject indicated by the conception is possible aposteriori. And here the cause of the presentdifficulty becomes apparent. If the questionregarded an object of sense merely, it would beimpossible for me to confound the conceptionwith the existence of a thing. For the conceptionmerely enables me to cogitate an object as ac-cording with the general conditions of expe-rience; while the existence of the object permitsme to cogitate it as contained in the sphere ofactual experience. At the same time, this con-nection with the world of experience does notin the least augment the conception, although apossible perception has been added to the ex-perience of the mind. But if we cogitate existen-ce by the pure category alone, it is not to bewondered at, that we should find ourselvesunable to present any criterion sufficient todistinguish it from mere possibility.

Whatever be the content of our conception ofan object, it is necessary to go beyond it, if wewish to predicate existence of the object. In thecase of sensuous objects, this is attained bytheir connection according to empirical lawswith some one of my perceptions; but there isno means of cognizing the existence of objectsof pure thought, because it must be cognizedcompletely a priori. But all our knowledge ofexistence (be it immediately by perception, orby inferences connecting some object with aperception) belongs entirely to the sphere ofexperience—which is in perfect unity with it-self; and although an existence out of this sphe-re cannot be absolutely declared to be impossi-ble, it is a hypothesis the truth of which wehave no means of ascertaining.

The notion of a Supreme Being is in many res-pects a highly useful idea; but for the very rea-son that it is an idea, it is incapable of enlargingour cognition with regard to the existence of

things. It is not even sufficient to instruct us asto the possibility of a being which we do notknow to exist. The analytical criterion of possi-bility, which consists in the absence of contra-diction in propositions, cannot be denied it. Butthe connection of real properties in a thing is asynthesis of the possibility of which an a priorijudgement cannot be formed, because theserealities are not presented to us specifically;and even if this were to happen, a judgementwould still be impossible, because the criterionof the possibility of synthetical cognitions mustbe sought for in the world of experience, towhich the object of an idea cannot belong. Andthus the celebrated Leibnitz has utterly failed inhis attempt to establish upon a priori groundsthe possibility of this sublime ideal being.

The celebrated ontological or Cartesian argu-ment for the existence of a Supreme Being istherefore insufficient; and we may as well hopeto increase our stock of knowledge by the aid of

mere ideas, as the merchant to augment hiswealth by the addition of noughts to his cashaccount.

SECTION V. Of the Impossibility of aCosmological Proof of the Existence of God.

It was by no means a natural course of procee-ding, but, on the contrary, an invention entirelydue to the subtlety of the schools, to attempt todraw from a mere idea a proof of the existenceof an object corresponding to it. Such a coursewould never have been pursued, were it not forthat need of reason which requires it to suppo-se the existence of a necessary being as a basisfor the empirical regress, and that, as this ne-cessity must be unconditioned and a priori,reason is bound to discover a conception whichshall satisfy, if possible, this requirement, andenable us to attain to the a priori cognition of

such a being. This conception was thought tobe found in the idea of an ens realissimum, andthus this idea was employed for the attainmentof a better defined knowledge of a necessarybeing, of the existence of which we were con-vinced, or persuaded, on other grounds. Thusreason was seduced from her natural courage;and, instead of concluding with the conceptionof an ens realissimum, an attempt was made tobegin with it, for the purpose of inferring fromit that idea of a necessary existence which itwas in fact called in to complete. Thus arosethat unfortunate ontological argument, whichneither satisfies the healthy common sense ofhumanity, nor sustains the scientific examina-tion of the philosopher.

The cosmological proof, which we are about toexamine, retains the connection between abso-lute necessity and the highest reality; but, ins-tead of reasoning from this highest reality to anecessary existence, like the preceding argu-

ment, it concludes from the given unconditio-ned necessity of some being its unlimited reali-ty. The track it pursues, whether rational orsophistical, is at least natural, and not only goesfar to persuade the common understanding,but shows itself deserving of respect from thespeculative intellect; while it contains, at thesame time, the outlines of all the argumentsemployed in natural theology—argumentswhich always have been, and still will be, inuse and authority. These, however adorned,and hid under whatever embellishments ofrhetoric and sentiment, are at bottom identicalwith the arguments we are at present to dis-cuss. This proof, termed by Leibnitz the argu-mentum a contingentia mundi, I shall now laybefore the reader, and subject to a strict exami-nation.

It is framed in the following manner: If somet-hing exists, an absolutely necessary being mustlikewise exist. Now I, at least, exist. Consequen-

tly, there exists an absolutely necessary being.The minor contains an experience, the majorreasons from a general experience to the exis-tence of a necessary being.* Thus this argumentreally begins at experience, and is not comple-tely a priori, or ontological. The object of allpossible experience being the world, it is calledthe cosmological proof. It contains no referenceto any peculiar property of sensuous objects, bywhich this world of sense might be distinguis-hed from other possible worlds; and in thisrespect it differs from the physico-theologicalproof, which is based upon the consideration ofthe peculiar constitution of our sensuousworld.

[*Footnote: This inference is too well known torequire more detailed discussion. It is basedupon the spurious transcendental law of causa-lity, that everything which is contingent has acause, which, if itself contingent, must alsohave a cause; and so on, till the series of subor-

dinated causes must end with an absolutelynecessary cause, without which it would notpossess completeness.]

The proof proceeds thus: A necessary being canbe determined only in one way, that is, it can bedetermined by only one of all possible opposedpredicates; consequently, it must be completelydetermined in and by its conception. But thereis only a single conception of a thing possible,which completely determines the thing a priori:that is, the conception of the ens realissimum. Itfollows that the conception of the ens realissi-mum is the only conception by and in whichwe can cogitate a necessary being. Consequen-tly, a Supreme Being necessarily exists.

In this cosmological argument are assembled somany sophistical propositions that speculativereason seems to have exerted in it all her dialec-tical skill to produce a transcendental illusionof the most extreme character. We shall post-pone an investigation of this argument for the

present, and confine ourselves to exposing thestratagem by which it imposes upon us an oldargument in a new dress, and appeals to theagreement of two witnesses, the one with thecredentials of pure reason, and the other withthose of empiricism; while, in fact, it is only theformer who has changed his dress and voice,for the purpose of passing himself off for anadditional witness. That it may possess a securefoundation, it bases its conclusions upon expe-rience, and thus appears to be completely dis-tinct from the ontological argument, which pla-ces its confidence entirely in pure a priori con-ceptions. But this experience merely aids reasonin making one step—to the existence of a neces-sary being. What the properties of this beingare cannot be learned from experience; andtherefore reason abandons it altogether, andpursues its inquiries in the sphere of pure con-ception, for the purpose of discovering whatthe properties of an absolutely necessary beingought to be, that is, what among all possible

things contain the conditions (requisita) of ab-solute necessity. Reason believes that it hasdiscovered these requisites in the conception ofan ens realissimum—and in it alone, and henceconcludes: The ens realissimum is an absolutelynecessary being. But it is evident that reasonhas here presupposed that the conception of anens realissimum is perfectly adequate to theconception of a being of absolute necessity, thatis, that we may infer the existence of the latterfrom that of the former—a proposition whichformed the basis of the ontological argument,and which is now employed in the support ofthe cosmological argument, contrary to thewish and professions of its inventors. For theexistence of an absolutely necessary being isgiven in conceptions alone. But if I say: "Theconception of the ens realissimum is a concep-tion of this kind, and in fact the only conceptionwhich is adequate to our idea of a necessarybeing," I am obliged to admit, that the lattermay be inferred from the former. Thus it is

properly the ontological argument which figu-res in the cosmological, and constitutes thewhole strength of the latter; while the spuriousbasis of experience has been of no further usethan to conduct us to the conception of absolutenecessity, being utterly insufficient to demons-trate the presence of this attribute in any de-terminate existence or thing. For when we pro-pose to ourselves an aim of this character, wemust abandon the sphere of experience, andrise to that of pure conceptions, which we exa-mine with the purpose of discovering whetherany one contains the conditions of the possibili-ty of an absolutely necessary being. But if thepossibility of such a being is thus demonstra-ted, its existence is also proved; for we maythen assert that, of all possible beings there isone which possesses the attribute of necessity—in other words, this being possesses an absolu-tely necessary existence.

All illusions in an argument are more easilydetected when they are presented in the formalmanner employed by the schools, which wenow proceed to do.

If the proposition: "Every absolutely necessarybeing is likewise an ens realissimum," is correct(and it is this which constitutes the nervus pro-bandi of the cosmological argument), it must,like all affirmative judgements, be capable ofconversion—the conversio per accidens, atleast. It follows, then, that some entia realissimaare absolutely necessary beings. But no ensrealissimum is in any respect different fromanother, and what is valid of some is valid ofall. In this present case, therefore, I may employsimple conversion, and say: "Every ens realis-simum is a necessary being." But as this propo-sition is determined a priori by the conceptionscontained in it, the mere conception of an ensrealissimum must possess the additional attri-bute of absolute necessity. But this is exactly

what was maintained in the ontological argu-ment, and not recognized by the cosmological,although it formed the real ground of its dis-guised and illusory reasoning.

Thus the second mode employed by speculati-ve reason of demonstrating the existence of aSupreme Being, is not only, like the first, illuso-ry and inadequate, but possesses the additionalblemish of an ignoratio elenchi—professing toconduct us by a new road to the desired goal,but bringing us back, after a short circuit, to theold path which we had deserted at its call.

I mentioned above that this cosmological ar-gument contains a perfect nest of dialecticalassumptions, which transcendental criticismdoes not find it difficult to expose and todissipate. I shall merely enumerate these,leaving it to the reader, who must by this timebe well practised in such matters, to investigatethe fallacies residing therein.

The following fallacies, for example, are disco-verable in this mode of proof: 1. The transcen-dental principle: "Everything that is contingentmust have a cause"—a principle without signi-ficance, except in the sensuous world. For thepurely intellectual conception of the contingentcannot produce any synthetical proposition,like that of causality, which is itself withoutsignificance or distinguishing characteristicexcept in the phenomenal world. But in thepresent case it is employed to help us beyondthe limits of its sphere. 2. "From the impossibili-ty of an infinite ascending series of causes inthe world of sense a first cause is inferred"; aconclusion which the principles of the em-ployment of reason do not justify even in thesphere of experience, and still less when anattempt is made to pass the limits of this sphe-re. 3. Reason allows itself to be satisfied uponinsufficient grounds, with regard to the com-pletion of this series. It removes all conditions(without which, however, no conception of

Necessity can take place); and, as after this it isbeyond our power to form any other concep-tions, it accepts this as a completion of the con-ception it wishes to form of the series. 4. Thelogical possibility of a conception of the total ofreality (the criterion of this possibility being theabsence of contradiction) is confounded withthe transcendental, which requires a principleof the practicability of such a synthesis—a prin-ciple which again refers us to the world of ex-perience. And so on.

The aim of the cosmological argument is toavoid the necessity of proving the existence of anecessary being priori from mere conceptions—a proof which must be ontological, and ofwhich we feel ourselves quite incapable. Withthis purpose, we reason from an actual existen-ce—an experience in general, to an absolutelynecessary condition of that existence. It is inthis case unnecessary to demonstrate its possi-bility. For after having proved that it exists, the

question regarding its possibility is super-fluous. Now, when we wish to define morestrictly the nature of this necessary being, wedo not look out for some being the conceptionof which would enable us to comprehend thenecessity of its being—for if we could do this,an empirical presupposition would be unneces-sary; no, we try to discover merely the negativecondition (conditio sine qua non), withoutwhich a being would not be absolutely necessa-ry. Now this would be perfectly admissible inevery sort of reasoning, from a consequence toits principle; but in the present case it unfortu-nately happens that the condition of absolutenecessity can be discovered in but a singlebeing, the conception of which must conse-quently contain all that is requisite for demons-trating the presence of absolute necessity, andthus entitle me to infer this absolute necessity apriori. That is, it must be possible to reasonconversely, and say: The thing, to which theconception of the highest reality belongs, is

absolutely necessary. But if I cannot reasonthus—and I cannot, unless I believe in the suffi-ciency of the ontological argument—I find in-surmountable obstacles in my new path, andam really no farther than the point from whichI set out. The conception of a Supreme Beingsatisfies all questions a priori regarding theinternal determinations of a thing, and is forthis reason an ideal without equal or parallel,the general conception of it indicating it as atthe same time an ens individuum among allpossible things. But the conception does notsatisfy the question regarding its existence—which was the purpose of all our inquiries; and,although the existence of a necessary beingwere admitted, we should find it impossible toanswer the question: What of all things in theworld must be regarded as such?

It is certainly allowable to admit the existenceof an all-sufficient being—a cause of all possi-ble effects—for the purpose of enabling reason

to introduce unity into its mode and grounds ofexplanation with regard to phenomena. But toassert that such a being necessarily exists, is nolonger the modest enunciation of an admissiblehypothesis, but the boldest declaration of anapodeictic certainty; for the cognition of thatwhich is absolutely necessary must itself pos-sess that character.

The aim of the transcendental ideal formed bythe mind is either to discover a conceptionwhich shall harmonize with the idea of absolu-te necessity, or a conception which shall containthat idea. If the one is possible, so is the other;for reason recognizes that alone as absolutelynecessary which is necessary from its concep-tion. But both attempts are equally beyond ourpower—we find it impossible to satisfy the un-derstanding upon this point, and as impossibleto induce it to remain at rest in relation to thisincapacity.

Unconditioned necessity, which, as the ultimatesupport and stay of all existing things, is anindispensable requirement of the mind, is anabyss on the verge of which human reasontrembles in dismay. Even the idea of eternity,terrible and sublime as it is, as depicted byHaller, does not produce upon the mental vi-sion such a feeling of awe and terror; for, alt-hough it measures the duration of things, itdoes not support them. We cannot bear, norcan we rid ourselves of the thought that abeing, which we regard as the greatest of allpossible existences, should say to himself: I amfrom eternity to eternity; beside me there isnothing, except that which exists by my will;whence then am I? Here all sinks away fromunder us; and the greatest, as the smallest, per-fection, hovers without stay or footing in pre-sence of the speculative reason, which finds itas easy to part with the one as with the other.

Many physical powers, which evidence theirexistence by their effects, are perfectly inscruta-ble in their nature; they elude all our powers ofobservation. The transcendental object whichforms the basis of phenomena, and, in connec-tion with it, the reason why our sensibility pos-sesses this rather than that particular kind ofconditions, are and must ever remain hiddenfrom our mental vision; the fact is there, thereason of the fact we cannot see. But an ideal ofpure reason cannot be termed mysterious orinscrutable, because the only credential of itsreality is the need of it felt by reason, for thepurpose of giving completeness to the world ofsynthetical unity. An ideal is not even given asa cogitable object, and therefore cannot be ins-crutable; on the contrary, it must, as a mereidea, be based on the constitution of reasonitself, and on this account must be capable ofexplanation and solution. For the very essenceof reason consists in its ability to give an ac-count, of all our conceptions, opinions, and

assertions—upon objective, or, when they hap-pen to be illusory and fallacious, upon subjecti-ve grounds.

Detection and Explanation of the DialecticalIllusion in all Transcendental Arguments forthe Existence of a Necessary Being.

Both of the above arguments are transcenden-tal; in other words, they do not proceed uponempirical principles. For, although the cosmo-logical argument professed to lay a basis ofexperience for its edifice of reasoning, it did notground its procedure upon the peculiar consti-tution of experience, but upon pure principlesof reason—in relation to an existence given byempirical consciousness; utterly abandoning itsguidance, however, for the purpose of suppor-ting its assertions entirely upon pure concep-tions. Now what is the cause, in these transcen-

dental arguments, of the dialectical, but natu-ral, illusion, which connects the conceptions ofnecessity and supreme reality, and hypostatizesthat which cannot be anything but an idea?What is the cause of this unavoidable step onthe part of reason, of admitting that some oneamong all existing things must be necessary,while it falls back from the assertion of the exis-tence of such a being as from an abyss? Andhow does reason proceed to explain this ano-maly to itself, and from the wavering conditionof a timid and reluctant approbation—alwaysagain withdrawn—arrive at a calm and settledinsight into its cause?

It is something very remarkable that, on thesupposition that something exists, I cannotavoid the inference that something exists neces-sarily. Upon this perfectly natural—but not onthat account reliable—inference does the cos-mological argument rest. But, let me form anyconception whatever of a thing, I find that I

cannot cogitate the existence of the thing asabsolutely necessary, and that nothing preventsme—be the thing or being what it may—fromcogitating its non-existence. I may thus be obli-ged to admit that all existing things have a ne-cessary basis, while I cannot cogitate any singleor individual thing as necessary. In otherwords, I can never complete the regressthrough the conditions of existence, withoutadmitting the existence of a necessary being;but, on the other hand, I cannot make a com-mencement from this being.

If I must cogitate something as existing neces-sarily as the basis of existing things, and yet amnot permitted to cogitate any individual thingas in itself necessary, the inevitable inference isthat necessity and contingency are not proper-ties of things themselves- otherwise an internalcontradiction would result; that consequentlyneither of these principles are objective, butmerely subjective principles of reason—the one

requiring us to seek for a necessary ground foreverything that exists, that is, to be satisfiedwith no other explanation than that which iscomplete a priori, the other forbidding us everto hope for the attainment of this completeness,that is, to regard no member of the empiricalworld as unconditioned. In this mode of vie-wing them, both principles, in their purely heu-ristic and regulative character, and as concer-ning merely the formal interest of reason, arequite consistent with each other. The one says:"You must philosophize upon nature," as ifthere existed a necessary primal basis of allexisting things, solely for the purpose of intro-ducing systematic unity into your knowledge,by pursuing an idea of this character—a foun-dation which is arbitrarily admitted to be ulti-mate; while the other warns you to consider noindividual determination, concerning the exis-tence of things, as such an ultimate foundation,that is, as absolutely necessary, but to keep theway always open for further progress in the

deduction, and to treat every determination asdetermined by some other. But if all that weperceive must be regarded as conditionallynecessary, it is impossible that anything whichis empirically given should be absolutely ne-cessary.

It follows from this that you must accept theabsolutely necessary as out of and beyond theworld, inasmuch as it is useful only as a princi-ple of the highest possible unity in experience,and you cannot discover any such necessaryexistence in the would, the second rule requi-ring you to regard all empirical causes of unityas themselves deduced.

The philosophers of antiquity regarded all theforms of nature as contingent; while matter wasconsidered by them, in accordance with thejudgement of the common reason of mankind,as primal and necessary. But if they had regar-ded matter, not relatively—as the substratumof phenomena, but absolutely and in itself—as

an independent existence, this idea of absolutenecessity would have immediately disappea-red. For there is nothing absolutely connectingreason with such an existence; on the contrary,it can annihilate it in thought, always and wit-hout self-contradiction. But in thought alonelay the idea of absolute necessity. A regulativeprinciple must, therefore, have been at thefoundation of this opinion. In fact, extensionand impenetrability—which together constituteour conception of matter—form the supremeempirical principle of the unity of phenomena,and this principle, in so far as it is empiricallyunconditioned, possesses the property of a re-gulative principle. But, as every determinationof matter which constitutes what is real in it—and consequently impenetrability—is an effect,which must have a cause, and is for this reasonalways derived, the notion of matter cannotharmonize with the idea of a necessary being,in its character of the principle of all derivedunity. For every one of its real properties, being

derived, must be only conditionally necessary,and can therefore be annihilated in thought;and thus the whole existence of matter can beso annihilated or suppressed. If this were notthe case, we should have found in the world ofphenomena the highest ground or condition ofunity—which is impossible, according to thesecond regulative principle. It follows that mat-ter, and, in general, all that forms part of theworld of sense, cannot be a necessary primalbeing, nor even a principle of empirical unity,but that this being or principle must have itsplace assigned without the world. And, in thisway, we can proceed in perfect confidence todeduce the phenomena of the world and theirexistence from other phenomena, just as if thereexisted no necessary being; and we can at thesame time, strive without ceasing towards theattainment of completeness for our deduction,just as if such a being—the supreme conditionof all existences—were presupposed by themind.

These remarks will have made it evident to thereader that the ideal of the Supreme Being, farfrom being an enouncement of the existence ofa being in itself necessary, is nothing more thana regulative principle of reason, requiring us toregard all connection existing between pheno-mena as if it had its origin from an all-sufficientnecessary cause, and basing upon this the ruleof a systematic and necessary unity in the ex-planation of phenomena. We cannot, at the sa-me time, avoid regarding, by a transcendentalsubreptio, this formal principle as constitutive,and hypostatizing this unity. Precisely similaris the case with our notion of space. Space is theprimal condition of all forms, which are proper-ly just so many different limitations of it; andthus, although it is merely a principle of sensi-bility, we cannot help regarding it as an absolu-tely necessary and self-subsistent thing—as anobject given a priori in itself. In the same way,it is quite natural that, as the systematic unityof nature cannot be established as a principle

for the empirical employment of reason, unlessit is based upon the idea of an ens realissimum,as the supreme cause, we should regard thisidea as a real object, and this object, in its cha-racter of supreme condition, as absolutely ne-cessary, and that in this way a regulativeshould be transformed into a constitutive prin-ciple. This interchange becomes evident when Iregard this supreme being, which, relatively tothe world, was absolutely (unconditionally)necessary, as a thing per se. In this case, I find itimpossible to represent this necessity in or byany conception, and it exists merely in my ownmind, as the formal condition of thought, butnot as a material and hypostatic condition ofexistence.

SECTION VI. Of the Impossibility of aPhysico-Theological Proof.

If, then, neither a pure conception nor the gene-ral experience of an existing being can providea sufficient basis for the proof of the existenceof the Deity, we can make the attempt by theonly other mode—that of grounding our argu-ment upon a determinate experience of thephenomena of the present world, their constitu-tion and disposition, and discover whether wecan thus attain to a sound conviction of theexistence of a Supreme Being. This argumentwe shall term the physico-theological argu-ment. If it is shown to be insufficient, speculati-ve reason cannot present us with any satisfac-tory proof of the existence of a being corres-ponding to our transcendental idea.

It is evident from the remarks that have beenmade in the preceding sections, that an answerto this question will be far from being difficult

or unconvincing. For how can any experiencebe adequate with an idea? The very essence ofan idea consists in the fact that no experiencecan ever be discovered congruent or adequatewith it. The transcendental idea of a necessaryand all-sufficient being is so immeasurablygreat, so high above all that is empirical, whichis always conditioned, that we hope in vain tofind materials in the sphere of experience suffi-ciently ample for our conception, and in vainseek the unconditioned among things that areconditioned, while examples, nay, even gui-dance is denied us by the laws of empiricalsynthesis.

If the Supreme Being forms a link in the chainof empirical conditions, it must be a member ofthe empirical series, and, like the lower mem-bers which it precedes, have its origin in somehigher member of the series. If, on the otherhand, we disengage it from the chain, and cogi-tate it as an intelligible being, apart from the

series of natural causes—how shall reason brid-ge the abyss that separates the latter from theformer? All laws respecting the regress fromeffects to causes, all synthetical additions to ourknowledge relate solely to possible experienceand the objects of the sensuous world, and,apart from them, are without significance.

The world around us opens before our view somagnificent a spectacle of order, variety, beau-ty, and conformity to ends, that whether wepursue our observations into the infinity ofspace in the one direction, or into its illimitabledivisions in the other, whether we regard theworld in its greatest or its least manifestations-even after we have attained to the highest sum-mit of knowledge which our weak minds canreach, we find that language in the presence ofwonders so inconceivable has lost its force, andnumber its power to reckon, nay, even thoughtfails to conceive adequately, and our concep-tion of the whole dissolves into an astonish-

ment without power of expression—all the mo-re eloquent that it is dumb. Everywhere aroundus we observe a chain of causes and effects, ofmeans and ends, of death and birth; and, asnothing has entered of itself into the conditionin which we find it, we are constantly referredto some other thing, which itself suggests thesame inquiry regarding its cause, and thus theuniverse must sink into the abyss of nothing-ness, unless we admit that, besides this infinitechain of contingencies, there exists somethingthat is primal and self-subsistent—somethingwhich, as the cause of this phenomenal world,secures its continuance and preservation.

This highest cause—what magnitude shall weattribute to it? Of the content of the world weare ignorant; still less can we estimate its mag-nitude by comparison with the sphere of thepossible. But this supreme cause being a neces-sity of the human mind, what is there to pre-vent us from attributing to it such a degree of

perfection as to place it above the sphere of allthat is possible? This we can easily do, althoughonly by the aid of the faint outline of an abs-tract conception, by representing this being toourselves as containing in itself, as an indivi-dual substance, all possible perfection—a con-ception which satisfies that requirement of rea-son which demands parsimony in principles,which is free from self-contradiction, whicheven contributes to the extension of the em-ployment of reason in experience, by means ofthe guidance afforded by this idea to order andsystem, and which in no respect conflicts withany law of experience.

This argument always deserves to be mentio-ned with respect. It is the oldest, the clearest,and that most in conformity with the commonreason of humanity. It animates the study ofnature, as it itself derives its existence anddraws ever new strength from that source. Itintroduces aims and ends into a sphere in

which our observation could not of itself havediscovered them, and extends our knowledgeof nature, by directing our attention to a unity,the principle of which lies beyond nature. Thisknowledge of nature again reacts upon thisidea—its cause; and thus our belief in a divineauthor of the universe rises to the power of anirresistible conviction.

For these reasons it would be utterly hopelessto attempt to rob this argument of the authorityit has always enjoyed. The mind, unceasinglyelevated by these considerations, which, alt-hough empirical, are so remarkably powerful,and continually adding to their force, will notsuffer itself to be depressed by the doubts sug-gested by subtle speculation; it tears itself outof this state of uncertainty, the moment it castsa look upon the wondrous forms of nature andthe majesty of the universe, and rises fromheight to height, from condition to condition,

till it has elevated itself to the supreme andunconditioned author of all.

But although we have nothing to object to thereasonableness and utility of this procedure,but have rather to commend and encourage it,we cannot approve of the claims which thisargument advances to demonstrative certaintyand to a reception upon its own merits, apartfrom favour or support by other arguments.Nor can it injure the cause of morality to en-deavour to lower the tone of the arrogant sop-hist, and to teach him that modesty and mode-ration which are the properties of a belief thatbrings calm and content into the mind, withoutprescribing to it an unworthy subjection. Imaintain, then, that the physico-theologicalargument is insufficient of itself to prove theexistence of a Supreme Being, that it must en-trust this to the ontological argument—towhich it serves merely as an introduction, andthat, consequently, this argument contains the

only possible ground of proof (possessed byspeculative reason) for the existence of thisbeing.

The chief momenta in the physico-theologicalargument are as follow: 1. We observe in theworld manifest signs of an arrangement full ofpurpose, executed with great wisdom, and ar-gument in whole of a content indescribablyvarious, and of an extent without limits. 2. Thisarrangement of means and ends is entirely fo-reign to the things existing in the world—itbelongs to them merely as a contingent attribu-te; in other words, the nature of different thingscould not of itself, whatever means were em-ployed, harmoniously tend towards certainpurposes, were they not chosen and directedfor these purposes by a rational and disposingprinciple, in accordance with certain funda-mental ideas. 3. There exists, therefore, a subli-me and wise cause (or several), which is notmerely a blind, all-powerful nature, producing

the beings and events which fill the world inunconscious fecundity, but a free and intelli-gent cause of the world. 4. The unity of thiscause may be inferred from the unity of thereciprocal relation existing between the parts ofthe world, as portions of an artistic edifice—aninference which all our observation favours,and all principles of analogy support.

In the above argument, it is inferred from theanalogy of certain products of nature with tho-se of human art, when it compels Nature tobend herself to its purposes, as in the case of ahouse, a ship, or a watch, that the same kind ofcausality—namely, understanding and will—resides in nature. It is also declared that theinternal possibility of this freely-acting nature(which is the source of all art, and perhaps alsoof human reason) is derivable from another andsuperhuman art—a conclusion which wouldperhaps be found incapable of standing the testof subtle transcendental criticism. But to neither

of these opinions shall we at present object. Weshall only remark that it must be confessedthat, if we are to discuss the subject of cause atall, we cannot proceed more securely than withthe guidance of the analogy subsisting betweennature and such products of design—thesebeing the only products whose causes and mo-des of organization are completely known tous. Reason would be unable to satisfy her ownrequirements, if she passed from a causalitywhich she does know, to obscure and inde-monstrable principles of explanation which shedoes not know.

According to the physico-theological argument,the connection and harmony existing in theworld evidence the contingency of the formmerely, but not of the matter, that is, of thesubstance of the world. To establish the truth ofthe latter opinion, it would be necessary to pro-ve that all things would be in themselves inca-pable of this harmony and order, unless they

were, even as regards their substance, the pro-duct of a supreme wisdom. But this would re-quire very different grounds of proof from tho-se presented by the analogy with human art.This proof can at most, therefore, demonstratethe existence of an architect of the world, who-se efforts are limited by the capabilities of thematerial with which he works, but not of acreator of the world, to whom all things aresubject. Thus this argument is utterly insuffi-cient for the task before us—a demonstration ofthe existence of an all-sufficient being. If wewish to prove the contingency of matter, wemust have recourse to a transcendental argu-ment, which the physico-theological was cons-tructed expressly to avoid.

We infer, from the order and design visible inthe universe, as a disposition of a thoroughlycontingent character, the existence of a causeproportionate thereto. The conception of thiscause must contain certain determinate quali-

ties, and it must therefore be regarded as theconception of a being which possesses all po-wer, wisdom, and so on, in one word, all per-fection—the conception, that is, of an all-sufficient being. For the predicates of verygreat, astonishing, or immeasurable power andexcellence, give us no determinate conceptionof the thing, nor do they inform us what thething may be in itself. They merely indicate therelation existing between the magnitude of theobject and the observer, who compares it withhimself and with his own power of comprehen-sion, and are mere expressions of praise andreverence, by which the object is either magni-fied, or the observing subject depreciated inrelation to the object. Where we have to dowith the magnitude (of the perfection) of athing, we can discover no determinate concep-tion, except that which comprehends all possi-ble perfection or completeness, and it is onlythe total (omnitudo) of reality which is comple-

tely determined in and through its conceptionalone.

Now it cannot be expected that any one will bebold enough to declare that he has a perfectinsight into the relation which the magnitude ofthe world he contemplates bears (in its extentas well as in its content) to omnipotence, intothat of the order and design in the world to thehighest wisdom, and that of the unity of theworld to the absolute unity of a Supreme Being.Physico-theology is therefore incapable of pre-senting a determinate conception of a supremecause of the world, and is therefore insufficientas a principle of theology—a theology which isitself to be the basis of religion.

The attainment of absolute totality is complete-ly impossible on the path of empiricism. Andyet this is the path pursued in the physico-theological argument. What means shall weemploy to bridge the abyss?

After elevating ourselves to admiration of themagnitude of the power, wisdom, and otherattributes of the author of the world, and fin-ding we can advance no further, we leave theargument on empirical grounds, and proceedto infer the contingency of the world from theorder and conformity to aims that are observa-ble in it. From this contingency we infer, by thehelp of transcendental conceptions alone, theexistence of something absolutely necessary;and, still advancing, proceed from the concep-tion of the absolute necessity of the first causeto the completely determined or determiningconception thereof—the conception of an all-embracing reality. Thus the physico-theological, failing in its undertaking, recurs inits embarrassment to the cosmological argu-ment; and, as this is merely the ontological ar-gument in disguise, it executes its design solelyby the aid of pure reason, although it at firstprofessed to have no connection with this facul-

ty and to base its entire procedure upon expe-rience alone.

The physico-theologians have therefore no rea-son to regard with such contempt the transcen-dental mode of argument, and to look downupon it, with the conceit of clear-sighted obser-vers of nature, as the brain-cobweb of obscurespeculatists. For, if they reflect upon and exa-mine their own arguments, they will find that,after following for some time the path of natureand experience, and discovering themselves nonearer their object, they suddenly leave thispath and pass into the region of pure possibili-ty, where they hope to reach upon the wings ofideas what had eluded all their empirical inves-tigations. Gaining, as they think, a firm footingafter this immense leap, they extend their de-terminate conception—into the possession ofwhich they have come, they know not how—over the whole sphere of creation, and explaintheir ideal, which is entirely a product of pure

reason, by illustrations drawn from experien-ce—though in a degree miserably unworthy ofthe grandeur of the object, while they refuse toacknowledge that they have arrived at this cog-nition or hypothesis by a very different roadfrom that of experience.

Thus the physico-theological is based upon thecosmological, and this upon the ontologicalproof of the existence of a Supreme Being; andas besides these three there is no other pathopen to speculative reason, the ontologicalproof, on the ground of pure conceptions ofreason, is the only possible one, if any proof ofa proposition so far transcending the empiricalexercise of the understanding is possible at all.

SECTION VII. Critique of all Theology ba-sed upon Speculative Principles of Reason.

If by the term theology I understand the cogni-tion of a primal being, that cognition is basedeither upon reason alone (theologia rationalis)or upon revelation (theologia revelata). Theformer cogitates its object either by means ofpure transcendental conceptions, as an ens ori-ginarium, realissimum, ens entium, and is ter-med transcendental theology; or, by means of aconception derived from the nature of our ownmind, as a supreme intelligence, and must thenbe entitled natural theology. The person whobelieves in a transcendental theology alone, istermed a deist; he who acknowledges the pos-sibility of a natural theology also, a theist. Theformer admits that we can cognize by pure rea-son alone the existence of a Supreme Being, butat the same time maintains that our conceptionof this being is purely transcendental, and thatall we can say of it is that it possesses all reality,

without being able to define it more closely.The second asserts that reason is capable ofpresenting us, from the analogy with nature,with a more definite conception of this being,and that its operations, as the cause of allthings, are the results of intelligence and freewill. The former regards the Supreme Being asthe cause of the world—whether by the necessi-ty of his nature, or as a free agent, is left unde-termined; the latter considers this being as theauthor of the world.

Transcendental theology aims either at infe-rring the existence of a Supreme Being from ageneral experience, without any closer referen-ce to the world to which this experience be-longs, and in this case it is called cosmotheolo-gy; or it endeavours to cognize the existence ofsuch a being, through mere conceptions, wit-hout the aid of experience, and is then termedontotheology.

Natural theology infers the attributes and theexistence of an author of the world, from theconstitution of, the order and unity observablein, the world, in which two modes of causalitymust be admitted to exist—those of nature andfreedom. Thus it rises from this world to a su-preme intelligence, either as the principle of allnatural, or of all moral order and perfection. Inthe former case it is termed physico-theology,in the latter, ethical or moral-theology.*

[*Footnote: Not theological ethics; for thisscience contains ethical laws, which presupposethe existence of a Supreme Governor of theworld; while moral-theology, on the contrary,is the expression of a conviction of the existenceof a Supreme Being, founded upon ethicallaws.]

As we are wont to understand by the term Godnot merely an eternal nature, the operations ofwhich are insensate and blind, but a SupremeBeing, who is the free and intelligent author of

all things, and as it is this latter view alone thatcan be of interest to humanity, we might, instrict rigour, deny to the deist any belief in Godat all, and regard him merely as a maintainer ofthe existence of a primal being or thing—thesupreme cause of all other things. But, as noone ought to be blamed, merely because hedoes not feel himself justified in maintaining acertain opinion, as if he altogether denied itstruth and asserted the opposite, it is more co-rrect—as it is less harsh—to say, the deist belie-ves in a God, the theist in a living God (summaintelligentia). We shall now proceed to investi-gate the sources of all these attempts of reasonto establish the existence of a Supreme Being.

It may be sufficient in this place to define theo-retical knowledge or cognition as knowledge ofthat which is, and practical knowledge asknowledge of that which ought to be. In thisview, the theoretical employment of reason isthat by which I cognize a priori (as necessary)

that something is, while the practical is that bywhich I cognize a priori what ought to happen.Now, if it is an indubitably certain, though atthe same time an entirely conditioned truth,that something is, or ought to happen, either acertain determinate condition of this truth isabsolutely necessary, or such a condition maybe arbitrarily presupposed. In the former casethe condition is postulated (per thesin), in thelatter supposed (per hypothesin). There arecertain practical laws—those of morality—which are absolutely necessary. Now, if theselaws necessarily presuppose the existence ofsome being, as the condition of the possibilityof their obligatory power, this being must bepostulated, because the conditioned, fromwhich we reason to this determinate condition,is itself cognized a priori as absolutely necessa-ry. We shall at some future time show that themoral laws not merely presuppose the existen-ce of a Supreme Being, but also, as themselvesabsolutely necessary in a different relation,

demand or postulate it—although only from apractical point of view. The discussion of thisargument we postpone for the present.

When the question relates merely to that whichis, not to that which ought to be, the conditio-ned which is presented in experience is alwayscogitated as contingent. For this reason its con-dition cannot be regarded as absolutely neces-sary, but merely as relatively necessary, or rat-her as needful; the condition is in itself and apriori a mere arbitrary presupposition in aid ofthe cognition, by reason, of the conditioned. If,then, we are to possess a theoretical cognitionof the absolute necessity of a thing, we cannotattain to this cognition otherwise than a prioriby means of conceptions; while it is impossiblein this way to cognize the existence of a causewhich bears any relation to an existence givenin experience.

Theoretical cognition is speculative when itrelates to an object or certain conceptions of an

object which is not given and cannot be disco-vered by means of experience. It is opposed tothe cognition of nature, which concerns onlythose objects or predicates which can be pre-sented in a possible experience.

The principle that everything which happens(the empirically contingent) must have a cause,is a principle of the cognition of nature, but notof speculative cognition. For, if we change itinto an abstract principle, and deprive it of itsreference to experience and the empirical, weshall find that it cannot with justice be regardedany longer as a synthetical proposition, andthat it is impossible to discover any mode oftransition from that which exists to somethingentirely different—termed cause. Nay, more,the conception of a cause likewise that of thecontingent—loses, in this speculative mode ofemploying it, all significance, for its objectivereality and meaning are comprehensible fromexperience alone.

When from the existence of the universe andthe things in it the existence of a cause of theuniverse is inferred, reason is proceeding not inthe natural, but in the speculative method. Forthe principle of the former enounces, not thatthings themselves or substances, but only thatwhich happens or their states—as empiricallycontingent, have a cause: the assertion that theexistence of substance itself is contingent is notjustified by experience, it is the assertion of areason employing its principles in a speculativemanner. If, again, I infer from the form of theuniverse, from the way in which all things areconnected and act and react upon each other,the existence of a cause entirely distinct fromthe universe—this would again be a judgementof purely speculative reason; because the objectin this case—the cause—can never be an objectof possible experience. In both these cases theprinciple of causality, which is valid only in thefield of experience—useless and even meanin-

gless beyond this region, would be divertedfrom its proper destination.

Now I maintain that all attempts of reason toestablish a theology by the aid of speculationalone are fruitless, that the principles of reasonas applied to nature do not conduct us to anytheological truths, and, consequently, that arational theology can have no existence, unlessit is founded upon the laws of morality. For allsynthetical principles of the understanding arevalid only as immanent in experience; while thecognition of a Supreme Being necessitates theirbeing employed transcendentally, and of thisthe understanding is quite incapable. If the em-pirical law of causality is to conduct us to aSupreme Being, this being must belong to thechain of empirical objects—in which case itwould be, like all phenomena, itself conditio-ned. If the possibility of passing the limits ofexperience be admitted, by means of the dyna-mical law of the relation of an effect to its cause,

what kind of conception shall we obtain by thisprocedure? Certainly not the conception of aSupreme Being, because experience never pre-sents us with the greatest of all possible effects,and it is only an effect of this character thatcould witness to the existence of a correspon-ding cause. If, for the purpose of fully satisf-ying the requirements of Reason, we recognizeher right to assert the existence of a perfect andabsolutely necessary being, this can be admit-ted only from favour, and cannot be regardedas the result or irresistible demonstration. Thephysico-theological proof may add weight toothers—if other proofs there are—by connec-ting speculation with experience; but in itself itrather prepares the mind for theological cogni-tion, and gives it a right and natural direction,than establishes a sure foundation for theology.

It is now perfectly evident that transcendentalquestions admit only of transcendental ans-wers—those presented a priori by pure concep-

tions without the least empirical admixture. Butthe question in the present case is evidentlysynthetical—it aims at the extension of our cog-nition beyond the bounds of experience—itrequires an assurance respecting the existenceof a being corresponding with the idea in ourminds, to which no experience can ever be ade-quate. Now it has been abundantly proved thatall a priori synthetical cognition is possible onlyas the expression of the formal conditions of apossible experience; and that the validity of allprinciples depends upon their immanence inthe field of experience, that is, their relation toobjects of empirical cognition or phenomena.Thus all transcendental procedure in referenceto speculative theology is without result.

If any one prefers doubting the conclusivenessof the proofs of our analytic to losing the per-suasion of the validity of these old and timehonoured arguments, he at least cannot declineanswering the question—how he can pass the

limits of all possible experience by the help ofmere ideas. If he talks of new arguments, or ofimprovements upon old arguments, I requesthim to spare me. There is certainly no greatchoice in this sphere of discussion, as all specu-lative arguments must at last look for supportto the ontological, and I have, therefore, verylittle to fear from the argumentative fecundityof the dogmatical defenders of a non-sensuousreason. Without looking upon myself as a re-markably combative person, I shall not declinethe challenge to detect the fallacy and destroythe pretensions of every attempt of speculativetheology. And yet the hope of better fortunenever deserts those who are accustomed to thedogmatical mode of procedure. I shall, therefo-re, restrict myself to the simple and equitabledemand that such reasoners will demonstrate,from the nature of the human mind as well asfrom that of the other sources of knowledge,how we are to proceed to extend our cognitioncompletely a priori, and to carry it to that point

where experience abandons us, and no meansexist of guaranteeing the objective reality of ourconceptions. In whatever way the understan-ding may have attained to a conception, theexistence of the object of the conception cannotbe discovered in it by analysis, because thecognition of the existence of the object dependsupon the object's being posited and given initself apart from the conception. But it is utterlyimpossible to go beyond our conception, wit-hout the aid of experience—which presents tothe mind nothing but phenomena, or to attainby the help of mere conceptions to a convictionof the existence of new kinds of objects or su-pernatural beings.

But although pure speculative reason is farfrom sufficient to demonstrate the existence ofa Supreme Being, it is of the highest utility incorrecting our conception of this being—on thesupposition that we can attain to the cognitionof it by some other means—in making it consis-

tent with itself and with all other conceptions ofintelligible objects, clearing it from all that isincompatible with the conception of an enssummun, and eliminating from it all limitationsor admixtures of empirical elements.

Transcendental theology is still therefore, not-withstanding its objective insufficiency, of im-portance in a negative respect; it is useful as atest of the procedure of reason when engagedwith pure ideas, no other than a transcendentalstandard being in this case admissible. For if,from a practical point of view, the hypothesis ofa Supreme and All-sufficient Being is to main-tain its validity without opposition, it must beof the highest importance to define this concep-tion in a correct and rigorous manner—as thetranscendental conception of a necessary being,to eliminate all phenomenal elements (anthro-pomorphism in its most extended significa-tion), and at the same time to overflow all con-tradictory assertions—be they atheistic, deistic,

or anthropomorphic. This is of course very ea-sy; as the same arguments which demonstratedthe inability of human reason to affirm the exis-tence of a Supreme Being must be alike suffi-cient to prove the invalidity of its denial. For itis impossible to gain from the pure speculationof reason demonstration that there exists noSupreme Being, as the ground of all that exists,or that this being possesses none of those pro-perties which we regard as analogical with thedynamical qualities of a thinking being, or that,as the anthropomorphists would have us belie-ve, it is subject to all the limitations which sen-sibility imposes upon those intelligences whichexist in the world of experience.

A Supreme Being is, therefore, for the specula-tive reason, a mere ideal, though a faultlessone—a conception which perfects and crownsthe system of human cognition, but the objecti-ve reality of which can neither be proved nordisproved by pure reason. If this defect is ever

supplied by a moral theology, the problematictranscendental theology which has preceded,will have been at least serviceable as demons-trating the mental necessity existing for theconception, by the complete determination of itwhich it has furnished, and the ceaseless testingof the conclusions of a reason often deceived bysense, and not always in harmony with its ownideas. The attributes of necessity, infinitude,unity, existence apart from the world (and notas a world soul), eternity (free from conditionsof time), omnipresence (free from conditions ofspace), omnipotence, and others, are puretranscendental predicates; and thus the accura-te conception of a Supreme Being, which everytheology requires, is furnished by transcenden-tal theology alone.

APPENDIX.

Of the Regulative Employment of the Ideas ofPure Reason.

The result of all the dialectical attempts of purereason not only confirms the truth of what wehave already proved in our TranscendentalAnalytic, namely, that all inferences whichwould lead us beyond the limits of experienceare fallacious and groundless, but it at the sametime teaches us this important lesson, thathuman reason has a natural inclination tooverstep these limits, and that transcendentalideas are as much the natural property of thereason as categories are of the understanding.There exists this difference, however, that whilethe categories never mislead us, outward ob-jects being always in perfect harmony there-with, ideas are the parents of irresistible illu-sions, the severest and most subtle criticism

being required to save us from the fallacieswhich they induce.

Whatever is grounded in the nature of our po-wers will be found to be in harmony with thefinal purpose and proper employment of thesepowers, when once we have discovered theirtrue direction and aim. We are entitled to sup-pose, therefore, that there exists a mode of em-ploying transcendental ideas which is properand immanent; although, when we mistaketheir meaning, and regard them as conceptionsof actual things, their mode of application istranscendent and delusive. For it is not the ideaitself, but only the employment of the idea inrelation to possible experience, that is transcen-dent or immanent. An idea is employed trans-cendently, when it is applied to an object false-ly believed to be adequate with and to corres-pond to it; imminently, when it is applied sole-ly to the employment of the understanding inthe sphere of experience. Thus all errors of su-

breptio—of misapplication, are to be ascribedto defects of judgement, and not to understan-ding or reason.

Reason never has an immediate relation to anobject; it relates immediately to the understan-ding alone. It is only through the understan-ding that it can be employed in the field of ex-perience. It does not form conceptions of ob-jects, it merely arranges them and gives to themthat unity which they are capable of possessingwhen the sphere of their application has beenextended as widely as possible. Reason availsitself of the conception of the understanding forthe sole purpose of producing totality in thedifferent series. This totality the understandingdoes not concern itself with; its only occupationis the connection of experiences, by which se-ries of conditions in accordance with concep-tions are established. The object of reason is,therefore, the understanding and its properdestination. As the latter brings unity into the

diversity of objects by means of its conceptions,so the former brings unity into the diversity ofconceptions by means of ideas; as it sets thefinal aim of a collective unity to the operationsof the understanding, which without this occu-pies itself with a distributive unity alone.

I accordingly maintain that transcendentalideas can never be employed as constitutiveideas, that they cannot be conceptions of ob-jects, and that, when thus considered, they as-sume a fallacious and dialectical character. But,on the other hand, they are capable of an admi-rable and indispensably necessary applicationto objects—as regulative ideas, directing theunderstanding to a certain aim, the guidinglines towards which all its laws follow, and inwhich they all meet in one point. This point—though a mere idea (focus imaginarius), that is,not a point from which the conceptions of theunderstanding do really proceed, for it lies be-yond the sphere of possible experience—serves,

notwithstanding, to give to these conceptionsthe greatest possible unity combined with thegreatest possible extension. Hence arises thenatural illusion which induces us to believethat these lines proceed from an object whichlies out of the sphere of empirical cognition,just as objects reflected in a mirror appear to bebehind it. But this illusion—which we mayhinder from imposing upon us—is necessaryand unavoidable, if we desire to see, not onlythose objects which lie before us, but thosewhich are at a great distance behind us; that isto say, when, in the present case, we direct theaims of the understanding, beyond every givenexperience, towards an extension as great ascan possibly be attained.

If we review our cognitions in their entire ex-tent, we shall find that the peculiar business ofreason is to arrange them into a system, that isto say, to give them connection according to aprinciple. This unity presupposes an idea—the

idea of the form of a whole (of cognition), pre-ceding the determinate cognition of the parts,and containing the conditions which determinea priori to every part its place and relation tothe other parts of the whole system. This idea,accordingly, demands complete unity in thecognition of the understanding—not the unityof a contingent aggregate, but that of a systemconnected according to necessary laws. It can-not be affirmed with propriety that this idea isa conception of an object; it is merely a concep-tion of the complete unity of the conceptions ofobjects, in so far as this unity is available to theunderstanding as a rule. Such conceptions ofreason are not derived from nature; on the con-trary, we employ them for the interrogationand investigation of nature, and regard ourcognition as defective so long as it is not ade-quate to them. We admit that such a thing aspure earth, pure water, or pure air, is not to bediscovered. And yet we require these concep-tions (which have their origin in the reason, so

far as regards their absolute purity and comple-teness) for the purpose of determining the sha-re which each of these natural causes has inevery phenomenon. Thus the different kinds ofmatter are all referred to earths, as mereweight; to salts and inflammable bodies, aspure force; and finally, to water and air, as thevehicula of the former, or the machines emplo-yed by them in their operations—for the pur-pose of explaining the chemical action andreaction of bodies in accordance with the ideaof a mechanism. For, although not actually soexpressed, the influence of such ideas of reasonis very observable in the procedure of naturalphilosophers.

If reason is the faculty of deducing the particu-lar from the general, and if the general be cer-tain in se and given, it is only necessary that thejudgement should subsume the particular un-der the general, the particular being thus neces-sarily determined. I shall term this the demons-

trative or apodeictic employment of reason. If,however, the general is admitted as problema-tical only, and is a mere idea, the particular caseis certain, but the universality of the rule whichapplies to this particular case remains a pro-blem. Several particular cases, the certainty ofwhich is beyond doubt, are then taken and exa-mined, for the purpose of discovering whetherthe rule is applicable to them; and if it appearsthat all the particular cases which can be collec-ted follow from the rule, its universality is infe-rred, and at the same time, all the causes whichhave not, or cannot be presented to our obser-vation, are concluded to be of the same charac-ter with those which we have observed. This Ishall term the hypothetical employment of thereason.

The hypothetical exercise of reason by the aidof ideas employed as problematical conceptionsis properly not constitutive. That is to say, if weconsider the subject strictly, the truth of the

rule, which has been employed as an hypot-hesis, does not follow from the use that is madeof it by reason. For how can we know all thepossible cases that may arise? some of whichmay, however, prove exceptions to the univer-sality of the rule. This employment of reason ismerely regulative, and its sole aim is the intro-duction of unity into the aggregate of our parti-cular cognitions, and thereby the approxima-ting of the rule to universality.

The object of the hypothetical employment ofreason is therefore the systematic unity of cog-nitions; and this unity is the criterion of thetruth of a rule. On the other hand, this systema-tic unity—as a mere idea—is in fact merely aunity projected, not to be regarded as given,but only in the light of a problem—a problemwhich serves, however, as a principle for thevarious and particular exercise of the unders-tanding in experience, directs it with regard tothose cases which are not presented to our ob-

servation, and introduces harmony and consis-tency into all its operations.

All that we can be certain of from the aboveconsiderations is that this systematic unity is alogical principle, whose aim is to assist the un-derstanding, where it cannot of itself attain torules, by means of ideas, to bring all these va-rious rules under one principle, and thus toensure the most complete consistency and con-nection that can be attained. But the assertionthat objects and the understanding by whichthey are cognized are so constituted as to bedetermined to systematic unity, that this maybe postulated a priori, without any reference tothe interest of reason, and that we are justifiedin declaring all possible cognitions—empiricaland others—to possess systematic unity, and tobe subject to general principles from which,notwithstanding their various character, theyare all derivable such an assertion can be foun-ded only upon a transcendental principle of

reason, which would render this systematicunity not subjectively and logically—in its cha-racter of a method, but objectively necessary.

We shall illustrate this by an example. The con-ceptions of the understanding make us ac-quainted, among many other kinds of unity,with that of the causality of a substance, whichis termed power. The different phenomenalmanifestations of the same substance appear atfirst view to be so very dissimilar that we areinclined to assume the existence of just as manydifferent powers as there are different effects—as, in the case of the human mind, we have fee-ling, consciousness, imagination, memory, wit,analysis, pleasure, desire and so on. Now weare required by a logical maxim to reduce thesedifferences to as small a number as possible, bycomparing them and discovering the hiddenidentity which exists. We must inquire, forexample, whether or not imagination (connec-ted with consciousness), memory, wit, and ana-

lysis are not merely different forms of unders-tanding and reason. The idea of a fundamentalpower, the existence of which no effort of logiccan assure us of, is the problem to be solved,for the systematic representation of the existingvariety of powers. The logical principle of rea-son requires us to produce as great a unity as ispossible in the system of our cognitions; andthe more the phenomena of this and the otherpower are found to be identical, the more pro-bable does it become, that they are nothing butdifferent manifestations of one and the samepower, which may be called, relatively spea-king, a fundamental power. And so with othercases.

These relatively fundamental powers mustagain be compared with each other, to discover,if possible, the one radical and absolutely fun-damental power of which they are but the ma-nifestations. But this unity is purely hypotheti-cal. It is not maintained, that this unity does

really exist, but that we must, in the interest ofreason, that is, for the establishment of princi-ples for the various rules presented by expe-rience, try to discover and introduce it, so far asis practicable, into the sphere of our cognitions.

But the transcendental employment of the un-derstanding would lead us to believe that thisidea of a fundamental power is not problemati-cal, but that it possesses objective reality, andthus the systematic unity of the various powersor forces in a substance is demanded by theunderstanding and erected into an apodeicticor necessary principle. For, without havingattempted to discover the unity of the variouspowers existing in nature, nay, even after allour attempts have failed, we notwithstandingpresuppose that it does exist, and may be, soo-ner or later, discovered. And this reason does,not only, as in the case above adduced, withregard to the unity of substance, but wheremany substances, although all to a certain ex-

tent homogeneous, are discoverable, as in thecase of matter in general. Here also does reasonpresuppose the existence of the systematic uni-ty of various powers—inasmuch as particularlaws of nature are subordinate to general laws;and parsimony in principles is not merely aneconomical principle of reason, but an essentiallaw of nature.

We cannot understand, in fact, how a logicalprinciple of unity can of right exist, unless wepresuppose a transcendental principle, bywhich such a systematic unit—as a property ofobjects themselves—is regarded as necessary apriori. For with what right can reason, in itslogical exercise, require us to regard the varietyof forces which nature displays, as in effect adisguised unity, and to deduce them from onefundamental force or power, when she is free toadmit that it is just as possible that all forcesshould be different in kind, and that a systema-tic unity is not conformable to the design of

nature? In this view of the case, reason wouldbe proceeding in direct opposition to her owndestination, by setting as an aim an idea whichentirely conflicts with the procedure and arran-gement of nature. Neither can we assert thatreason has previously inferred this unity fromthe contingent nature of phenomena. For thelaw of reason which requires us to seek for thisunity is a necessary law, inasmuch as without itwe should not possess a faculty of reason, norwithout reason a consistent and self-accordantmode of employing the understanding, nor, inthe absence of this, any proper and sufficientcriterion of empirical truth. In relation to thiscriterion, therefore, we must suppose the ideaof the systematic unity of nature to possess ob-jective validity and necessity.

We find this transcendental presuppositionlurking in different forms in the principles ofphilosophers, although they have neither re-cognized it nor confessed to themselves its pre-

sence. That the diversities of individual thingsdo not exclude identity of species, that the va-rious species must be considered as merelydifferent determinations of a few genera, andthese again as divisions of still higher races,and so on—that, accordingly, a certain systema-tic unity of all possible empirical conceptions,in so far as they can be deduced from higherand more general conceptions, must be soughtfor, is a scholastic maxim or logical principle,without which reason could not be employedby us. For we can infer the particular from thegeneral, only in so far as general properties ofthings constitute the foundation upon whichthe particular rest.

That the same unity exists in nature is presup-posed by philosophers in the well-known scho-lastic maxim, which forbids us unnecessarily toaugment the number of entities or principles(entia praeter necessitatem non esse multipli-canda). This maxim asserts that nature herself

assists in the establishment of this unity of rea-son, and that the seemingly infinite diversity ofphenomena should not deter us from the ex-pectation of discovering beneath this diversitya unity of fundamental properties, of which theaforesaid variety is but a more or less determi-ned form. This unity, although a mere idea,thinkers have found it necessary rather to mo-derate the desire than to encourage it. It wasconsidered a great step when chemists wereable to reduce all salts to two main genera—acids and alkalis; and they regard this differen-ce as itself a mere variety, or different manifes-tation of one and the same fundamental mate-rial. The different kinds of earths (stones andeven metals) chemists have endeavoured toreduce to three, and afterwards to two; but still,not content with this advance, they cannot butthink that behind these diversities there lurksbut one genus—nay, that even salts and earthshave a common principle. It might be conjectu-red that this is merely an economical plan of

reason, for the purpose of sparing itself trouble,and an attempt of a purely hypothetical charac-ter, which, when successful, gives an appearan-ce of probability to the principle of explanationemployed by the reason. But a selfish purposeof this kind is easily to be distinguished fromthe idea, according to which every one presup-poses that this unity is in accordance with thelaws of nature, and that reason does not in thiscase request, but requires, although we are qui-te unable to determine the proper limits of thisunity.

If the diversity existing in phenomena—a di-versity not of form (for in this they may be si-milar) but of content—were so great that thesubtlest human reason could never by compa-rison discover in them the least similarity(which is not impossible), in this case the logi-cal law of genera would be without foundation,the conception of a genus, nay, all general con-ceptions would be impossible, and the faculty

of the understanding, the exercise of which isrestricted to the world of conceptions, couldnot exist. The logical principle of genera, accor-dingly, if it is to be applied to nature (by whichI mean objects presented to our senses), pre-supposes a transcendental principle. In accor-dance with this principle, homogeneity is ne-cessarily presupposed in the variety of pheno-mena (although we are unable to determine apriori the degree of this homogeneity), becausewithout it no empirical conceptions, and conse-quently no experience, would be possible.

The logical principle of genera, which demandsidentity in phenomena, is balanced by anotherprinciple—that of species, which requires varie-ty and diversity in things, notwithstandingtheir accordance in the same genus, and directsthe understanding to attend to the one no lessthan to the other. This principle (of the facultyof distinction) acts as a check upon the reasonand reason exhibits in this respect a double and

conflicting interest—on the one hand, the inter-est in the extent (the interest of generality) inrelation to genera; on the other, that of the con-tent (the interest of individuality) in relation tothe variety of species. In the former case, theunderstanding cogitates more under its concep-tions, in the latter it cogitates more in them.This distinction manifests itself likewise in thehabits of thought peculiar to natural philosop-hers, some of whom—the remarkably specula-tive heads—may be said to be hostile to hetero-geneity in phenomena, and have their eyes al-ways fixed on the unity of genera, while ot-hers—with a strong empirical tendency—aimunceasingly at the analysis of phenomena, andalmost destroy in us the hope of ever being ableto estimate the character of these according togeneral principles.

The latter mode of thought is evidently basedupon a logical principle, the aim of which is thesystematic completeness of all cognitions. This

principle authorizes me, beginning at the ge-nus, to descend to the various and diverse con-tained under it; and in this way extension, as inthe former case unity, is assured to the system.For if we merely examine the sphere of the con-ception which indicates a genus, we cannotdiscover how far it is possible to proceed in thedivision of that sphere; just as it is impossible,from the consideration of the space occupied bymatter, to determine how far we can proceed inthe division of it. Hence every genus must con-tain different species, and these again differentsubspecies; and as each of the latter must itselfcontain a sphere (must be of a certain extent, asa conceptus communis), reason demands thatno species or sub-species is to be considered asthe lowest possible. For a species or sub-species, being always a conception, which con-tains only what is common to a number of dif-ferent things, does not completely determineany individual thing, or relate immediately toit, and must consequently contain other con-

ceptions, that is, other sub-species under it.This law of specification may be thus expres-sed: entium varietates non temere sunt mi-nuendae.

But it is easy to see that this logical law wouldlikewise be without sense or application, wereit not based upon a transcendental law of speci-fication, which certainly does not require thatthe differences existing phenomena should beinfinite in number, for the logical principle,which merely maintains the indeterminatenessof the logical sphere of a conception, in relationto its possible division, does not authorize thisstatement; while it does impose upon the un-derstanding the duty of searching for subspe-cies to every species, and minor differences inevery difference. For, were there no lower con-ceptions, neither could there be any higher.Now the understanding cognizes only bymeans of conceptions; consequently, how farsoever it may proceed in division, never by

mere intuition, but always by lower and lowerconceptions. The cognition of phenomena intheir complete determination (which is possibleonly by means of the understanding) requiresan unceasingly continued specification of con-ceptions, and a progression to ever smaller dif-ferences, of which abstraction bad been madein the conception of the species, and still morein that of the genus.

This law of specification cannot be deducedfrom experience; it can never present us with aprinciple of so universal an application. Empi-rical specification very soon stops in its distinc-tion of diversities, and requires the guidance ofthe transcendental law, as a principle of thereason—a law which imposes on us the necessi-ty of never ceasing in our search for differences,even although these may not present themsel-ves to the senses. That absorbent earths are ofdifferent kinds could only be discovered byobeying the anticipatory law of reason, which

imposes upon the understanding the task ofdiscovering the differences existing betweenthese earths, and supposes that nature is richerin substances than our senses would indicate.The faculty of the understanding belongs to usjust as much under the presupposition of diffe-rences in the objects of nature, as under thecondition that these objects are homogeneous,because we could not possess conceptions, normake any use of our understanding, were notthe phenomena included under these concep-tions in some respects dissimilar, as well assimilar, in their character.

Reason thus prepares the sphere of the unders-tanding for the operations of this faculty: 1. Bythe principle of the homogeneity of the diversein higher genera; 2. By the principle of the va-riety of the homogeneous in lower species; and,to complete the systematic unity, it adds, 3. Alaw of the affinity of all conceptions whichprescribes a continuous transition from one

species to every other by the gradual increaseof diversity. We may term these the principlesof the homogeneity, the specification, and thecontinuity of forms. The latter results from theunion of the two former, inasmuch as we re-gard the systematic connection as complete inthought, in the ascent to higher genera, as wellas in the descent to lower species. For all diver-sities must be related to each other, as they allspring from one highest genus, descendingthrough the different gradations of a more andmore extended determination.

We may illustrate the systematic unity produ-ced by the three logical principles in the follo-wing manner. Every conception may be regar-ded as a point, which, as the standpoint of aspectator, has a certain horizon, which may besaid to enclose a number of things that may beviewed, so to speak, from that centre. Withinthis horizon there must be an infinite numberof other points, each of which has its own hori-

zon, smaller and more circumscribed; in otherwords, every species contains sub-species, ac-cording to the principle of specification, and thelogical horizon consists of smaller horizons(subspecies), but not of points (individuals),which possess no extent. But different horizonsor genera, which include under them so manyconceptions, may have one common horizon,from which, as from a mid-point, they may besurveyed; and we may proceed thus, till wearrive at the highest genus, or universal andtrue horizon, which is determined by the hig-hest conception, and which contains under it-self all differences and varieties, as genera, spe-cies, and subspecies.

To this highest standpoint I am conducted bythe law of homogeneity, as to all lower andmore variously-determined conceptions by thelaw of specification. Now as in this way thereexists no void in the whole extent of all possibleconceptions, and as out of the sphere of these

the mind can discover nothing, there arisesfrom the presupposition of the universal hori-zon above mentioned, and its complete divi-sion, the principle: Non datur vacuum forma-rum. This principle asserts that there are notdifferent primitive and highest genera, whichstand isolated, so to speak, from each other, butall the various genera are mere divisions andlimitations of one highest and universal genus;and hence follows immediately the principle:Datur continuum formarum. This principleindicates that all differences of species limiteach other, and do not admit of transition fromone to another by a saltus, but only throughsmaller degrees of the difference between theone species and the other. In one word, thereare no species or sub-species which (in the viewof reason) are the nearest possible to each other;intermediate species or sub-species being al-ways possible, the difference of which fromeach of the former is always smaller than thedifference existing between these.

The first law, therefore, directs us to avoid thenotion that there exist different primal genera,and enounces the fact of perfect homogeneity;the second imposes a check upon this tendencyto unity and prescribes the distinction of sub-species, before proceeding to apply our generalconceptions to individuals. The third unitesboth the former, by enouncing the fact ofhomogeneity as existing even in the most va-rious diversity, by means of the gradual transi-tion from one species to another. Thus it indica-tes a relationship between the different bran-ches or species, in so far as they all spring fromthe same stem.

But this logical law of the continuum specierum(formarum logicarum) presupposes a transcen-dental principle (lex continui in natura), wit-hout which the understanding might be ledinto error, by following the guidance of theformer, and thus perhaps pursuing a path con-trary to that prescribed by nature. This law

must, consequently, be based upon pure trans-cendental, and not upon empirical, considera-tions. For, in the latter case, it would come laterthan the system; whereas it is really itself theparent of all that is systematic in our cognitionof nature. These principles are not mere hypot-heses employed for the purpose of experimen-ting upon nature; although when any suchconnection is discovered, it forms a solidground for regarding the hypothetical unity asvalid in the sphere of nature—and thus theyare in this respect not without their use. But wego farther, and maintain that it is manifest thatthese principles of parsimony in fundamentalcauses, variety in effects, and affinity in phe-nomena, are in accordance both with reasonand nature, and that they are not mere methodsor plans devised for the purpose of assisting usin our observation of the external world.

But it is plain that this continuity of forms is amere idea, to which no adequate object can be

discovered in experience. And this for two rea-sons. First, because the species in nature arereally divided, and hence form quanta discreta;and, if the gradual progression through theiraffinity were continuous, the intermediatemembers lying between two given species mustbe infinite in number, which is impossible. Se-condly, because we cannot make any determi-nate empirical use of this law, inasmuch as itdoes not present us with any criterion of affini-ty which could aid us in determining how farwe ought to pursue the graduation of differen-ces: it merely contains a general indication thatit is our duty to seek for and, if possible, to dis-cover them.

When we arrange these principles of systematicunity in the order conformable to their em-ployment in experience, they will stand thus:Variety, Affinity, Unity, each of them, as ideas,being taken in the highest degree of their com-pleteness. Reason presupposes the existence of

cognitions of the understanding, which have adirect relation to experience, and aims at theideal unity of these cognitions—a unity whichfar transcends all experience or empirical no-tions. The affinity of the diverse, notwithstan-ding the differences existing between its parts,has a relation to things, but a still closer one tothe mere properties and powers of things. Forexample, imperfect experience may representthe orbits of the planets as circular. But we dis-cover variations from this course, and we pro-ceed to suppose that the planets revolve in apath which, if not a circle, is of a character verysimilar to it. That is to say, the movements ofthose planets which do not form a circle willapproximate more or less to the properties of acircle, and probably form an ellipse. The pathsof comets exhibit still greater variations, for, sofar as our observation extends, they do not re-turn upon their own course in a circle or ellipse.But we proceed to the conjecture that cometsdescribe a parabola, a figure which is closely

allied to the ellipse. In fact, a parabola is merelyan ellipse, with its longer axis produced to anindefinite extent. Thus these principles conductus to a unity in the genera of the forms of theseorbits, and, proceeding farther, to a unity asregards the cause of the motions of the heaven-ly bodies—that is, gravitation. But we go onextending our conquests over nature, and en-deavour to explain all seeming deviations fromthese rules, and even make additions to oursystem which no experience can ever substan-tiate—for example, the theory, in affinity withthat of ellipses, of hyperbolic paths of comets,pursuing which, these bodies leave our solarsystem and, passing from sun to sun, unite themost distant parts of the infinite universe,which is held together by the same movingpower.

The most remarkable circumstance connectedwith these principles is that they seem to betranscendental, and, although only containing

ideas for the guidance of the empirical exerciseof reason, and although this empirical em-ployment stands to these ideas in an asympto-tic relation alone (to use a mathematical term),that is, continually approximate, without everbeing able to attain to them, they possess, not-withstanding, as a priori synthetical proposi-tions, objective though undetermined validity,and are available as rules for possible experien-ce. In the elaboration of our experience, theymay also be employed with great advantage, asheuristic [Footnote: From the Greek, eurhioko.]principles. A transcendental deduction of themcannot be made; such a deduction being alwaysimpossible in the case of ideas, as has been al-ready shown.

We distinguished, in the Transcendental Analy-tic, the dynamical principles of the understan-ding, which are regulative principles of intui-tion, from the mathematical, which are consti-tutive principles of intuition. These dynamical

laws are, however, constitutive in relation toexperience, inasmuch as they render the con-ceptions without which experience could notexist possible a priori. But the principles of pu-re reason cannot be constitutive even in regardto empirical conceptions, because no sensuousschema corresponding to them can be discove-red, and they cannot therefore have an object inconcreto. Now, if I grant that they cannot beemployed in the sphere of experience, as consti-tutive principles, how shall I secure for thememployment and objective validity as regulati-ve principles, and in what way can they be soemployed?

The understanding is the object of reason, assensibility is the object of the understanding.The production of systematic unity in all theempirical operations of the understanding isthe proper occupation of reason; just as it is thebusiness of the understanding to connect thevarious content of phenomena by means of

conceptions, and subject them to empiricallaws. But the operations of the understandingare, without the schemata of sensibility, unde-termined; and, in the same manner, the unity ofreason is perfectly undetermined as regards theconditions under which, and the extent towhich, the understanding ought to carry thesystematic connection of its conceptions. But,although it is impossible to discover in intui-tion a schema for the complete systematic unityof all the conceptions of the understanding,there must be some analogon of this schema.This analogon is the idea of the maximum ofthe division and the connection of our cogni-tion in one principle. For we may have a de-terminate notion of a maximum and an absolu-tely perfect, all the restrictive conditions whichare connected with an indeterminate and va-rious content having been abstracted. Thus theidea of reason is analogous with a sensuousschema, with this difference, that the applica-tion of the categories to the schema of reason

does not present a cognition of any object (as isthe case with the application of the categoriesto sensuous schemata), but merely provides uswith a rule or principle for the systematic unityof the exercise of the understanding. Now, asevery principle which imposes upon the exerci-se of the understanding a priori compliancewith the rule of systematic unity also relates,although only in an indirect manner, to an ob-ject of experience, the principles of pure reasonwill also possess objective reality and validityin relation to experience. But they will not aimat determining our knowledge in regard to anyempirical object; they will merely indicate theprocedure, following which the empirical anddeterminate exercise of the understanding maybe in complete harmony and connection withitself—a result which is produced by its beingbrought into harmony with the principle ofsystematic unity, so far as that is possible, anddeduced from it.

I term all subjective principles, which are notderived from observation of the constitution ofan object, but from the interest which Reasonhas in producing a certain completeness in hercognition of that object, maxims of reason. Thusthere are maxims of speculative reason, whichare based solely upon its speculative interest,although they appear to be objective principles.

When principles which are really regulative areregarded as constitutive, and employed as ob-jective principles, contradictions must arise; butif they are considered as mere maxims, there isno room for contradictions of any kind, as theythen merely indicate the different interests ofreason, which occasion differences in the modeof thought. In effect, Reason has only one singleinterest, and the seeming contradiction existingbetween her maxims merely indicates a diffe-rence in, and a reciprocal limitation of, the met-hods by which this interest is satisfied.

This reasoner has at heart the interest of diver-sity—in accordance with the principle of speci-fication; another, the interest of unity—in ac-cordance with the principle of aggregation.Each believes that his judgement rests upon athorough insight into the subject he is exami-ning, and yet it has been influenced solely by agreater or less degree of adherence to some oneof the two principles, neither of which are ob-jective, but originate solely from the interest ofreason, and on this account to be termedmaxims rather than principles. When I observeintelligent men disputing about the distinctivecharacteristics of men, animals, or plants, andeven of minerals, those on the one side assu-ming the existence of certain national characte-ristics, certain well-defined and hereditary dis-tinctions of family, race, and so on, while theother side maintain that nature has endowedall races of men with the same faculties anddispositions, and that all differences are but theresult of external and accidental circumstan-

ces—I have only to consider for a moment thereal nature of the subject of discussion, to arriveat the conclusion that it is a subject far too deepfor us to judge of, and that there is little proba-bility of either party being able to speak from aperfect insight into and understanding of thenature of the subject itself. Both have, in reality,been struggling for the twofold interest of rea-son; the one maintaining the one interest, theother the other. But this difference between themaxims of diversity and unity may easily bereconciled and adjusted; although, so long asthey are regarded as objective principles, theymust occasion not only contradictions and po-lemic, but place hinderances in the way of theadvancement of truth, until some means is dis-covered of reconciling these conflicting inte-rests, and bringing reason into union and har-mony with itself.

The same is the case with the so-called law dis-covered by Leibnitz, and supported with re-

markable ability by Bonnet—the law of the con-tinuous gradation of created beings, which isnothing more than an inference from the prin-ciple of affinity; for observation and study ofthe order of nature could never present it to themind as an objective truth. The steps of thisladder, as they appear in experience, are too farapart from each other, and the so-called pettydifferences between different kinds of animalsare in nature commonly so wide separationsthat no confidence can be placed in such views(particularly when we reflect on the great varie-ty of things, and the ease with which we candiscover resemblances), and no faith in the lawswhich are said to express the aims and purpo-ses of nature. On the other hand, the method ofinvestigating the order of nature in the light ofthis principle, and the maxim which requires usto regard this order—it being still undetermi-ned how far it extends—as really existing innature, is beyond doubt a legitimate and exce-llent principle of reason—a principle which

extends farther than any experience or observa-tion of ours and which, without giving us anypositive knowledge of anything in the region ofexperience, guides us to the goal of systematicunity.

Of the Ultimate End of the Natural Dialectic ofHuman Reason.

The ideas of pure reason cannot be, of themsel-ves and in their own nature, dialectical; it isfrom their misemployment alone that fallaciesand illusions arise. For they originate in thenature of reason itself, and it is impossible thatthis supreme tribunal for all the rights andclaims of speculation should be itself undeser-ving of confidence and promotive of error. It isto be expected, therefore, that these ideas havea genuine and legitimate aim. It is true, the mobof sophists raise against reason the cry of in-

consistency and contradiction, and affect todespise the government of that faculty, becausethey cannot understand its constitution, whileit is to its beneficial influences alone that theyowe the position and the intelligence whichenable them to criticize and to blame its proce-dure.

We cannot employ an a priori conception withcertainty, until we have made a transcendentaldeduction therefore. The ideas of pure reasondo not admit of the same kind of deduction asthe categories. But if they are to possess theleast objective validity, and to represent anyt-hing but mere creations of thought (entia ratio-nis ratiocinantis), a deduction of them must bepossible. This deduction will complete the criti-cal task imposed upon pure reason; and it is tothis part Of our labours that we now proceed.

There is a great difference between a thing'sbeing presented to the mind as an object in anabsolute sense, or merely as an ideal object. In

the former case I employ my conceptions todetermine the object; in the latter case nothingis present to the mind but a mere schema,which does not relate directly to an object, noteven in a hypothetical sense, but which is use-ful only for the purpose of representing otherobjects to the mind, in a mediate and indirectmanner, by means of their relation to the ideain the intellect. Thus I say the conception of asupreme intelligence is a mere idea; that is tosay, its objective reality does not consist in thefact that it has an immediate relation to an ob-ject (for in this sense we have no means of esta-blishing its objective validity), it is merely aschema constructed according to the necessaryconditions of the unity of reason—the schemaof a thing in general, which is useful towardsthe production of the highest degree of syste-matic unity in the empirical exercise of reason,in which we deduce this or that object of expe-rience from the imaginary object of this idea, asthe ground or cause of the said object of expe-

rience. In this way, the idea is properly a heu-ristic, and not an ostensive, conception; it doesnot give us any information respecting theconstitution of an object, it merely indicateshow, under the guidance of the idea, we oughtto investigate the constitution and the relationsof objects in the world of experience. Now, if itcan be shown that the three kinds of transcen-dental ideas (psychological, cosmological, andtheological), although not relating directly toany object nor determining it, do nevertheless,on the supposition of the existence of an idealobject, produce systematic unity in the laws ofthe empirical employment of the reason, andextend our empirical cognition, without everbeing inconsistent or in opposition with it- itmust be a necessary maxim of reason to regula-te its procedure according to these ideas. Andthis forms the transcendental deduction of allspeculative ideas, not as constitutive principlesof the extension of our cognition beyond thelimits of our experience, but as regulative prin-

ciples of the systematic unity of empirical cog-nition, which is by the aid of these ideas arran-ged and emended within its own proper limits,to an extent unattainable by the operation ofthe principles of the understanding alone.

I shall make this plainer. Guided by the princi-ples involved in these ideas, we must, in thefirst place, so connect all the phenomena, ac-tions, and feelings of the mind, as if it were asimple substance, which, endowed with perso-nal identity, possesses a permanent existence(in this life at least), while its states, amongwhich those of the body are to be included asexternal conditions, are in continual change.Secondly, in cosmology, we must investigatethe conditions of all natural phenomena, inter-nal as well as external, as if they belonged to achain infinite and without any prime or supre-me member, while we do not, on this account,deny the existence of intelligible grounds ofthese phenomena, although we never employ

them to explain phenomena, for the simplereason that they are not objects of our cogni-tion. Thirdly, in the sphere of theology, wemust regard the whole system of possible expe-rience as forming an absolute, but dependentand sensuously-conditioned unity, and at thesame time as based upon a sole, supreme, andall-sufficient ground existing apart from theworld itself—a ground which is a self-subsistent, primeval and creative reason, inrelation to which we so employ our reason inthe field of experience, as if all objects drewtheir origin from that archetype of all reason. Inother words, we ought not to deduce the inter-nal phenomena of the mind from a simplethinking substance, but deduce them from eachother under the guidance of the regulative ideaof a simple being; we ought not to deduce thephenomena, order, and unity of the universefrom a supreme intelligence, but merely drawfrom this idea of a supremely wise cause the

rules which must guide reason in its connectionof causes and effects.

Now there is nothing to hinder us from admit-ting these ideas to possess an objective and hy-perbolic existence, except the cosmologicalideas, which lead reason into an antinomy: thepsychological and theological ideas are not an-tinomial. They contain no contradiction; andhow, then, can any one dispute their objectivereality, since he who denies it knows as littleabout their possibility as we who affirm? Andyet, when we wish to admit the existence of athing, it is not sufficient to convince ourselvesthat there is no positive obstacle in the way; forit cannot be allowable to regard mere creationsof thought, which transcend, though they donot contradict, all our conceptions, as real anddeterminate objects, solely upon the authorityof a speculative reason striving to compass itsown aims. They cannot, therefore, be admittedto be real in themselves; they can only possess a

comparative reality—that of a schema of theregulative principle of the systematic unity ofall cognition. They are to be regarded not asactual things, but as in some measure analo-gous to them. We abstract from the object of theidea all the conditions which limit the exerciseof our understanding, but which, on the otherhand, are the sole conditions of our possessinga determinate conception of any given thing.And thus we cogitate a something, of the realnature of which we have not the least concep-tion, but which we represent to ourselves asstanding in a relation to the whole system ofphenomena, analogous to that in which phe-nomena stand to each other.

By admitting these ideal beings, we do not rea-lly extend our cognitions beyond the objects ofpossible experience; we extend merely the em-pirical unity of our experience, by the aid ofsystematic unity, the schema of which is fur-nished by the idea, which is therefore valid—

not as a constitutive, but as a regulative princi-ple. For although we posit a thing correspon-ding to the idea—a something, an actual exis-tence—we do not on that account aim at theextension of our cognition by means of trans-cendent conceptions. This existence is purelyideal, and not objective; it is the mere expres-sion of the systematic unity which is to be theguide of reason in the field of experience. Thereare no attempts made at deciding what theground of this unity may be, or what the realnature of this imaginary being.

Thus the transcendental and only determinateconception of God, which is presented to us byspeculative reason, is in the strictest sense deis-tic. In other words, reason does not assure us ofthe objective validity of the conception; it mere-ly gives us the idea of something, on which thesupreme and necessary unity of all experienceis based. This something we cannot, followingthe analogy of a real substance, cogitate other-

wise than as the cause of all things operating inaccordance with rational laws, if we regard it asan individual object; although we should restcontented with the idea alone as a regulativeprinciple of reason, and make no attempt atcompleting the sum of the conditions imposedby thought. This attempt is, indeed, inconsis-tent with the grand aim of complete systematicunity in the sphere of cognition—a unity towhich no bounds are set by reason.

Hence it happens that, admitting a divinebeing, I can have no conception of the internalpossibility of its perfection, or of the necessityof its existence. The only advantage of this ad-mission is that it enables me to answer all otherquestions relating to the contingent, and to givereason the most complete satisfaction as re-gards the unity which it aims at attaining in theworld of experience. But I cannot satisfy reasonwith regard to this hypothesis itself; and thisproves that it is not its intelligence and insight

into the subject, but its speculative interest alo-ne which induces it to proceed from a pointlying far beyond the sphere of our cognition,for the purpose of being able to consider allobjects as parts of a systematic whole.

Here a distinction presents itself, in regard tothe way in which we may cogitate a presuppo-sition—a distinction which is somewhat subtle,but of great importance in transcendental phi-losophy. I may have sufficient grounds to ad-mit something, or the existence of something,in a relative point of view (suppositio relativa),without being justified in admitting it in anabsolute sense (suppositio absoluta). This dis-tinction is undoubtedly requisite, in the case ofa regulative principle, the necessity of whichwe recognize, though we are ignorant of thesource and cause of that necessity, and whichwe assume to be based upon some ultimateground, for the purpose of being able to cogita-te the universality of the principle in a more

determinate way. For example, I cogitate theexistence of a being corresponding to a puretranscendental idea. But I cannot admit thatthis being exists absolutely and in itself, becau-se all of the conceptions by which I can cogitatean object in a determinate manner fall short ofassuring me of its existence; nay, the conditionsof the objective validity of my conceptions areexcluded by the idea—by the very fact of itsbeing an idea. The conceptions of reality, subs-tance, causality, nay, even that of necessity inexistence, have no significance out of the sphe-re of empirical cognition, and cannot, beyondthat sphere, determine any object. They may,accordingly, be employed to explain the possi-bility of things in the world of sense, but theyare utterly inadequate to explain the possibilityof the universe itself considered as a whole;because in this case the ground of explanationmust lie out of and beyond the world, and can-not, therefore, be an object of possible experien-ce. Now, I may admit the existence of an in-

comprehensible being of this nature—the objectof a mere idea, relatively to the world of sense;although I have no ground to admit its existen-ce absolutely and in itself. For if an idea (that ofa systematic and complete unity, of which Ishall presently speak more particularly) lies atthe foundation of the most extended empiricalemployment of reason, and if this idea cannotbe adequately represented in concreto, alt-hough it is indispensably necessary for the ap-proximation of empirical unity to the highestpossible degree—I am not only authorized, butcompelled, to realize this idea, that is, to posit areal object corresponding thereto. But I cannotprofess to know this object; it is to me merely asomething, to which, as the ground of systema-tic unity in cognition, I attribute such propertiesas are analogous to the conceptions employedby the understanding in the sphere of experien-ce. Following the analogy of the notions of rea-lity, substance, causality, and necessity, I cogi-tate a being, which possesses all these attributes

in the highest degree; and, as this idea is theoffspring of my reason alone, I cogitate thisbeing as self-subsistent reason, and as the causeof the universe operating by means of ideas ofthe greatest possible harmony and unity. Thus Iabstract all conditions that would limit myidea, solely for the purpose of rendering syste-matic unity possible in the world of empiricaldiversity, and thus securing the widest possibleextension for the exercise of reason in thatsphere. This I am enabled to do, by regardingall connections and relations in the world ofsense, as if they were the dispositions of a su-preme reason, of which our reason is but a faintimage. I then proceed to cogitate this SupremeBeing by conceptions which have, properly, nomeaning or application, except in the world ofsense. But as I am authorized to employ thetranscendental hypothesis of such a being in arelative respect alone, that is, as the substratumof the greatest possible unity in experience—Imay attribute to a being which I regard as dis-

tinct from the world, such properties as belongsolely to the sphere of sense and experience.For I do not desire, and am not justified in desi-ring, to cognize this object of my idea, as itexists in itself; for I possess no conceptions suf-ficient for or task, those of reality, substance,causality, nay, even that of necessity in existen-ce, losing all significance, and becoming merelythe signs of conceptions, without content andwithout applicability, when I attempt to carrythem beyond the limits of the world of sense. Icogitate merely the relation of a perfectly unk-nown being to the greatest possible systematicunity of experience, solely for the purpose ofemploying it as the schema of the regulativeprinciple which directs reason in its empiricalexercise.

It is evident, at the first view, that we cannotpresuppose the reality of this transcendentalobject, by means of the conceptions of reality,substance, causality, and so on, because these

conceptions cannot be applied to anything thatis distinct from the world of sense. Thus thesupposition of a Supreme Being or cause is pu-rely relative; it is cogitated only in behalf of thesystematic unity of experience; such a being isbut a something, of whose existence in itself wehave not the least conception. Thus, too, it be-comes sufficiently manifest why we requiredthe idea of a necessary being in relation to ob-jects given by sense, although we can neverhave the least conception of this being, or of itsabsolute necessity.

And now we can clearly perceive the result ofour transcendental dialectic, and the properaim of the ideas of pure reason—which becomedialectical solely from misunderstanding andinconsiderateness. Pure reason is, in fact, occu-pied with itself, and not with any object. Ob-jects are not presented to it to be embraced inthe unity of an empirical conception; it is onlythe cognitions of the understanding that are

presented to it, for the purpose of receiving theunity of a rational conception, that is, of beingconnected according to a principle. The unity ofreason is the unity of system; and this systema-tic unity is not an objective principle, extendingits dominion over objects, but a subjectivemaxim, extending its authority over the empiri-cal cognition of objects. The systematic connec-tion which reason gives to the empirical em-ployment of the understanding not only ad-vances the extension of that employment, butensures its correctness, and thus the principleof a systematic unity of this nature is also objec-tive, although only in an indefinite respect(principium vagum). It is not, however, a cons-titutive principle, determining an object towhich it directly relates; it is merely a regulati-ve principle or maxim, advancing and strengt-hening the empirical exercise of reason, by theopening up of new paths of which the unders-tanding is ignorant, while it never conflicts

with the laws of its exercise in the sphere ofexperience.

But reason cannot cogitate this systematic uni-ty, without at the same time cogitating an ob-ject of the idea—an object that cannot be pre-sented in any experience, which contains noconcrete example of a complete systematic uni-ty. This being (ens rationis ratiocinatae) is the-refore a mere idea and is not assumed to be athing which is real absolutely and in itself. Onthe contrary, it forms merely the problematicalfoundation of the connection which the mindintroduces among the phenomena of the sen-suous world. We look upon this connection, inthe light of the above-mentioned idea, as if itdrew its origin from the supposed being whichcorresponds to the idea. And yet all we aim atis the possession of this idea as a secure foun-dation for the systematic unity of experience—aunity indispensable to reason, advantageous to

the understanding, and promotive of the inte-rests of empirical cognition.

We mistake the true meaning of this idea whenwe regard it as an enouncement, or even as ahypothetical declaration of the existence of areal thing, which we are to regard as the originor ground of a systematic constitution of theuniverse. On the contrary, it is left completelyundetermined what the nature or properties ofthis so-called ground may be. The idea is mere-ly to be adopted as a point of view, from whichthis unity, so essential to reason and so benefi-cial to the understanding, may be regarded asradiating. In one word, this transcendentalthing is merely the schema of a regulative prin-ciple, by means of which Reason, so far as inher lies, extends the dominion of systematicunity over the whole sphere of experience.

The first object of an idea of this kind is the ego,considered merely as a thinking nature or soul.If I wish to investigate the properties of a thin-

king being, I must interrogate experience. But Ifind that I can apply none of the categories tothis object, the schema of these categories,which is the condition of their application,being given only in sensuous intuition. But Icannot thus attain to the cognition of a systema-tic unity of all the phenomena of the internalsense. Instead, therefore, of an empirical con-ception of what the soul really is, reason takesthe conception of the empirical unity of allthought, and, by cogitating this unity as un-conditioned and primitive, constructs the ratio-nal conception or idea of a simple substancewhich is in itself unchangeable, possessing per-sonal identity, and in connection with otherreal things external to it; in one word, it cons-tructs the idea of a simple self-subsistent inte-lligence. But the real aim of reason in this pro-cedure is the attainment of principles of syste-matic unity for the explanation of the pheno-mena of the soul. That is, reason desires to beable to represent all the determinations of the

internal sense as existing in one subject, all po-wers as deduced from one fundamental power,all changes as mere varieties in the condition ofa being which is permanent and always thesame, and all phenomena in space as entirelydifferent in their nature from the procedure ofthought. Essential simplicity (with the otherattributes predicated of the ego) is regarded asthe mere schema of this regulative principle; itis not assumed that it is the actual ground ofthe properties of the soul. For these propertiesmay rest upon quite different grounds, ofwhich we are completely ignorant; just as theabove predicates could not give us any know-ledge of the soul as it is in itself, even if we re-garded them as valid in respect of it, inasmuchas they constitute a mere idea, which cannot berepresented in concreto. Nothing but good canresult from a psychological idea of this kind, ifwe only take proper care not to consider it asmore than an idea; that is, if we regard it asvalid merely in relation to the employment of

reason, in the sphere of the phenomena of thesoul. Under the guidance of this idea, or prin-ciple, no empirical laws of corporeal phenome-na are called in to explain that which is a phe-nomenon of the internal sense alone; no windyhypotheses of the generation, annihilation, andpalingenesis of souls are admitted. Thus theconsideration of this object of the internal senseis kept pure, and unmixed with heterogeneouselements; while the investigation of reasonaims at reducing all the grounds of explanationemployed in this sphere of knowledge to a sin-gle principle. All this is best effected, nay, can-not be effected otherwise than by means ofsuch a schema, which requires us to regard thisideal thing as an actual existence. The psycho-logical idea is, therefore, meaningless andinapplicable, except as the schema of a regula-tive conception. For, if I ask whether the soul isnot really of a spiritual nature—it is a questionwhich has no meaning. From such a conceptionhas been abstracted, not merely all corporeal

nature, but all nature, that is, all the predicatesof a possible experience; and consequently, allthe conditions which enable us to cogitate anobject to this conception have disappeared. But,if these conditions are absent, it is evident thatthe conception is meaningless.

The second regulative idea of speculative rea-son is the conception of the universe. For natu-re is properly the only object presented to us, inregard to which reason requires regulativeprinciples. Nature is twofold—thinking andcorporeal nature. To cogitate the latter in re-gard to its internal possibility, that is, to deter-mine the application of the categories to it, noidea is required—no representation whichtranscends experience. In this sphere, therefore,an idea is impossible, sensuous intuition beingour only guide; while, in the sphere of psycho-logy, we require the fundamental idea (I),which contains a priori a certain form ofthought namely, the unity of the ego. Pure rea-

son has, therefore, nothing left but nature ingeneral, and the completeness of conditions innature in accordance with some principle. Theabsolute totality of the series of these condi-tions is an idea, which can never be fully reali-zed in the empirical exercise of reason, while itis serviceable as a rule for the procedure of rea-son in relation to that totality. It requires us, inthe explanation of given phenomena (in theregress or ascent in the series), to proceed as ifthe series were infinite in itself, that is, wereprolonged in indefinitum,; while on the otherhand, where reason is regarded as itself thedetermining cause (in the region of freedom),we are required to proceed as if we had notbefore us an object of sense, but of the pureunderstanding. In this latter case, the condi-tions do not exist in the series of phenomena,but may be placed quite out of and beyond it,and the series of conditions may be regarded asif it had an absolute beginning from an intelli-gible cause. All this proves that the cosmologi-

cal ideas are nothing but regulative principles,and not constitutive; and that their aim is not torealize an actual totality in such series. The fulldiscussion of this subject will be found in itsproper place in the chapter on the antinomy ofpure reason.

The third idea of pure reason, containing thehypothesis of a being which is valid merely as arelative hypothesis, is that of the one and all-sufficient cause of all cosmological series, inother words, the idea of God. We have not theslightest ground absolutely to admit the exis-tence of an object corresponding to this idea;for what can empower or authorize us to affirmthe existence of a being of the highest perfec-tion—a being whose existence is absolutelynecessary—merely because we possess the con-ception of such a being? The answer is: It is theexistence of the world which renders this hy-pothesis necessary. But this answer makes itperfectly evident that the idea of this being, like

all other speculative ideas, is essentially not-hing more than a demand upon reason that itshall regulate the connection which it and itssubordinate faculties introduce into the phe-nomena of the world by principles of systema-tic unity and, consequently, that it shall regardall phenomena as originating from one all-embracing being, as the supreme and all-sufficient cause. From this it is plain that theonly aim of reason in this procedure is the esta-blishment of its own formal rule for the exten-sion of its dominion in the world of experience;that it does not aim at an extension of its cogni-tion beyond the limits of experience; and that,consequently, this idea does not contain anyconstitutive principle.

The highest formal unity, which is based uponideas alone, is the unity of all things—a unity inaccordance with an aim or purpose; and thespeculative interest of reason renders it neces-sary to regard all order in the world as if it ori-

ginated from the intention and design of a su-preme reason. This principle unfolds to theview of reason in the sphere of experience newand enlarged prospects, and invites it to con-nect the phenomena of the world according toteleological laws, and in this way to attain tothe highest possible degree of systematic unity.The hypothesis of a supreme intelligence, as thesole cause of the universe—an intelligencewhich has for us no more than an ideal existen-ce—is accordingly always of the greatest servi-ce to reason. Thus, if we presuppose, in relationto the figure of the earth (which is round, butsomewhat flattened at the poles),* or that ofmountains or seas, wise designs on the part ofan author of the universe, we cannot fail to ma-ke, by the light of this supposition, a greatnumber of interesting discoveries. If we keep tothis hypothesis, as a principle which is purelyregulative, even error cannot be very detrimen-tal. For, in this case, error can have no moreserious consequences than that, where we ex-

pected to discover a teleological connection(nexus finalis), only a mechanical or physicalconnection appears. In such a case, we merelyfail to find the additional form of unity we ex-pected, but we do not lose the rational unitywhich the mind requires in its procedure inexperience. But even a miscarriage of this sortcannot affect the law in its general and teleolo-gical relations. For although we may convict ananatomist of an error, when he connects thelimb of some animal with a certain purpose, itis quite impossible to prove in a single case thatany arrangement of nature, be it what it may, isentirely without aim or design. And thus medi-cal physiology, by the aid of a principle presen-ted to it by pure reason, extends its very limitedempirical knowledge of the purposes of thedifferent parts of an organized body so far thatit may be asserted with the utmost confidence,and with the approbation of all reflecting men,that every organ or bodily part of an animal hasits use and answers a certain design. Now, this

is a supposition which, if regarded as of a cons-titutive character, goes much farther than anyexperience or observation of ours can justify.Hence it is evident that it is nothing more thana regulative principle of reason, which aims atthe highest degree of systematic unity, by theaid of the idea of a causality according to de-sign in a supreme cause—a cause which it re-gards as the highest intelligence.

[*Footnote: The advantages which a circularform, in the case of the earth, has over everyother, are well known. But few are aware thatthe slight flattening at the poles, which gives itthe figure of a spheroid, is the only cause whichprevents the elevations of continents or even ofmountains, perhaps thrown up by some inter-nal convulsion, from continually altering theposition of the axis of the earth—and that tosome considerable degree in a short time. Thegreat protuberance of the earth under the Equa-tor serves to overbalance the impetus of all ot-

her masses of earth, and thus to preserve theaxis of the earth, so far as we can observe, in itspresent position. And yet this wise arrange-ment has been unthinkingly explained from theequilibrium of the formerly fluid mass.]

If, however, we neglect this restriction of theidea to a purely regulative influence, reason isbetrayed into numerous errors. For it has thenleft the ground of experience, in which aloneare to be found the criteria of truth, and hasventured into the region of the incomprehensi-ble and unsearchable, on the heights of which itloses its power and collectedness, because it hascompletely severed its connection with expe-rience.

The first error which arises from our employingthe idea of a Supreme Being as a constitutive(in repugnance to the very nature of an idea),and not as a regulative principle, is the error ofinactive reason (ignava ratio).* We may so termevery principle which requires us to regard our

investigations of nature as absolutely complete,and allows reason to cease its inquiries, as if ithad fully executed its task. Thus the psycholo-gical idea of the ego, when employed as a cons-titutive principle for the explanation of the phe-nomena of the soul, and for the extension ofour knowledge regarding this subject beyondthe limits of experience—even to the conditionof the soul after death—is convenient enoughfor the purposes of pure reason, but detrimen-tal and even ruinous to its interests in the sphe-re of nature and experience. The dogmatizingspiritualist explains the unchanging unity ofour personality through all changes of condi-tion from the unity of a thinking substance, theinterest which we take in things and events thatcan happen only after our death, from a cons-ciousness of the immaterial nature of our thin-king subject, and so on. Thus he dispenses withall empirical investigations into the cause ofthese internal phenomena, and with all possibleexplanations of them upon purely natural

grounds; while, at the dictation of a transcen-dent reason, he passes by the immanent sourcesof cognition in experience, greatly to his ownease and convenience, but to the sacrifice of all,genuine insight and intelligence. These prejudi-cial consequences become still more evident, inthe case of the dogmatical treatment of our ideaof a Supreme Intelligence, and the theologicalsystem of nature (physico-theology) which isfalsely based upon it. For, in this case, the aimswhich we observe in nature, and often thosewhich we merely fancy to exist, make the inves-tigation of causes a very easy task, by directingus to refer such and such phenomena immedia-tely to the unsearchable will and counsel of theSupreme Wisdom, while we ought to investiga-te their causes in the general laws of the me-chanism of matter. We are thus recommendedto consider the labour of reason as ended, whenwe have merely dispensed with its employ-ment, which is guided surely and safely onlyby the order of nature and the series of changes

in the world- which are arranged according toimmanent and general laws. This error may beavoided, if we do not merely consider from theview-point of final aims certain parts of nature,such as the division and structure of a conti-nent, the constitution and direction of certainmountain-chains, or even the organization exis-ting in the vegetable and animal kingdoms, butlook upon this systematic unity of nature in aperfectly general way, in relation to the idea ofa Supreme Intelligence. If we pursue this advi-ce, we lay as a foundation for all investigationthe conformity to aims of all phenomena ofnature in accordance with universal laws, forwhich no particular arrangement of nature isexempt, but only cognized by us with more orless difficulty; and we possess a regulativeprinciple of the systematic unity of a teleologi-cal connection, which we do not attempt toanticipate or predetermine. All that we do, andought to do, is to follow out the physico-mechanical connection in nature according to

general laws, with the hope of discovering,sooner or later, the teleological connection also.Thus, and thus only, can the principle of finalunity aid in the extension of the employment ofreason in the sphere of experience, withoutbeing in any case detrimental to its interests.

[*Footnote: This was the term applied by theold dialecticians to a sophistical argument,which ran thus: If it is your fate to die of thisdisease, you will die, whether you employ aphysician or not. Cicero says that this mode ofreasoning has received this appellation, becau-se, if followed, it puts an end to the employ-ment of reason in the affairs of life. For a simi-lar reason, I have applied this designation tothe sophistical argument of pure reason.]

The second error which arises from the miscon-ception of the principle of systematic unity isthat of perverted reason (perversa ratio, uste-ron roteron rationis). The idea of systematicunity is available as a regulative principle in

the connection of phenomena according to ge-neral natural laws; and, how far soever wehave to travel upon the path of experience todiscover some fact or event, this idea requiresus to believe that we have approached all themore nearly to the completion of its use in thesphere of nature, although that completion cannever be attained. But this error reverses theprocedure of reason. We begin by hypostati-zing the principle of systematic unity, and bygiving an anthropomorphic determination tothe conception of a Supreme Intelligence, andthen proceed forcibly to impose aims upon na-ture. Thus not only does teleology, which oughtto aid in the completion of unity in accordancewith general laws, operate to the destruction ofits influence, but it hinders reason from attai-ning its proper aim, that is, the proof, uponnatural grounds, of the existence of a supremeintelligent cause. For, if we cannot presupposesupreme finality in nature a priori, that is, asessentially belonging to nature, how can we be

directed to endeavour to discover this unityand, rising gradually through its different de-grees, to approach the supreme perfection of anauthor of all—a perfection which is absolutelynecessary, and therefore cognizable a priori?The regulative principle directs us to presuppo-se systematic unity absolutely and, consequen-tly, as following from the essential nature ofthings—but only as a unity of nature, not mere-ly cognized empirically, but presupposed apriori, although only in an indeterminate man-ner. But if I insist on basing nature upon thefoundation of a supreme ordaining Being, theunity of nature is in effect lost. For, in this case,it is quite foreign and unessential to the natureof things, and cannot be cognized from the ge-neral laws of nature. And thus arises a viciouscircular argument, what ought to have beenproved having been presupposed.

To take the regulative principle of systematicunity in nature for a constitutive principle, and

to hypostatize and make a cause out of thatwhich is properly the ideal ground of the con-sistent and harmonious exercise of reason, in-volves reason in inextricable embarrassments.The investigation of nature pursues its ownpath under the guidance of the chain of naturalcauses, in accordance with the general laws ofnature, and ever follows the light of the idea ofan author of the universe—not for the purposeof deducing the finality, which it constantlypursues, from this Supreme Being, but to attainto the cognition of his existence from the finali-ty which it seeks in the existence of the pheno-mena of nature, and, if possible, in that of allthings to cognize this being, consequently, asabsolutely necessary. Whether this latter pur-pose succeed or not, the idea is and must al-ways be a true one, and its employment, whenmerely regulative, must always be accompa-nied by truthful and beneficial results.

Complete unity, in conformity with aims, cons-titutes absolute perfection. But if we do not findthis unity in the nature of the things which goto constitute the world of experience, that is, ofobjective cognition, consequently in the univer-sal and necessary laws of nature, how can weinfer from this unity the idea of the supremeand absolutely necessary perfection of a primalbeing, which is the origin of all causality? Thegreatest systematic unity, and consequentlyteleological unity, constitutes the very founda-tion of the possibility of the most extended em-ployment of human reason. The idea of unity istherefore essentially and indissolubly connec-ted with the nature of our reason. This idea is alegislative one; and hence it is very natural thatwe should assume the existence of a legislativereason corresponding to it, from which the sys-tematic unity of nature- the object of the opera-tions of reason—must be derived.

In the course of our discussion of the antino-mies, we stated that it is always possible toanswer all the questions which pure reasonmay raise; and that the plea of the limited natu-re of our cognition, which is unavoidable andproper in many questions regarding naturalphenomena, cannot in this case be admitted,because the questions raised do not relate to thenature of things, but are necessarily originatedby the nature of reason itself, and relate to itsown internal constitution. We can now esta-blish this assertion, which at first sight appea-red so rash, in relation to the two questions inwhich reason takes the greatest interest, andthus complete our discussion of the dialectic ofpure reason.

If, then, the question is asked, in relation totranscendental theology,* first, whether there isanything distinct from the world, which con-tains the ground of cosmical order and connec-tion according to general laws? The answer is:

Certainly. For the world is a sum of phenome-na; there must, therefore, be some transcenden-tal basis of these phenomena, that is, a basiscogitable by the pure understanding alone. If,secondly, the question is asked whether thisbeing is substance, whether it is of the greatestreality, whether it is necessary, and so forth? Ianswer that this question is utterly withoutmeaning. For all the categories which aid me informing a conception of an object cannot beemployed except in the world of sense, and arewithout meaning when not applied to objectsof actual or possible experience. Out of thissphere, they are not properly conceptions, butthe mere marks or indices of conceptions,which we may admit, although they cannot,without the help of experience, help us to un-derstand any subject or thing. If, thirdly, thequestion is whether we may not cogitate thisbeing, which is distinct from the world, in ana-logy with the objects of experience? The answeris: Undoubtedly, but only as an ideal, and not

as a real object. That is, we must cogitate it onlyas an unknown substratum of the systematicunity, order, and finality of the world—a unitywhich reason must employ as the regulativeprinciple of its investigation of nature. Nay,more, we may admit into the idea certain anth-ropomorphic elements, which are promotive ofthe interests of this regulative principle. For it isno more than an idea, which does not relatedirectly to a being distinct from the world, butto the regulative principle of the systematicunity of the world, by means, however, of aschema of this unity—the schema of a SupremeIntelligence, who is the wisely-designing aut-hor of the universe. What this basis of cosmicalunity may be in itself, we know not—we cannotdiscover from the idea; we merely know howwe ought to employ the idea of this unity, inrelation to the systematic operation of reason inthe sphere of experience.

[*Footnote: After what has been said of the psy-chological idea of the ego and its proper em-ployment as a regulative principle of the opera-tions of reason, I need not enter into detailsregarding the transcendental illusion by whichthe systematic unity of all the various pheno-mena of the internal sense is hypostatized. Theprocedure is in this case very similar to thatwhich has been discussed in our remarks onthe theological ideal.]

But, it will be asked again, can we on thesegrounds, admit the existence of a wise and om-nipotent author of the world? Without doubt;and not only so, but we must assume the exis-tence of such a being. But do we thus extendthe limits of our knowledge beyond the field ofpossible experience? By no means. For we havemerely presupposed a something, of which wehave no conception, which we do not know asit is in itself; but, in relation to the systematicdisposition of the universe, which we must

presuppose in all our observation of nature, wehave cogitated this unknown being in analogywith an intelligent existence (an empirical con-ception), that is to say, we have endowed itwith those attributes, which, judging from thenature of our own reason, may contain theground of such a systematic unity. This idea istherefore valid only relatively to the employ-ment in experience of our reason. But if we at-tribute to it absolute and objective validity, weoverlook the fact that it is merely an ideal beingthat we cogitate; and, by setting out from a ba-sis which is not determinable by considerationsdrawn from experience, we place ourselves in aposition which incapacitates us from applyingthis principle to the empirical employment ofreason.

But, it will be asked further, can I make any useof this conception and hypothesis in my inves-tigations into the world and nature? Yes, forthis very purpose was the idea established by

reason as a fundamental basis. But may I re-gard certain arrangements, which seemed tohave been made in conformity with some fixedaim, as the arrangements of design, and lookupon them as proceeding from the divine will,with the intervention, however, of certain otherparticular arrangements disposed to that end?Yes, you may do so; but at the same time youmust regard it as indifferent, whether it is as-serted that divine wisdom has disposed allthings in conformity with his highest aims, orthat the idea of supreme wisdom is a regulativeprinciple in the investigation of nature, and atthe same time a principle of the systematic uni-ty of nature according to general laws, even inthose cases where we are unable to discoverthat unity. In other words, it must be perfectlyindifferent to you whether you say, when youhave discovered this unity: God has wisely wi-lled it so; or: Nature has wisely arranged this.For it was nothing but the systematic unity,which reason requires as a basis for the investi-

gation of nature, that justified you in acceptingthe idea of a supreme intelligence as a schemafor a regulative principle; and, the farther youadvance in the discovery of design and finality,the more certain the validity of your idea. But,as the whole aim of this regulative principlewas the discovery of a necessary and systema-tic unity in nature, we have, in so far as we at-tain this, to attribute our success to the idea of aSupreme Being; while, at the same time, wecannot, without involving ourselves in contra-dictions, overlook the general laws of nature, asit was in reference to them alone that this ideawas employed. We cannot, I say, overlook thegeneral laws of nature, and regard this confor-mity to aims observable in nature as contingentor hyperphysical in its origin; inasmuch as the-re is no ground which can justify us in the ad-mission of a being with such properties distinctfrom and above nature. All that we are authori-zed to assert is that this idea may be employedas a principle, and that the properties of the

being which is assumed to correspond to it maybe regarded as systematically connected in ana-logy with the causal determination of pheno-mena.

For the same reasons we are justified in intro-ducing into the idea of the supreme cause otheranthropomorphic elements (for without thesewe could not predicate anything of it); we mayregard it as allowable to cogitate this cause as abeing with understanding, the feelings of plea-sure and displeasure, and faculties of desireand will corresponding to these. At the sametime, we may attribute to this being infiniteperfection—a perfection which necessarilytranscends that which our knowledge of theorder and design in the world authorize us topredicate of it. For the regulative law of syste-matic unity requires us to study nature on thesupposition that systematic and final unity ininfinitum is everywhere discoverable, even inthe highest diversity. For, although we may

discover little of this cosmical perfection, it be-longs to the legislative prerogative of reason torequire us always to seek for and to expect it;while it must always be beneficial to instituteall inquiries into nature in accordance with thisprinciple. But it is evident that, by this idea of asupreme author of all, which I place as thefoundation of all inquiries into nature, I do notmean to assert the existence of such a being, orthat I have any knowledge of its existence; and,consequently, I do not really deduce anythingfrom the existence of this being, but merelyfrom its idea, that is to say, from the nature ofthings in this world, in accordance with thisidea. A certain dim consciousness of the trueuse of this idea seems to have dictated to thephilosophers of all times the moderate langua-ge used by them regarding the cause of theworld. We find them employing the expres-sions wisdom and care of nature, and divinewisdom, as synonymous—nay, in purely spe-culative discussions, preferring the former, be-

cause it does not carry the appearance of grea-ter pretensions than such as we are entitled tomake, and at the same time directs reason to itsproper field of action—nature and her pheno-mena.

Thus, pure reason, which at first seemed topromise us nothing less than the extension ofour cognition beyond the limits of experience,is found, when thoroughly examined, to con-tain nothing but regulative principles, the vir-tue and function of which is to introduce intoour cognition a higher degree of unity than theunderstanding could of itself. These principles,by placing the goal of all our struggles at sogreat a distance, realize for us the most tho-rough connection between the different parts ofour cognition, and the highest degree of syste-matic unity. But, on the other hand, if misun-derstood and employed as constitutive princi-ples of transcendent cognition, they become theparents of illusions and contradictions, while

pretending to introduce us to new regions ofknowledge.

Thus all human cognition begins with intui-tions, proceeds from thence to conceptions, andends with ideas. Although it possesses, in rela-tion to all three elements, a priori sources ofcognition, which seemed to transcend the limitsof all experience, a thoroughgoing criticismdemonstrates that speculative reason can never,by the aid of these elements, pass the bounds ofpossible experience, and that the proper desti-nation of this highest faculty of cognition is toemploy all methods, and all the principles ofthese methods, for the purpose of penetratinginto the innermost secrets of nature, by the aidof the principles of unity (among all kinds ofwhich teleological unity is the highest), while itought not to attempt to soar above the sphereof experience, beyond which there lies noughtfor us but the void inane. The critical examina-tion, in our Transcendental Analytic, of all the

propositions which professed to extend cogni-tion beyond the sphere of experience, complete-ly demonstrated that they can only conduct usto a possible experience. If we were not dis-trustful even of the clearest abstract theorems,if we were not allured by specious and invitingprospects to escape from the constraining po-wer of their evidence, we might spare ourselvesthe laborious examination of all the dialecticalarguments which a transcendent reason addu-ces in support of its pretensions; for we shouldknow with the most complete certainty that,however honest such professions might be,they are null and valueless, because they relateto a kind of knowledge to which no man can byany possibility attain. But, as there is no end todiscussion, if we cannot discover the true causeof the illusions by which even the wisest aredeceived, and as the analysis of all our trans-cendent cognition into its elements is of itself ofno slight value as a psychological study, whileit is a duty incumbent on every philosopher—it

was found necessary to investigate the dialecti-cal procedure of reason in its primary sources.And as the inferences of which this dialectic isthe parent are not only deceitful, but naturallypossess a profound interest for humanity, itwas advisable at the same time, to give a fullaccount of the momenta of this dialectical pro-cedure, and to deposit it in the archives ofhuman reason, as a warning to all future me-taphysicians to avoid these causes of speculati-ve error.

II.

TRANSCENDENTAL DOCTRINE OFMETHOD.

If we regard the sum of the cognition of purespeculative reason as an edifice, the idea ofwhich, at least, exists in the human mind, it

may be said that we have in the TranscendentalDoctrine of Elements examined the materialsand determined to what edifice these belong,and what its height and stability. We havefound, indeed, that, although we had purposedto build for ourselves a tower which shouldreach to Heaven, the supply of materials suffi-ced merely for a habitation, which was spa-cious enough for all terrestrial purposes, andhigh enough to enable us to survey the levelplain of experience, but that the bold underta-king designed necessarily failed for want ofmaterials—not to mention the confusion oftongues, which gave rise to endless disputesamong the labourers on the plan of the edifice,and at last scattered them over all the world,each to erect a separate building for himself,according to his own plans and his own inclina-tions. Our present task relates not to the mate-rials, but to the plan of an edifice; and, as wehave had sufficient warning not to ventureblindly upon a design which may be found to

transcend our natural powers, while, at thesame time, we cannot give up the intention oferecting a secure abode for the mind, we mustproportion our design to the material which ispresented to us, and which is, at the same time,sufficient for all our wants.

I understand, then, by the transcendental doc-trine of method, the determination of the for-mal conditions of a complete system of purereason. We shall accordingly have to treat ofthe discipline, the canon, the architectonic, and,finally, the history of pure reason. This part ofour Critique will accomplish, from the trans-cendental point of view, what has been usuallyattempted, but miserably executed, under thename of practical logic. It has been badly execu-ted, I say, because general logic, not being limi-ted to any particular kind of cognition (noteven to the pure cognition of the understan-ding) nor to any particular objects, it cannot,without borrowing from other sciences, do mo-

re than present merely the titles or signs of pos-sible methods and the technical expressions,which are employed in the systematic parts ofall sciences; and thus the pupil is made ac-quainted with names, the meaning and applica-tion of which he is to learn only at some futuretime.

CHAPTER I. The Discipline of Pure Rea-son.

Negative judgements—those which are so notmerely as regards their logical form, but in res-pect of their content—are not commonly heldin especial respect. They are, on the contrary,regarded as jealous enemies of our insatiabledesire for knowledge; and it almost requires anapology to induce us to tolerate, much less toprize and to respect them.

All propositions, indeed, may be logically ex-pressed in a negative form; but, in relation tothe content of our cognition, the peculiar pro-vince of negative judgements is solely to pre-vent error. For this reason, too, negative propo-sitions, which are framed for the purpose ofcorrecting false cognitions where error is abso-lutely impossible, are undoubtedly true, butinane and senseless; that is, they are in realitypurposeless and, for this reason, often veryridiculous. Such is the proposition of theschoolman that Alexander could not have sub-dued any countries without an army.

But where the limits of our possible cognitionare very much contracted, the attraction to newfields of knowledge great, the illusions towhich the mind is subject of the most deceptivecharacter, and the evil consequences of error ofno inconsiderable magnitude—the negativeelement in knowledge, which is useful only toguard us against error, is of far more importan-

ce than much of that positive instruction whichmakes additions to the sum of our knowledge.The restraint which is employed to repress, andfinally to extirpate the constant inclination todepart from certain rules, is termed discipline.It is distinguished from culture, which aims atthe formation of a certain degree of skill, wit-hout attempting to repress or to destroy anyother mental power, already existing. In thecultivation of a talent, which has given eviden-ce of an impulse towards self-development,discipline takes a negative,* culture and doctri-ne a positive, part.

[*Footnote: I am well aware that, in the langua-ge of the schools, the term discipline is usuallyemployed as synonymous with instruction. Butthere are so many cases in which it is necessaryto distinguish the notion of the former, as acourse of corrective training, from that of thelatter, as the communication of knowledge, and

the nature of things itself demands the appro-priation of the most suitable expressions forthis distinction, that it is my desire that theformer terms should never be employed in anyother than a negative signification.]

That natural dispositions and talents (such asimagination and wit), which ask a free and un-limited development, require in many respectsthe corrective influence of discipline, every onewill readily grant. But it may well appear stran-ge that reason, whose proper duty it is to pres-cribe rules of discipline to all the other powersof the mind, should itself require this correcti-ve. It has, in fact, hitherto escaped this humilia-tion, only because, in presence of its magnifi-cent pretensions and high position, no onecould readily suspect it to be capable of substi-tuting fancies for conceptions, and words forthings.

Reason, when employed in the field of expe-rience, does not stand in need of criticism, be-cause its principles are subjected to the conti-nual test of empirical observations. Nor is criti-cism requisite in the sphere of mathematics,where the conceptions of reason must alwaysbe presented in concreto in pure intuition, andbaseless or arbitrary assertions are discoveredwithout difficulty. But where reason is not heldin a plain track by the influence of empirical orof pure intuition, that is, when it is employed inthe transcendental sphere of pure conceptions,it stands in great need of discipline, to restrainits propensity to overstep the limits of possibleexperience and to keep it from wandering intoerror. In fact, the utility of the philosophy ofpure reason is entirely of this negative charac-ter. Particular errors may be corrected by parti-cular animadversions, and the causes of theseerrors may be eradicated by criticism. But whe-re we find, as in the case of pure reason, a com-plete system of illusions and fallacies, closely

connected with each other and depending upongrand general principles, there seems to be re-quired a peculiar and negative code of mentallegislation, which, under the denomination of adiscipline, and founded upon the nature of rea-son and the objects of its exercise, shall consti-tute a system of thorough examination andtesting, which no fallacy will be able to withs-tand or escape from, under whatever disguiseor concealment it may lurk.

But the reader must remark that, in this thesecond division of our transcendental Critiquethe discipline of pure reason is not directed tothe content, but to the method of the cognitionof pure reason. The former task has been com-pleted in the doctrine of elements. But there isso much similarity in the mode of employingthe faculty of reason, whatever be the object towhich it is applied, while, at the same time, itsemployment in the transcendental sphere is soessentially different in kind from every other,

that, without the warning negative influence ofa discipline specially directed to that end, theerrors are unavoidable which spring from theunskillful employment of the methods whichare originated by reason but which are out ofplace in this sphere.

SECTION I. The Discipline of Pure Reasonin the Sphere of Dogmatism.

The science of mathematics presents the mostbrilliant example of the extension of the sphereof pure reason without the aid of experience.Examples are always contagious; and they exertan especial influence on the same faculty,which naturally flatters itself that it will havethe same good fortune in other case as fell to itslot in one fortunate instance. Hence pure reasonhopes to be able to extend its empire in thetranscendental sphere with equal success and

security, especially when it applies the samemethod which was attended with such brilliantresults in the science of mathematics. It is, the-refore, of the highest importance for us to knowwhether the method of arriving at demonstra-tive certainty, which is termed mathematical,be identical with that by which we endeavourto attain the same degree of certainty in philo-sophy, and which is termed in that sciencedogmatical.

Philosophical cognition is the cognition of rea-son by means of conceptions; mathematicalcognition is cognition by means of the construc-tion of conceptions. The construction of a con-ception is the presentation a priori of the intui-tion which corresponds to the conception. Forthis purpose a non-empirical intuition is requi-site, which, as an intuition, is an individualobject; while, as the construction of a concep-tion (a general representation), it must be seento be universally valid for all the possible intui-

tions which rank under that conception. Thus Iconstruct a triangle, by the presentation of theobject which corresponds to this conception,either by mere imagination, in pure intuition,or upon paper, in empirical intuition, in bothcases completely a priori, without borrowingthe type of that figure from any experience. Theindividual figure drawn upon paper is empiri-cal; but it serves, notwithstanding, to indicatethe conception, even in its universality, becausein this empirical intuition we keep our eye me-rely on the act of the construction of the con-ception, and pay no attention to the variousmodes of determining it, for example, its size,the length of its sides, the size of its angles, the-se not in the least affecting the essential charac-ter of the conception.

Philosophical cognition, accordingly, regardsthe particular only in the general; mathematicalthe general in the particular, nay, in the indivi-dual. This is done, however, entirely a priori

and by means of pure reason, so that, as thisindividual figure is determined under certainuniversal conditions of construction, the objectof the conception, to which this individual figu-re corresponds as its schema, must be cogitatedas universally determined.

The essential difference of these two modes ofcognition consists, therefore, in this formal qua-lity; it does not regard the difference of the mat-ter or objects of both. Those thinkers who aimat distinguishing philosophy from mathematicsby asserting that the former has to do with qua-lity merely, and the latter with quantity, havemistaken the effect for the cause. The reasonwhy mathematical cognition can relate only toquantity is to be found in its form alone. For itis the conception of quantities only that is ca-pable of being constructed, that is, presented apriori in intuition; while qualities cannot begiven in any other than an empirical intuition.Hence the cognition of qualities by reason is

possible only through conceptions. No one canfind an intuition which shall correspond to theconception of reality, except in experience; itcannot be presented to the mind a priori andantecedently to the empirical consciousness of areality. We can form an intuition, by means ofthe mere conception of it, of a cone, without theaid of experience; but the colour of the cone wecannot know except from experience. I cannotpresent an intuition of a cause, except in anexample which experience offers to me. Besi-des, philosophy, as well as mathematics, treatsof quantities; as, for example, of totality, infini-ty, and so on. Mathematics, too, treats of thedifference of lines and surfaces—as spaces ofdifferent quality, of the continuity of exten-sion—as a quality thereof. But, although insuch cases they have a common object, the mo-de in which reason considers that object is verydifferent in philosophy from what it is in mat-hematics. The former confines itself to the ge-neral conceptions; the latter can do nothing

with a mere conception, it hastens to intuition.In this intuition it regards the conception inconcreto, not empirically, but in an a priori in-tuition, which it has constructed; and in which,all the results which follow from the generalconditions of the construction of the conceptionare in all cases valid for the object of the cons-tructed conception.

Suppose that the conception of a triangle isgiven to a philosopher and that he is requiredto discover, by the philosophical method, whatrelation the sum of its angles bears to a rightangle. He has nothing before him but the con-ception of a figure enclosed within three rightlines, and, consequently, with the same numberof angles. He may analyse the conception of aright line, of an angle, or of the number three aslong as he pleases, but he will not discover anyproperties not contained in these conceptions.But, if this question is proposed to a geometri-cian, he at once begins by constructing a trian-

gle. He knows that two right angles are equal tothe sum of all the contiguous angles which pro-ceed from one point in a straight line; and hegoes on to produce one side of his triangle, thusforming two adjacent angles which are togetherequal to two right angles. He then divides theexterior of these angles, by drawing a line para-llel with the opposite side of the triangle, andimmediately perceives that he has thus got anexterior adjacent angle which is equal to theinterior. Proceeding in this way, through achain of inferences, and always on the groundof intuition, he arrives at a clear and universallyvalid solution of the question.

But mathematics does not confine itself to theconstruction of quantities (quanta), as in thecase of geometry; it occupies itself with purequantity also (quantitas), as in the case of alge-bra, where complete abstraction is made of theproperties of the object indicated by the con-ception of quantity. In algebra, a certain met-

hod of notation by signs is adopted, and theseindicate the different possible constructions ofquantities, the extraction of roots, and so on.After having thus denoted the general concep-tion of quantities, according to their differentrelations, the different operations by whichquantity or number is increased or diminishedare presented in intuition in accordance withgeneral rules. Thus, when one quantity is to bedivided by another, the signs which denoteboth are placed in the form peculiar to the ope-ration of division; and thus algebra, by meansof a symbolical construction of quantity, just asgeometry, with its ostensive or geometricalconstruction (a construction of the objectsthemselves), arrives at results which discursivecognition cannot hope to reach by the aid ofmere conceptions.

Now, what is the cause of this difference in thefortune of the philosopher and the mathemati-cian, the former of whom follows the path of

conceptions, while the latter pursues that ofintuitions, which he represents, a priori, in co-rrespondence with his conceptions? The causeis evident from what has been already demons-trated in the introduction to this Critique. Wedo not, in the present case, want to discoveranalytical propositions, which may be produ-ced merely by analysing our conceptions—forin this the philosopher would have the advan-tage over his rival; we aim at the discovery ofsynthetical propositions—such synthetical pro-positions, moreover, as can be cognized a prio-ri. I must not confine myself to that which Iactually cogitate in my conception of a triangle,for this is nothing more than the mere defini-tion; I must try to go beyond that, and to arriveat properties which are not contained in, alt-hough they belong to, the conception. Now,this is impossible, unless I determine the objectpresent to my mind according to the condi-tions, either of empirical, or of pure, intuition.In the former case, I should have an empirical

proposition (arrived at by actual measurementof the angles of the triangle), which would pos-sess neither universality nor necessity; but thatwould be of no value. In the latter, I proceed bygeometrical construction, by means of which Icollect, in a pure intuition, just as I would in anempirical intuition, all the various propertieswhich belong to the schema of a triangle in ge-neral, and consequently to its conception, andthus construct synthetical propositions whichpossess the attribute of universality.

It would be vain to philosophize upon thetriangle, that is, to reflect on it discursively; Ishould get no further than the definition withwhich I had been obliged to set out. There arecertainly transcendental synthetical proposi-tions which are framed by means of pure con-ceptions, and which form the peculiar distinc-tion of philosophy; but these do not relate toany particular thing, but to a thing in general,and enounce the conditions under which the

perception of it may become a part of possibleexperience. But the science of mathematics hasnothing to do with such questions, nor with thequestion of existence in any fashion; it is con-cerned merely with the properties of objects inthemselves, only in so far as these are connec-ted with the conception of the objects.

In the above example, we merely attempted toshow the great difference which exists betweenthe discursive employment of reason in thesphere of conceptions, and its intuitive exerciseby means of the construction of conceptions.The question naturally arises: What is the causewhich necessitates this twofold exercise of rea-son, and how are we to discover whether it isthe philosophical or the mathematical methodwhich reason is pursuing in an argument?

All our knowledge relates, finally, to possibleintuitions, for it is these alone that present ob-jects to the mind. An a priori or non-empiricalconception contains either a pure intuition—

and in this case it can be constructed; or it con-tains nothing but the synthesis of possible in-tuitions, which are not given a priori. In thislatter case, it may help us to form synthetical apriori judgements, but only in the discursivemethod, by conceptions, not in the intuitive, bymeans of the construction of conceptions.

The only a priori intuition is that of the pureform of phenomena- space and time. A concep-tion of space and time as quanta may be pre-sented a priori in intuition, that is, constructed,either alone with their quality (figure), or aspure quantity (the mere synthesis of the homo-geneous), by means of number. But the matterof phenomena, by which things are given inspace and time, can be presented only in per-ception, a posteriori. The only conceptionwhich represents a priori this empirical contentof phenomena is the conception of a thing ingeneral; and the a priori synthetical cognitionof this conception can give us nothing more

than the rule for the synthesis of that whichmay be contained in the corresponding a poste-riori perception; it is utterly inadequate to pre-sent an a priori intuition of the real object,which must necessarily be empirical.

Synthetical propositions, which relate to thingsin general, an a priori intuition of which is im-possible, are transcendental. For this reasontranscendental propositions cannot be framedby means of the construction of conceptions;they are a priori, and based entirely on concep-tions themselves. They contain merely the rule,by which we are to seek in the world of percep-tion or experience the synthetical unity of thatwhich cannot be intuited a priori. But they areincompetent to present any of the conceptionswhich appear in them in an a priori intuition;these can be given only a posteriori, in expe-rience, which, however, is itself possible onlythrough these synthetical principles.

If we are to form a synthetical judgement re-garding a conception, we must go beyond it, tothe intuition in which it is given. If we keep towhat is contained in the conception, the judge-ment is merely analytical—it is merely an ex-planation of what we have cogitated in the con-ception. But I can pass from the conception tothe pure or empirical intuition which corres-ponds to it. I can proceed to examine my con-ception in concreto, and to cognize, either apriori or a posterio, what I find in the object ofthe conception. The former—a priori cogni-tion—is rational-mathematical cognition bymeans of the construction of the conception; thelatter—a posteriori cognition—is purely empi-rical cognition, which does not possess the at-tributes of necessity and universality. Thus Imay analyse the conception I have of gold; but Igain no new information from this analysis, Imerely enumerate the different propertieswhich I had connected with the notion indica-ted by the word. My knowledge has gained in

logical clearness and arrangement, but no addi-tion has been made to it. But if I take the matterwhich is indicated by this name, and submit itto the examination of my senses, I am enabledto form several synthetical—although still em-pirical- propositions. The mathematical concep-tion of a triangle I should construct, that is, pre-sent a priori in intuition, and in this way attainto rational-synthetical cognition. But when thetranscendental conception of reality, or subs-tance, or power is presented to my mind, I findthat it does not relate to or indicate either anempirical or pure intuition, but that it indicatesmerely the synthesis of empirical intuitions,which cannot of course be given a priori. Thesynthesis in such a conception cannot proceed apriori—without the aid of experience—to theintuition which corresponds to the conception;and, for this reason, none of these conceptionscan produce a determinative synthetical propo-sition, they can never present more than a prin-ciple of the synthesis* of possible empirical

intuitions. A transcendental proposition is, the-refore, a synthetical cognition of reason bymeans of pure conceptions and the discursivemethod, and it renders possible all syntheticalunity in empirical cognition, though it cannotpresent us with any intuition a priori.

[*Footnote: In the case of the conception of cau-se, I do really go beyond the empirical concep-tion of an event—but not to the intuition whichpresents this conception in concreto, but only tothe time-conditions, which may be found inexperience to correspond to the conception. Myprocedure is, therefore, strictly according toconceptions; I cannot in a case of this kind em-ploy the construction of conceptions, becausethe conception is merely a rule for the synthesisof perceptions, which are not pure intuitions,and which, therefore, cannot be given a priori.]

There is thus a twofold exercise of reason. Bothmodes have the properties of universality andan a priori origin in common, but are, in theirprocedure, of widely different character. Thereason of this is that in the world of phenome-na, in which alone objects are presented to ourminds, there are two main elements—the formof intuition (space and time), which can be cog-nized and determined completely a priori, andthe matter or content—that which is presentedin space and time, and which, consequently,contains a something—an existence correspon-ding to our powers of sensation. As regards thelatter, which can never be given in a determina-te mode except by experience, there are no apriori notions which relate to it, except the un-determined conceptions of the synthesis of pos-sible sensations, in so far as these belong (in apossible experience) to the unity of conscious-ness. As regards the former, we can determineour conceptions a priori in intuition, inasmuchas we are ourselves the creators of the objects of

the conceptions in space and time- these objectsbeing regarded simply as quanta. In the onecase, reason proceeds according to conceptionsand can do nothing more than subject pheno-mena to these—which can only be determinedempirically, that is, a posteriori—in conformity,however, with those conceptions as the rules ofall empirical synthesis. In the other case, reasonproceeds by the construction of conceptions;and, as these conceptions relate to an a prioriintuition, they may be given and determined inpure intuition a priori, and without the aid ofempirical data. The examination and considera-tion of everything that exists in space or time—whether it is a quantum or not, in how far theparticular something (which fills space or time)is a primary substratum, or a mere determina-tion of some other existence, whether it relatesto anything else—either as cause or effect,whether its existence is isolated or in reciprocalconnection with and dependence upon others,the possibility of this existence, its reality and

necessity or opposites—all these form part ofthe cognition of reason on the ground of con-ceptions, and this cognition is termed philo-sophical. But to determine a priori an intuitionin space (its figure), to divide time into periods,or merely to cognize the quantity of an intui-tion in space and time, and to determine it bynumber—all this is an operation of reason bymeans of the construction of conceptions, andis called mathematical.

The success which attends the efforts of reasonin the sphere of mathematics naturally fostersthe expectation that the same good fortune willbe its lot, if it applies the mathematical methodin other regions of mental endeavour besidesthat of quantities. Its success is thus great, be-cause it can support all its conceptions by apriori intuitions and, in this way, make itself amaster, as it were, over nature; while pure phi-losophy, with its a priori discursive concep-tions, bungles about in the world of nature, and

cannot accredit or show any a priori evidenceof the reality of these conceptions. Masters inthe science of mathematics are confident of thesuccess of this method; indeed, it is a commonpersuasion that it is capable of being applied toany subject of human thought. They have hard-ly ever reflected or philosophized on their fa-vourite science—a task of great difficulty; andthe specific difference between the two modesof employing the faculty of reason has neverentered their thoughts. Rules current in thefield of common experience, and which com-mon sense stamps everywhere with its appro-val, are regarded by them as axiomatic. Fromwhat source the conceptions of space and time,with which (as the only primitive quanta) theyhave to deal, enter their minds, is a questionwhich they do not trouble themselves to ans-wer; and they think it just as unnecessary toexamine into the origin of the pure conceptionsof the understanding and the extent of theirvalidity. All they have to do with them is to

employ them. In all this they are perfectly right,if they do not overstep the limits of the sphereof nature. But they pass, unconsciously, fromthe world of sense to the insecure ground ofpure transcendental conceptions (instabilis te-llus, innabilis unda), where they can neitherstand nor swim, and where the tracks of theirfootsteps are obliterated by time; while themarch of mathematics is pursued on a broadand magnificent highway, which the latest pos-terity shall frequent without fear of danger orimpediment.

As we have taken upon us the task of determi-ning, clearly and certainly, the limits of purereason in the sphere of transcendentalism, andas the efforts of reason in this direction are per-sisted in, even after the plainest and most ex-pressive warnings, hope still beckoning us pastthe limits of experience into the splendours ofthe intellectual world—it becomes necessary tocut away the last anchor of this fallacious and

fantastic hope. We shall, accordingly, show thatthe mathematical method is unattended in thesphere of philosophy by the least advantage—except, perhaps, that it more plainly exhibits itsown inadequacy—that geometry and philo-sophy are two quite different things, althoughthey go band in hand in hand in the field ofnatural science, and, consequently, that theprocedure of the one can never be imitated bythe other.

The evidence of mathematics rests upon defini-tions, axioms, and demonstrations. I shall besatisfied with showing that none of these formscan be employed or imitated in philosophy inthe sense in which they are understood by mat-hematicians; and that the geometrician, if heemploys his method in philosophy, will suc-ceed only in building card-castles, while theemployment of the philosophical method inmathematics can result in nothing but mereverbiage. The essential business of philosophy,

indeed, is to mark out the limits of the science;and even the mathematician, unless his talent isnaturally circumscribed and limited to this par-ticular department of knowledge, cannot turn adeaf ear to the warnings of philosophy, or sethimself above its direction.

I. Of Definitions. A definition is, as the termitself indicates, the representation, upon prima-ry grounds, of the complete conception of athing within its own limits.* Accordingly, anempirical conception cannot be defined, it canonly be explained. For, as there are in such aconception only a certain number of marks orsigns, which denote a certain class of sensuousobjects, we can never be sure that we do notcogitate under the word which indicates thesame object, at one time a greater, at another asmaller number of signs. Thus, one person maycogitate in his conception of gold, in addition toits properties of weight, colour, malleability,that of resisting rust, while another person may

be ignorant of this quality. We employ certainsigns only so long as we require them for thesake of distinction; new observations abstractsome and add new ones, so that an empiricalconception never remains within permanentlimits. It is, in fact, useless to define a concep-tion of this kind. If, for example, we are spea-king of water and its properties, we do not stopat what we actually think by the word water,but proceed to observation and experiment;and the word, with the few signs attached to it,is more properly a designation than a concep-tion of the thing. A definition in this case wouldevidently be nothing more than a determina-tion of the word. In the second place, no a prio-ri conception, such as those of substance, cause,right, fitness, and so on, can be defined. For Ican never be sure, that the clear representationof a given conception (which is given in a con-fused state) has been fully developed, until Iknow that the representation is adequate withits object. But, inasmuch as the conception, as it

is presented to the mind, may contain a numberof obscure representations, which we do notobserve in our analysis, although we employthem in our application of the conception, I cannever be sure that my analysis is complete,while examples may make this probable, alt-hough they can never demonstrate the fact.Instead of the word definition, I should ratheremploy the term exposition— a more modestexpression, which the critic may accept withoutsurrendering his doubts as to the completenessof the analysis of any such conception. As, the-refore, neither empirical nor a priori concep-tions are capable of definition, we have to seewhether the only other kind of conceptions—arbitrary conceptions—can be subjected to thismental operation. Such a conception can al-ways be defined; for I must know thoroughlywhat I wished to cogitate in it, as it was I whocreated it, and it was not given to my mindeither by the nature of my understanding or byexperience. At the same time, I cannot say that,

by such a definition, I have defined a real ob-ject. If the conception is based upon empiricalconditions, if, for example, I have a conceptionof a clock for a ship, this arbitrary conceptiondoes not assure me of the existence or even ofthe possibility of the object. My definition ofsuch a conception would with more proprietybe termed a declaration of a project than a defi-nition of an object. There are no other concep-tions which can bear definition, except thosewhich contain an arbitrary synthesis, which canbe constructed a priori. Consequently, thescience of mathematics alone possesses defini-tions. For the object here thought is presented apriori in intuition; and thus it can never containmore or less than the conception, because theconception of the object has been given by thedefinition—and primarily, that is, without de-riving the definition from any other source.Philosophical definitions are, therefore, merelyexpositions of given conceptions, while mat-hematical definitions are constructions of con-

ceptions originally formed by the mind itself;the former are produced by analysis, the com-pleteness of which is never demonstrativelycertain, the latter by a synthesis. In a mathema-tical definition the conception is formed, in aphilosophical definition it is only explained.From this it follows:

[*Footnote: The definition must describe theconception completely that is, omit none of themarks or signs of which it composed; within itsown limits, that is, it must be precise, and enu-merate no more signs than belong to the con-ception; and on primary grounds, that is to say,the limitations of the bounds of the conceptionmust not be deduced from other conceptions,as in this case a proof would be necessary, andthe so-called definition would be incapable oftaking its place at the bead of all the judge-ments we have to form regarding an object.]

(a) That we must not imitate, in philosophy, themathematical usage of commencing with defi-

nitions—except by way of hypothesis or expe-riment. For, as all so-called philosophical defi-nitions are merely analyses of given concep-tions, these conceptions, although only in aconfused form, must precede the analysis; andthe incomplete exposition must precede thecomplete, so that we may be able to draw cer-tain inferences from the characteristics whichan incomplete analysis has enabled us to disco-ver, before we attain to the complete expositionor definition of the conception. In one word, afull and clear definition ought, in philosophy,rather to form the conclusion than the com-mencement of our labours.* In mathematics, onthe contrary, we cannot have a conception priorto the definition; it is the definition which givesus the conception, and it must for this reasonform the commencement of every chain ofmathematical reasoning.

[*Footnote: Philosophy abounds in faulty defi-nitions, especially such as contain some of the

elements requisite to form a complete defini-tion. If a conception could not be employed inreasoning before it had been defined, it wouldfare ill with all philosophical thought. But, asincompletely defined conceptions may alwaysbe employed without detriment to truth, so faras our analysis of the elements contained inthem proceeds, imperfect definitions, that is,propositions which are properly not defini-tions, but merely approximations thereto, maybe used with great advantage. In mathematics,definition belongs ad esse, in philosophy admelius esse. It is a difficult task to construct aproper definition. Jurists are still without acomplete definition of the idea of right.]

(b) Mathematical definitions cannot be erro-neous. For the conception is given only in andthrough the definition, and thus it contains on-ly what has been cogitated in the definition. Butalthough a definition cannot be incorrect, asregards its content, an error may sometimes,

although seldom, creep into the form. Thiserror consists in a want of precision. Thus thecommon definition of a circle—that it is a cur-ved line, every point in which is equally distantfrom another point called the centre—is faulty,from the fact that the determination indicatedby the word curved is superfluous. For thereought to be a particular theorem, which may beeasily proved from the definition, to the effectthat every line, which has all its points at equaldistances from another point, must be a curvedline—that is, that not even the smallest part ofit can be straight. Analytical definitions, on theother hand, may be erroneous in many res-pects, either by the introduction of signs whichdo not actually exist in the conception, or bywanting in that completeness which forms theessential of a definition. In the latter case, thedefinition is necessarily defective, because wecan never be fully certain of the completenessof our analysis. For these reasons, the method

of definition employed in mathematics cannotbe imitated in philosophy.

2. Of Axioms. These, in so far as they are im-mediately certain, are a priori synthetical prin-ciples. Now, one conception cannot be connec-ted synthetically and yet immediately withanother; because, if we wish to proceed out ofand beyond a conception, a third mediatingcognition is necessary. And, as philosophy is acognition of reason by the aid of conceptionsalone, there is to be found in it no principlewhich deserves to be called an axiom. Mat-hematics, on the other hand, may possessaxioms, because it can always connect the pre-dicates of an object a priori, and without anymediating term, by means of the constructionof conceptions in intuition. Such is the casewith the proposition: Three points can alwayslie in a plane. On the other hand, no syntheticalprinciple which is based upon conceptions, canever be immediately certain (for example, the

proposition: Everything that happens has acause), because I require a mediating term toconnect the two conceptions of event and cau-se- namely, the condition of time-determinationin an experience, and I cannot cognize any suchprinciple immediately and from conceptionsalone. Discursive principles are, accordingly,very different from intuitive principles oraxioms. The former always require deduction,which in the case of the latter may be altogetherdispensed with. Axioms are, for this reason,always self-evident, while philosophical prin-ciples, whatever may be the degree of certaintythey possess, cannot lay any claim to such adistinction. No synthetical proposition of puretranscendental reason can be so evident, as isoften rashly enough declared, as the statement,twice two are four. It is true that in the AnalyticI introduced into the list of principles of thepure understanding, certain axioms of intui-tion; but the principle there discussed was notitself an axiom, but served merely to present

the principle of the possibility of axioms in ge-neral, while it was really nothing more than aprinciple based upon conceptions. For it is onepart of the duty of transcendental philosophyto establish the possibility of mathematics itself.Philosophy possesses, then, no axioms, and hasno right to impose its a priori principles uponthought, until it has established their authorityand validity by a thoroughgoing deduction.

3. Of Demonstrations. Only an apodeicticproof, based upon intuition, can be termed ademonstration. Experience teaches us what is,but it cannot convince us that it might not havebeen otherwise. Hence a proof upon empiricalgrounds cannot be apodeictic. A priori concep-tions, in discursive cognition, can never produ-ce intuitive certainty or evidence, however cer-tain the judgement they present may be. Mat-hematics alone, therefore, contains demonstra-tions, because it does not deduce its cognitionfrom conceptions, but from the construction of

conceptions, that is, from intuition, which canbe given a priori in accordance with concep-tions. The method of algebra, in equations,from which the correct answer is deduced byreduction, is a kind of construction—not geo-metrical, but by symbols- in which all concep-tions, especially those of the relations of quanti-ties, are represented in intuition by signs; andthus the conclusions in that science are securedfrom errors by the fact that every proof is sub-mitted to ocular evidence. Philosophical cogni-tion does not possess this advantage, it beingrequired to consider the general always in abs-tracto (by means of conceptions), while mat-hematics can always consider it in concreto (inan individual intuition), and at the same timeby means of a priori representation, wherebyall errors are rendered manifest to the senses.The former—discursive proofs—ought to betermed acroamatic proofs, rather than demons-trations, as only words are employed in them,while demonstrations proper, as the term itself

indicates, always require a reference to the in-tuition of the object.

It follows from all these considerations that it isnot consonant with the nature of philosophy,especially in the sphere of pure reason, to em-ploy the dogmatical method, and to adorn itselfwith the titles and insignia of mathematicalscience. It does not belong to that order, andcan only hope for a fraternal union with thatscience. Its attempts at mathematical evidenceare vain pretensions, which can only keep itback from its true aim, which is to detect theillusory procedure of reason when transgres-sing its proper limits, and by fully explainingand analysing our conceptions, to conduct usfrom the dim regions of speculation to the clearregion of modest self-knowledge. Reason mustnot, therefore, in its transcendental endeavours,look forward with such confidence, as if thepath it is pursuing led straight to its aim, norreckon with such security upon its premisses,

as to consider it unnecessary to take a stepback, or to keep a strict watch for errors, which,overlooked in the principles, may be detectedin the arguments themselves—in which case itmay be requisite either to determine these prin-ciples with greater strictness, or to change thementirely.

I divide all apodeictic propositions, whetherdemonstrable or immediately certain, into dog-mata and mathemata. A direct synthetical pro-position, based on conceptions, is a dogma; aproposition of the same kind, based on theconstruction of conceptions, is a mathema.Analytical judgements do not teach us any mo-re about an object than what was contained inthe conception we had of it; because they donot extend our cognition beyond our concep-tion of an object, they merely elucidate the con-ception. They cannot therefore be with proprie-ty termed dogmas. Of the two kinds of a priorisynthetical propositions above mentioned, only

those which are employed in philosophy can,according to the general mode of speech, bearthis name; those of arithmetic or geometrywould not be rightly so denominated. Thus thecustomary mode of speaking confirms the ex-planation given above, and the conclusion arri-ved at, that only those judgements which arebased upon conceptions, not on the construc-tion of conceptions, can be termed dogmatical.

Thus, pure reason, in the sphere of speculation,does not contain a single direct synthetical jud-gement based upon conceptions. By means ofideas, it is, as we have shown, incapable of pro-ducing synthetical judgements, which are ob-jectively valid; by means of the conceptions ofthe understanding, it establishes certain indubi-table principles, not, however, directly on thebasis of conceptions, but only indirectly bymeans of the relation of these conceptions tosomething of a purely contingent nature, name-ly, possible experience. When experience is

presupposed, these principles are apodeictica-lly certain, but in themselves, and directly, theycannot even be cognized a priori. Thus the gi-ven conceptions of cause and event will not besufficient for the demonstration of the proposi-tion: Every event has a cause. For this reason, itis not a dogma; although from another point ofview, that of experience, it is capable of beingproved to demonstration. The proper term forsuch a proposition is principle, and not theo-rem (although it does require to be proved),because it possesses the remarkable peculiarityof being the condition of the possibility of itsown ground of proof, that is, experience, and offorming a necessary presupposition in all empi-rical observation.

If then, in the speculative sphere of pure rea-son, no dogmata are to be found; all dogmaticalmethods, whether borrowed from mathematics,or invented by philosophical thinkers, are alikeinappropriate and inefficient. They only serve

to conceal errors and fallacies, and to deceivephilosophy, whose duty it is to see that reasonpursues a safe and straight path. A philosop-hical method may, however, be systematical.For our reason is, subjectively considered, itselfa system, and, in the sphere of mere concep-tions, a system of investigation according toprinciples of unity, the material being suppliedby experience alone. But this is not the properplace for discussing the peculiar method oftranscendental philosophy, as our present taskis simply to examine whether our faculties arecapable of erecting an edifice on the basis ofpure reason, and how far they may proceedwith the materials at their command.

SECTION II. The Discipline of Pure Rea-son in Polemics.

Reason must be subject, in all its operations, tocriticism, which must always be permitted toexercise its functions without restraint; other-wise its interests are imperilled and its influen-ce obnoxious to suspicion. There is nothing,however useful, however sacred it may be, thatcan claim exemption from the searching exami-nation of this supreme tribunal, which has norespect of persons. The very existence of reasondepends upon this freedom; for the voice ofreason is not that of a dictatorial and despoticpower, it is rather like the vote of the citizens ofa free state, every member of which must havethe privilege of giving free expression to hisdoubts, and possess even the right of veto.

But while reason can never decline to submititself to the tribunal of criticism, it has not al-ways cause to dread the judgement of this

court. Pure reason, however, when engaged inthe sphere of dogmatism, is not so thoroughlyconscious of a strict observance of its highestlaws, as to appear before a higher judicial rea-son with perfect confidence. On the contrary, itmust renounce its magnificent dogmatical pre-tensions in philosophy.

Very different is the case when it has to defenditself, not before a judge, but against an equal.If dogmatical assertions are advanced on thenegative side, in opposition to those made byreason on the positive side, its justification katauthrhopon is complete, although the proof ofits propositions is kat aletheian unsatisfactory.

By the polemic of pure reason I mean the de-fence of its propositions made by reason, inopposition to the dogmatical counter-propositions advanced by other parties. Thequestion here is not whether its own statementsmay not also be false; it merely regards the factthat reason proves that the opposite cannot be

established with demonstrative certainty, noreven asserted with a higher degree of probabili-ty. Reason does not hold her possessions uponsufferance; for, although she cannot show aperfectly satisfactory title to them, no one canprove that she is not the rightful possessor.

It is a melancholy reflection that reason, in itshighest exercise, falls into an antithetic; andthat the supreme tribunal for the settlement ofdifferences should not be at union with itself. Itis true that we had to discuss the question of anapparent antithetic, but we found that it wasbased upon a misconception. In conformitywith the common prejudice, phenomena wereregarded as things in themselves, and thus anabsolute completeness in their synthesis wasrequired in the one mode or in the other (it wasshown to be impossible in both); a demandentirely out of place in regard to phenomena.There was, then, no real self-contradiction ofreason in the propositions: The series of phe-

nomena given in themselves has an absolutelyfirst beginning; and: This series is absolutelyand in itself without beginning. The two propo-sitions are perfectly consistent with each other,because phenomena as phenomena are inthemselves nothing, and consequently the hy-pothesis that they are things in themselvesmust lead to self-contradictory inferences.

But there are cases in which a similar misun-derstanding cannot be provided against, andthe dispute must remain unsettled. Take, forexample, the theistic proposition: There is aSupreme Being; and on the other hand, the at-heistic counter-statement: There exists no Su-preme Being; or, in psychology: Everythingthat thinks possesses the attribute of absoluteand permanent unity, which is utterly differentfrom the transitory unity of material phenome-na; and the counter-proposition: The soul is notan immaterial unity, and its nature is transito-ry, like that of phenomena. The objects of these

questions contain no heterogeneous or contra-dictory elements, for they relate to things inthemselves, and not to phenomena. Therewould arise, indeed, a real contradiction, if rea-son came forward with a statement on the ne-gative side of these questions alone. As regardsthe criticism to which the grounds of proof onthe affirmative side must be subjected, it maybe freely admitted, without necessitating thesurrender of the affirmative propositions,which have, at least, the interest of reason intheir favour—an advantage which the oppositeparty cannot lay claim to.

I cannot agree with the opinion of several ad-mirable thinkers—Sulzer among the rest—that,in spite of the weakness of the arguments hit-herto in use, we may hope, one day, to see suf-ficient demonstrations of the two cardinal pro-positions of pure reason—the existence of aSupreme Being, and the immortality of thesoul. I am certain, on the contrary, that this will

never be the case. For on what ground can rea-son base such synthetical propositions, whichdo not relate to the objects of experience andtheir internal possibility? But it is also demons-tratively certain that no one will ever be able tomaintain the contrary with the least show ofprobability. For, as he can attempt such a proofsolely upon the basis of pure reason, he isbound to prove that a Supreme Being, and athinking subject in the character of a pure inte-lligence, are impossible. But where will he findthe knowledge which can enable him to enoun-ce synthetical judgements in regard to thingswhich transcend the region of experience? Wemay, therefore, rest assured that the oppositenever will be demonstrated. We need not, then,have recourse to scholastic arguments; we mayalways admit the truth of those propositionswhich are consistent with the speculative inte-rests of reason in the sphere of experience, andform, moreover, the only means of uniting thespeculative with the practical interest. Our op-

ponent, who must not be considered here as acritic solely, we can be ready to meet with anon liquet which cannot fail to disconcert him;while we cannot deny his right to a similar re-tort, as we have on our side the advantage ofthe support of the subjective maxim of reason,and can therefore look upon all his sophisticalarguments with calm indifference.

From this point of view, there is properly noantithetic of pure reason. For the only arena forsuch a struggle would be upon the field of puretheology and psychology; but on this groundthere can appear no combatant whom we needto fear. Ridicule and boasting can be his onlyweapons; and these may be laughed at, as merechild's play. This consideration restores to Rea-son her courage; for what source of confidencecould be found, if she, whose vocation it is todestroy error, were at variance with herself andwithout any reasonable hope of ever reaching astate of permanent repose?

Everything in nature is good for some purpose.Even poisons are serviceable; they destroy theevil effects of other poisons generated in oursystem, and must always find a place in everycomplete pharmacopoeia. The objections raisedagainst the fallacies and sophistries of specula-tive reason, are objections given by the natureof this reason itself, and must therefore have adestination and purpose which can only be forthe good of humanity. For what purpose hasProvidence raised many objects, in which wehave the deepest interest, so far above us, thatwe vainly try to cognize them with certainty,and our powers of mental vision are rather ex-cited than satisfied by the glimpses we maychance to seize? It is very doubtful whether it isfor our benefit to advance bold affirmationsregarding subjects involved in such obscurity;perhaps it would even be detrimental to ourbest interests. But it is undoubtedly always be-neficial to leave the investigating, as well as thecritical reason, in perfect freedom, and permit it

to take charge of its own interests, which areadvanced as much by its limitation, as by itsextension of its views, and which always sufferby the interference of foreign powers forcing it,against its natural tendencies, to bend to certainpreconceived designs.

Allow your opponent to say what he thinksreasonable, and combat him only with theweapons of reason. Have no anxiety for thepractical interests of humanity—these are neverimperilled in a purely speculative dispute. Sucha dispute serves merely to disclose the antino-my of reason, which, as it has its source in thenature of reason, ought to be thoroughly inves-tigated. Reason is benefited by the examinationof a subject on both sides, and its judgementsare corrected by being limited. It is not the mat-ter that may give occasion to dispute, but themanner. For it is perfectly permissible to em-ploy, in the presence of reason, the language ofa firmly rooted faith, even after we have been

obliged to renounce all pretensions to know-ledge.

If we were to ask the dispassionate DavidHume—a philosopher endowed, in a degreethat few are, with a well-balanced judgement:What motive induced you to spend so muchlabour and thought in undermining the conso-ling and beneficial persuasion that reason iscapable of assuring us of the existence, andpresenting us with a determinate conception ofa Supreme Being?—his answer would be: Not-hing but the desire of teaching reason to knowits own powers better, and, at the same time, adislike of the procedure by which that facultywas compelled to support foregone conclu-sions, and prevented from confessing the inter-nal weaknesses which it cannot but feel when itenters upon a rigid self-examination. If, on theother hand, we were to ask Priestley—a philo-sopher who had no taste for transcendentalspeculation, but was entirely devoted to the

principles of empiricism—what his motiveswere for overturning those two main pillars ofreligion—the doctrines of the freedom of thewill and the immortality of the soul (in his viewthe hope of a future life is but the expectation ofthe miracle of resurrection)- this philosopher,himself a zealous and pious teacher of religion,could give no other answer than this: I acted inthe interest of reason, which always suffers,when certain objects are explained and judgedby a reference to other supposed laws than tho-se of material nature—the only laws which weknow in a determinate manner. It would beunfair to decry the latter philosopher, who en-deavoured to harmonize his paradoxical opi-nions with the interests of religion, and to un-dervalue an honest and reflecting man, becausehe finds himself at a loss the moment he has leftthe field of natural science. The same gracemust be accorded to Hume, a man not lesswell-disposed, and quite as blameless in hismoral character, and who pushed his abstract

speculations to an extreme length, because, ashe rightly believed, the object of them lies enti-rely beyond the bounds of natural science, andwithin the sphere of pure ideas.

What is to be done to provide against the dan-ger which seems in the present case to menacethe best interests of humanity? The course to bepursued in reference to this subject is a perfec-tly plain and natural one. Let each thinker pur-sue his own path; if he shows talent, if he givesevidence of profound thought, in one word, ifhe shows that he possesses the power of reaso-ning—reason is always the gainer. If you haverecourse to other means, if you attempt to coer-ce reason, if you raise the cry of treason tohumanity, if you excite the feelings of thecrowd, which can neither understand nor sym-pathize with such subtle speculations—youwill only make yourselves ridiculous. For thequestion does not concern the advantage ordisadvantage which we are expected to reap

from such inquiries; the question is merely howfar reason can advance in the field of specula-tion, apart from all kinds of interest, and whet-her we may depend upon the exertions of spe-culative reason, or must renounce all relianceon it. Instead of joining the combatants, it isyour part to be a tranquil spectator of thestruggle—a laborious struggle for the partiesengaged, but attended, in its progress as well asin its result, with the most advantageous con-sequences for the interests of thought andknowledge. It is absurd to expect to be enligh-tened by Reason, and at the same time to pres-cribe to her what side of the question she mustadopt. Moreover, reason is sufficiently held incheck by its own power, the limits imposed onit by its own nature are sufficient; it is unneces-sary for you to place over it additional guards,as if its power were dangerous to the constitu-tion of the intellectual state. In the dialectic ofreason there is no victory gained which need inthe least disturb your tranquility.

The strife of dialectic is a necessity of reason,and we cannot but wish that it had been con-ducted long ere this with that perfect freedomwhich ought to be its essential condition. In thiscase, we should have had at an earlier period amatured and profound criticism, which musthave put an end to all dialectical disputes, byexposing the illusions and prejudices in whichthey originated.

There is in human nature an unworthy propen-sity—a propensity which, like everything thatsprings from nature, must in its final purposebe conducive to the good of humanity—to con-ceal our real sentiments, and to give expressiononly to certain received opinions, which areregarded as at once safe and promotive of thecommon good. It is true, this tendency, not onlyto conceal our real sentiments, but to professthose which may gain us favour in the eyes ofsociety, has not only civilized, but, in a certainmeasure, moralized us; as no one can break

through the outward covering of respectability,honour, and morality, and thus the seemingly-good examples which we which we see aroundus form an excellent school for moral impro-vement, so long as our belief in their genuine-ness remains unshaken. But this disposition torepresent ourselves as better than we are, andto utter opinions which are not our own, can benothing more than a kind of provisionaryarrangement of nature to lead us from the ru-deness of an uncivilized state, and to teach ushow to assume at least the appearance andmanner of the good we see. But when trueprinciples have been developed, and have ob-tained a sure foundation in our habit ofthought, this conventionalism must be attackedwith earnest vigour, otherwise it corrupts theheart, and checks the growth of good disposi-tions with the mischievous weed of air appea-rances.

I am sorry to remark the same tendency to mis-representation and hypocrisy in the sphere ofspeculative discussion, where there is lesstemptation to restrain the free expression ofthought. For what can be more prejudicial tothe interests of intelligence than to falsify ourreal sentiments, to conceal the doubts which wefeel in regard to our statements, or to maintainthe validity of grounds of proof which we wellknow to be insufficient? So long as mere perso-nal vanity is the source of these unworthy arti-fices—and this is generally the case in specula-tive discussions, which are mostly destitute ofpractical interest, and are incapable of completedemonstration—the vanity of the opposite par-ty exaggerates as much on the other side; andthus the result is the same, although it is notbrought about so soon as if the dispute hadbeen conducted in a sincere and upright spirit.But where the mass entertains the notion thatthe aim of certain subtle speculators is nothingless than to shake the very foundations of pu-

blic welfare and morality—it seems not onlyprudent, but even praise worthy, to maintainthe good cause by illusory arguments, ratherthan to give to our supposed opponents theadvantage of lowering our declarations to themoderate tone of a merely practical conviction,and of compelling us to confess our inability toattain to apodeictic certainty in speculative sub-jects. But we ought to reflect that there is not-hing, in the world more fatal to the maintenan-ce of a good cause than deceit, misrepresenta-tion, and falsehood. That the strictest laws ofhonesty should be observed in the discussion ofa purely speculative subject is the least requi-rement that can be made. If we could reckonwith security even upon so little, the conflict ofspeculative reason regarding the importantquestions of God, immortality, and freedom,would have been either decided long ago, orwould very soon be brought to a conclusion.But, in general, the uprightness of the defencestands in an inverse ratio to the goodness of the

cause; and perhaps more honesty and fairnessare shown by those who deny than by thosewho uphold these doctrines.

I shall persuade myself, then, that I have rea-ders who do not wish to see a righteous causedefended by unfair arguments. Such will nowrecognize the fact that, according to the princi-ples of this Critique, if we consider not what is,but what ought to be the case, there can be rea-lly no polemic of pure reason. For how can twopersons dispute about a thing, the reality ofwhich neither can present in actual or even inpossible experience? Each adopts the plan ofmeditating on his idea for the purpose of dra-wing from the idea, if he can, what is more thanthe idea, that is, the reality of the object which itindicates. How shall they settle the dispute,since neither is able to make his assertions di-rectly comprehensible and certain, but mustrestrict himself to attacking and confuting thoseof his opponent? All statements enounced by

pure reason transcend the conditions of possi-ble experience, beyond the sphere of which wecan discover no criterion of truth, while theyare at the same time framed in accordance withthe laws of the understanding, which are appli-cable only to experience; and thus it is the fateof all such speculative discussions that whilethe one party attacks the weaker side of hisopponent, he infallibly lays open his ownweaknesses.

The critique of pure reason may be regarded asthe highest tribunal for all speculative disputes;for it is not involved in these disputes, whichhave an immediate relation to certain objectsand not to the laws of the mind, but is institu-ted for the purpose of determining the rightsand limits of reason.

Without the control of criticism, reason is, as itwere, in a state of nature, and can only esta-blish its claims and assertions by war. Criti-cism, on the contrary, deciding all questions

according to the fundamental laws of its owninstitution, secures to us the peace of law andorder, and enables us to discuss all differencesin the more tranquil manner of a legal process.In the former case, disputes are ended by victo-ry, which both sides may claim and which isfollowed by a hollow armistice; in the latter, bya sentence, which, as it strikes at the root of allspeculative differences, ensures to all concer-ned a lasting peace. The endless disputes of adogmatizing reason compel us to look for somemode of arriving at a settled decision by a criti-cal investigation of reason itself; just as Hobbesmaintains that the state of nature is a state ofinjustice and violence, and that we must leaveit and submit ourselves to the constraint of law,which indeed limits individual freedom, butonly that it may consist with the freedom ofothers and with the common good of all.

This freedom will, among other things, permitof our openly stating the difficulties and doubts

which we are ourselves unable to solve, wit-hout being decried on that account as turbulentand dangerous citizens. This privilege formspart of the native rights of human reason,which recognizes no other judge than the uni-versal reason of humanity; and as this reason isthe source of all progress and improvement,such a privilege is to be held sacred and invio-lable. It is unwise, moreover, to denounce asdangerous any bold assertions against, or rashattacks upon, an opinion which is held by thelargest and most moral class of the community;for that would be giving them an importancewhich they do not deserve. When I hear thatthe freedom of the will, the hope of a futurelife, and the existence of God have been overth-rown by the arguments of some able writer, Ifeel a strong desire to read his book; for I expectthat he will add to my knowledge and impartgreater clearness and distinctness to my viewsby the argumentative power shown in his wri-tings. But I am perfectly certain, even before I

have opened the book, that he has not succee-ded in a single point, not because I believe I amin possession of irrefutable demonstrations ofthese important propositions, but because thistranscendental critique, which has disclosed tome the power and the limits of pure reason, hasfully convinced me that, as it is insufficient toestablish the affirmative, it is as powerless, andeven more so, to assure us of the truth of thenegative answer to these questions. From whatsource does this free-thinker derive his know-ledge that there is, for example, no SupremeBeing? This proposition lies out of the field ofpossible experience, and, therefore, beyond thelimits of human cognition. But I would not readat, all the answer which the dogmatical main-tainer of the good cause makes to his opponent,because I know well beforehand, that he willmerely attack the fallacious grounds of his ad-versary, without being able to establish his ownassertions. Besides, a new illusory argument, inthe construction of which talent and acuteness

are shown, is suggestive of new ideas and newtrains of reasoning, and in this respect the oldand everyday sophistries are quite useless.Again, the dogmatical opponent of religiongives employment to criticism, and enables usto test and correct its principles, while there isno occasion for anxiety in regard to the influen-ce and results of his reasoning.

But, it will be said, must we not warn the youthentrusted to academical care against such wri-tings, must we not preserve them from theknowledge of these dangerous assertions, untiltheir judgement is ripened, or rather until thedoctrines which we wish to inculcate are sofirmly rooted in their minds as to withstand allattempts at instilling the contrary dogmas, fromwhatever quarter they may come?

If we are to confine ourselves to the dogmaticalprocedure in the sphere of pure reason, andfind ourselves unable to settle such disputesotherwise than by becoming a party in them,

and setting counter-assertions against the sta-tements advanced by our opponents, there iscertainly no plan more advisable for the mo-ment, but, at the same time, none more absurdand inefficient for the future, than this retainingof the youthful mind under guardianship for atime, and thus preserving it—for so long atleast—from seduction into error. But when, at alater period, either curiosity, or the prevalentfashion of thought places such writings in theirhands, will the so-called convictions of theiryouth stand firm? The young thinker, who hasin his armoury none but dogmatical weaponswith which to resist the attacks of his opponent,and who cannot detect the latent dialecticwhich lies in his own opinions as well as inthose of the opposite party, sees the advance ofillusory arguments and grounds of proof whichhave the advantage of novelty, against as illu-sory grounds of proof destitute of this advan-tage, and which, perhaps, excite the suspicionthat the natural credulity of his youth has been

abused by his instructors. He thinks he can findno better means of showing that he has outgrown the discipline of his minority than bydespising those well-meant warnings, and,knowing no system of thought but that ofdogmatism, he drinks deep draughts of thepoison that is to sap the principles in which hisearly years were trained.

Exactly the opposite of the system here recom-mended ought to be pursued in academicalinstruction. This can only be effected, however,by a thorough training in the critical investiga-tion of pure reason. For, in order to bring theprinciples of this critique into exercise as soonas possible, and to demonstrate their perfecteven in the presence of the highest degree ofdialectical illusion, the student ought to exami-ne the assertions made on both sides of specu-lative questions step by step, and to test themby these principles. It cannot be a difficult taskfor him to show the fallacies inherent in these

propositions, and thus he begins early to feelhis own power of securing himself against theinfluence of such sophistical arguments, whichmust finally lose, for him, all their illusory po-wer. And, although the same blows whichoverturn the edifice of his opponent are as fatalto his own speculative structures, if such he haswished to rear; he need not feel any sorrow inregard to this seeming misfortune, as he hasnow before him a fair prospect into the practi-cal region in which he may reasonably hope tofind a more secure foundation for a rationalsystem.

There is, accordingly, no proper polemic in thesphere of pure reason. Both parties beat the airand fight with their own shadows, as they passbeyond the limits of nature, and can find notangible point of attack—no firm footing fortheir dogmatical conflict. Fight as vigorously asthey may, the shadows which they hew down,immediately start up again, like the heroes in

Walhalla, and renew the bloodless and uncea-sing contest.

But neither can we admit that there is any pro-per sceptical employment of pure reason, suchas might be based upon the principle of neutra-lity in all speculative disputes. To excite reasonagainst itself, to place weapons in the hands ofthe party on the one side as well as in those ofthe other, and to remain an undisturbed andsarcastic spectator of the fierce struggle thatensues, seems, from the dogmatical point ofview, to be a part fitting only a malevolent dis-position. But, when the sophist evidences aninvincible obstinacy and blindness, and a pridewhich no criticism can moderate, there is noother practicable course than to oppose to thispride and obstinacy similar feelings and pre-tensions on the other side, equally well or illfounded, so that reason, staggered by the re-flections thus forced upon it, finds it necessaryto moderate its confidence in such pretensions

and to listen to the advice of criticism. But wecannot stop at these doubts, much less regardthe conviction of our ignorance, not only as acure for the conceit natural to dogmatism, butas the settlement of the disputes in which rea-son is involved with itself. On the contrary,scepticism is merely a means of awakeningreason from its dogmatic dreams and exciting itto a more careful investigation into its ownpowers and pretensions. But, as scepticism ap-pears to be the shortest road to a permanentpeace in the domain of philosophy, and as it isthe track pursued by the many who aim at gi-ving a philosophical colouring to their con-temptuous dislike of all inquiries of this kind, Ithink it necessary to present to my readers thismode of thought in its true light.

Scepticism not a Permanent State for HumanReason.

The consciousness of ignorance—unless thisignorance is recognized to be absolutely neces-sary ought, instead of forming the conclusion ofmy inquiries, to be the strongest motive to thepursuit of them. All ignorance is either igno-rance of things or of the limits of knowledge. Ifmy ignorance is accidental and not necessary, itmust incite me, in the first case, to a dogmaticalinquiry regarding the objects of which I amignorant; in the second, to a critical investiga-tion into the bounds of all possible knowledge.But that my ignorance is absolutely necessaryand unavoidable, and that it consequently ab-solves from the duty of all further investiga-tion, is a fact which cannot be made out uponempirical grounds—from observation—butupon critical grounds alone, that is, by a tho-roughgoing investigation into the primarysources of cognition. It follows that the deter-mination of the bounds of reason can be madeonly on a priori grounds; while the empiricallimitation of reason, which is merely an inde-

terminate cognition of an ignorance that cannever be completely removed, can take placeonly a posteriori. In other words, our empiricalknowledge is limited by that which yet remainsfor us to know. The former cognition of ourignorance, which is possible only on a rationalbasis, is a science; the latter is merely a percep-tion, and we cannot say how far the inferencesdrawn from it may extend. If I regard the earth,as it really appears to my senses, as a flat surfa-ce, I am ignorant how far this surface extends.But experience teaches me that, how far soeverI go, I always see before me a space in which Ican proceed farther; and thus I know the li-mits—merely visual—of my actual knowledgeof the earth, although I am ignorant of the li-mits of the earth itself. But if I have got so far asto know that the earth is a sphere, and that itssurface is spherical, I can cognize a priori anddetermine upon principles, from my knowled-ge of a small part of this surface—say to theextent of a degree—the diameter and circumfe-

rence of the earth; and although I am ignorantof the objects which this surface contains, Ihave a perfect knowledge of its limits and ex-tent.

The sum of all the possible objects of our cogni-tion seems to us to be a level surface, with anapparent horizon—that which forms the limitof its extent, and which has been termed by usthe idea of unconditioned totality. To reach thislimit by empirical means is impossible, and allattempts to determine it a priori according to aprinciple, are alike in vain. But all the questionsraised by pure reason relate to that which liesbeyond this horizon, or, at least, in its bounda-ry line.

The celebrated David Hume was one of thosegeographers of human reason who believe thatthey have given a sufficient answer to all suchquestions by declaring them to lie beyond thehorizon of our knowledge—a horizon which,however, Hume was unable to determine. His

attention especially was directed to the princi-ple of causality; and he remarked with perfectjustice that the truth of this principle, and eventhe objective validity of the conception of acause, was not commonly based upon clearinsight, that is, upon a priori cognition. Hencehe concluded that this law does not derive itsauthority from its universality and necessity,but merely from its general applicability in thecourse of experience, and a kind of subjectivenecessity thence arising, which he termed habit.From the inability of reason to establish thisprinciple as a necessary law for the acquisitionof all experience, he inferred the nullity of allthe attempts of reason to pass the region of theempirical.

This procedure of subjecting the facta of reasonto examination, and, if necessary, to disappro-val, may be termed the censura of reason. Thiscensura must inevitably lead us to doubts re-garding all transcendent employment of prin-

ciples. But this is only the second step in ourinquiry. The first step in regard to the subjectsof pure reason, and which marks the infancy ofthat faculty, is that of dogmatism. The second,which we have just mentioned, is that of scepti-cism, and it gives evidence that our judgementhas been improved by experience. But a thirdstep is necessary—indicative of the maturityand manhood of the judgement, which nowlays a firm foundation upon universal and ne-cessary principles. This is the period of criti-cism, in which we do not examine the facta ofreason, but reason itself, in the whole extent ofits powers, and in regard to its capability of apriori cognition; and thus we determine notmerely the empirical and ever-shifting boundsof our knowledge, but its necessary and eternallimits. We demonstrate from indubitable prin-ciples, not merely our ignorance in respect tothis or that subject, but in regard to all possiblequestions of a certain class. Thus scepticism is aresting place for reason, in which it may reflect

on its dogmatical wanderings and gain someknowledge of the region in which it happens tobe, that it may pursue its way with greater cer-tainty; but it cannot be its permanent dwelling-place. It must take up its abode only in the re-gion of complete certitude, whether this relatesto the cognition of objects themselves, or to thelimits which bound all our cognition.

Reason is not to be considered as an indefinite-ly extended plane, of the bounds of which wehave only a general knowledge; it ought ratherto be compared to a sphere, the radius of whichmay be found from the curvature of its surfa-ce—that is, the nature of a priori syntheticalpropositions—and, consequently, its circumfe-rence and extent. Beyond the sphere of expe-rience there are no objects which it can cognize;nay, even questions regarding such suppositi-tious objects relate only to the subjective prin-ciples of a complete determination of the rela-

tions which exist between the understanding-conceptions which lie within this sphere.

We are actually in possession of a priori synt-hetical cognitions, as is proved by the existenceof the principles of the understanding, whichanticipate experience. If any one cannot com-prehend the possibility of these principles, hemay have some reason to doubt whether theyare really a priori; but he cannot on this accountdeclare them to be impossible, and affirm thenullity of the steps which reason may have ta-ken under their guidance. He can only say: Ifwe perceived their origin and their authentici-ty, we should be able to determine the extentand limits of reason; but, till we can do this, allpropositions regarding the latter are mere ran-dom assertions. In this view, the doubt respec-ting all dogmatical philosophy, which proceedswithout the guidance of criticism, is wellgrounded; but we cannot therefore deny toreason the ability to construct a sound philo-

sophy, when the way has been prepared by athorough critical investigation. All the concep-tions produced, and all the questions raised, bypure reason, do not lie in the sphere of expe-rience, but in that of reason itself, and hencethey must be solved, and shown to be eithervalid or inadmissible, by that faculty. We haveno right to decline the solution of such pro-blems, on the ground that the solution can bediscovered only from the nature of things, andunder pretence of the limitation of human fa-culties, for reason is the sole creator of all theseideas, and is therefore bound either to establishtheir validity or to expose their illusory nature.

The polemic of scepticism is properly directedagainst the dogmatist, who erects a system ofphilosophy without having examined the fun-damental objective principles on which it isbased, for the purpose of evidencing the futilityof his designs, and thus bringing him to aknowledge of his own powers. But, in itself,

scepticism does not give us any certain infor-mation in regard to the bounds of our know-ledge. All unsuccessful dogmatical attempts ofreason are facia, which it is always useful tosubmit to the censure of the sceptic. But thiscannot help us to any decision regarding theexpectations which reason cherishes of bettersuccess in future endeavours; the investigationsof scepticism cannot, therefore, settle the dispu-te regarding the rights and powers of humanreason.

Hume is perhaps the ablest and most ingeniousof all sceptical philosophers, and his writingshave, undoubtedly, exerted the most powerfulinfluence in awakening reason to a thoroughinvestigation into its own powers. It will, there-fore, well repay our labours to consider for alittle the course of reasoning which he followedand the errors into which he strayed, althoughsetting out on the path of truth and certitude.

Hume was probably aware, although he neverclearly developed the notion, that we proceedin judgements of a certain class beyond ourconception if the object. I have termed this kindof judgement synthetical. As regard the mannerin which I pass beyond my conception by theaid of experience, no doubts can be entertained.Experience is itself a synthesis of perceptions;and it employs perceptions to increment theconception, which I obtain by means of anotherperception. But we feel persuaded that we areable to proceed beyond a conception, and toextend our cognition a priori. We attempt thisin two ways—either, through the pure unders-tanding, in relation to that which may becomean object of experience, or, through pure rea-son, in relation to such properties of things, orof the existence of things, as can never be pre-sented in any experience. This sceptical philo-sopher did not distinguish these two kinds ofjudgements, as he ought to have done, but re-garded this augmentation of conceptions, and,

if we may so express ourselves, the sponta-neous generation of understanding and reason,independently of the impregnation of experien-ce, as altogether impossible. The so-called apriori principles of these faculties he conse-quently held to be invalid and imaginary, andregarded them as nothing but subjective habitsof thought originating in experience, and there-fore purely empirical and contingent rules, towhich we attribute a spurious necessity anduniversality. In support of this strange asser-tion, he referred us to the generally acknow-ledged principle of the relation between causeand effect. No faculty of the mind can conductus from the conception of a thing to the existen-ce of something else; and hence he believed hecould infer that, without experience, we pos-sess no source from which we can augment aconception, and no ground sufficient to justifyus in framing a judgement that is to extend ourcognition a priori. That the light of the sun,which shines upon a piece of wax, at the same

time melts it, while it hardens clay, no power ofthe understanding could infer from the concep-tions which we previously possessed of thesesubstances; much less is there any a priori lawthat could conduct us to such a conclusion,which experience alone can certify. On the ot-her hand, we have seen in our discussion oftranscendental logic, that, although we can ne-ver proceed immediately beyond the content ofthe conception which is given us, we can al-ways cognize completely a priori—in relation,however, to a third term, namely, possible ex-perience—the law of its connection with otherthings. For example, if I observe that a piece ofwax melts, I can cognize a priori that theremust have been something (the sun's heat) pre-ceding, which this law; although, without theaid of experience, I could not cognize a prioriand in a determinate manner either the causefrom the effect, or the effect from the cause.Hume was, therefore, wrong in inferring, fromthe contingency of the determination according

to law, the contingency of the law itself; and thepassing beyond the conception of a thing topossible experience (which is an a priori pro-ceeding, constituting the objective reality of theconception), he confounded with our synthesisof objects in actual experience, which is always,of course, empirical. Thus, too, he regarded theprinciple of affinity, which has its seat in theunderstanding and indicates a necessary con-nection, as a mere rule of association, lying inthe imitative faculty of imagination, which canpresent only contingent, and not objective con-nections.

The sceptical errors of this remarkably acutethinker arose principally from a defect, whichwas common to him with the dogmatists, na-mely, that he had never made a systematic re-view of all the different kinds of a priori synt-hesis performed by the understanding. Had hedone so, he would have found, to take oneexample among many, that the principle of

permanence was of this character, and that it,as well as the principle of causality, anticipatesexperience. In this way he might have been ableto describe the determinate limits of the a priorioperations of understanding and reason. But hemerely declared the understanding to be limi-ted, instead of showing what its limits were; hecreated a general mistrust in the power of ourfaculties, without giving us any determinateknowledge of the bounds of our necessary andunavoidable ignorance; he examined and con-demned some of the principles of the unders-tanding, without investigating all its powerswith the completeness necessary to criticism.He denies, with truth, certain powers to theunderstanding, but he goes further, and decla-res it to be utterly inadequate to the a prioriextension of knowledge, although he has notfully examined all the powers which reside inthe faculty; and thus the fate which alwaysovertakes scepticism meets him too. That is tosay, his own declarations are doubted, for his

objections were based upon facta, which arecontingent, and not upon principles, which canalone demonstrate the necessary invalidity ofall dogmatical assertions.

As Hume makes no distinction between thewell-grounded claims of the understandingand the dialectical pretensions of reason,against which, however, his attacks are mainlydirected, reason does not feel itself shut outfrom all attempts at the extension of a prioricognition, and hence it refuses, in spite of a fewchecks in this or that quarter, to relinquish suchefforts. For one naturally arms oneself to resistan attack, and becomes more obstinate in theresolve to establish the claims he has advanced.But a complete review of the powers of reason,and the conviction thence arising that we are inpossession of a limited field of action, while wemust admit the vanity of higher claims, puts anend to all doubt and dispute, and induces rea-

son to rest satisfied with the undisturbed pos-session of its limited domain.

To the uncritical dogmatist, who has not surve-yed the sphere of his understanding, nor de-termined, in accordance with principles, thelimits of possible cognition, who, consequently,is ignorant of his own powers, and believes hewill discover them by the attempts he makes inthe field of cognition, these attacks of scepti-cism are not only dangerous, but destructive.For if there is one proposition in his chain ofreasoning which be he cannot prove, or thefallacy in which he cannot evolve in accordancewith a principle, suspicion falls on all his sta-tements, however plausible they may appear.

And thus scepticism, the bane of dogmaticalphilosophy, conducts us to a sound investiga-tion into the understanding and the reason.When we are thus far advanced, we need fearno further attacks; for the limits of our domainare clearly marked out, and we can make no

claims nor become involved in any disputesregarding the region that lies beyond these li-mits. Thus the sceptical procedure in philosop-hy does not present any solution of the pro-blems of reason, but it forms an excellent exer-cise for its powers, awakening its circumspec-tion, and indicating the means whereby it maymost fully establish its claims to its legitimatepossessions.

SECTION III. The Discipline of Pure Rea-son in Hypothesis.

This critique of reason has now taught us thatall its efforts to extend the bounds of knowled-ge, by means of pure speculation, are utterlyfruitless. So much the wider field, it may ap-pear, lies open to hypothesis; as, where we can-not know with certainty, we are at liberty tomake guesses and to form suppositions.

Imagination may be allowed, under the strictsurveillance of reason, to invent suppositions;but, these must be based on something that isperfectly certain—and that is the possibility ofthe object. If we are well assured upon thispoint, it is allowable to have recourse to suppo-sition in regard to the reality of the object; butthis supposition must, unless it is utterlygroundless, be connected, as its ground of ex-planation, with that which is really given andabsolutely certain. Such a supposition is termeda hypothesis.

It is beyond our power to form the least con-ception a priori of the possibility of dynamicalconnection in phenomena; and the category ofthe pure understanding will not enable us toexcogitate any such connection, but merelyhelps us to understand it, when we meet with itin experience. For this reason we cannot, inaccordance with the categories, imagine or in-vent any object or any property of an object not

given, or that may not be given in experience,and employ it in a hypothesis; otherwise, weshould be basing our chain of reasoning uponmere chimerical fancies, and not upon concep-tions of things. Thus, we have no right to as-sume the existence of new powers, not existingin nature—for example, an understanding witha non-sensuous intuition, a force of attractionwithout contact, or some new kind of substan-ces occupying space, and yet without the pro-perty of impenetrability—and, consequently,we cannot assume that there is any other kindof community among substances than that ob-servable in experience, any kind of presencethan that in space, or any kind of duration thanthat in time. In one word, the conditions of pos-sible experience are for reason the only condi-tions of the possibility of things; reason cannotventure to form, independently of these condi-tions, any conceptions of things, because suchconceptions, although not self-contradictory,are without object and without application.

The conceptions of reason are, as we have al-ready shown, mere ideas, and do not relate toany object in any kind of experience. At thesame time, they do not indicate imaginary orpossible objects. They are purely problematicalin their nature and, as aids to the heuristic exer-cise of the faculties, form the basis of the regu-lative principles for the systematic employmentof the understanding in the field of experience.If we leave this ground of experience, they be-come mere fictions of thought, the possibility ofwhich is quite indemonstrable; and they can-not, consequently, be employed as hypothesesin the explanation of real phenomena. It is quiteadmissible to cogitate the soul as simple, for thepurpose of enabling ourselves to employ theidea of a perfect and necessary unity of all thefaculties of the mind as the principle of all ourinquiries into its internal phenomena, althoughwe cannot cognize this unity in concreto. But toassume that the soul is a simple substance (atranscendental conception) would be enoun-

cing a proposition which is not only indemons-trable—as many physical hypotheses are—buta proposition which is purely arbitrary, and inthe highest degree rash. The simple is neverpresented in experience; and, if by substance ishere meant the permanent object of sensuousintuition, the possibility of a simple phenome-non is perfectly inconceivable. Reason affordsno good grounds for admitting the existence ofintelligible beings, or of intelligible propertiesof sensuous things, although—as we have noconception either of their possibility or of theirimpossibility—it will always be out of our po-wer to affirm dogmatically that they do notexist. In the explanation of given phenomena,no other things and no other grounds of expla-nation can be employed than those which standin connection with the given phenomena ac-cording to the known laws of experience. Atranscendental hypothesis, in which a mereidea of reason is employed to explain the phe-nomena of nature, would not give us any better

insight into a phenomenon, as we should betrying to explain what we do not sufficientlyunderstand from known empirical principles,by what we do not understand at all. The prin-ciples of such a hypothesis might conduce tothe satisfaction of reason, but it would not as-sist the understanding in its application to ob-jects. Order and conformity to aims in the sphe-re of nature must be themselves explainedupon natural grounds and according to naturallaws; and the wildest hypotheses, if they areonly physical, are here more admissible than ahyperphysical hypothesis, such as that of a di-vine author. For such a hypothesis would in-troduce the principle of ignava ratio, whichrequires us to give up the search for causes thatmight be discovered in the course of experienceand to rest satisfied with a mere idea. As re-gards the absolute totality of the grounds ofexplanation in the series of these causes, thiscan be no hindrance to the understanding inthe case of phenomena; because, as they are to

us nothing more than phenomena, we have noright to look for anything like completeness inthe synthesis of the series of their conditions.

Transcendental hypotheses are therefore inad-missible; and we cannot use the liberty of em-ploying, in the absence of physical, hyperphy-sical grounds of explanation. And this for tworeasons; first, because such hypothesis do notadvance reason, but rather stop it in its pro-gress; secondly, because this licence wouldrender fruitless all its exertions in its own pro-per sphere, which is that of experience. For,when the explanation of natural phenomenahappens to be difficult, we have constantly athand a transcendental ground of explanation,which lifts us above the necessity of investiga-ting nature; and our inquiries are brought to aclose, not because we have obtained all the re-quisite knowledge, but because we abut upon aprinciple which is incomprehensible andwhich, indeed, is so far back in the track of

thought as to contain the conception of the ab-solutely primal being.

The next requisite for the admissibility of a hy-pothesis is its sufficiency. That is, it must de-termine a priori the consequences which aregiven in experience and which are supposed tofollow from the hypothesis itself. If we requireto employ auxiliary hypotheses, the suspicionnaturally arises that they are mere fictions; be-cause the necessity for each of them requiresthe same justification as in the case of the origi-nal hypothesis, and thus their testimony is in-valid. If we suppose the existence of an infinite-ly perfect cause, we possess sufficient groundsfor the explanation of the conformity to aims,the order and the greatness which we observein the universe; but we find ourselves obliged,when we observe the evil in the world and theexceptions to these laws, to employ new hy-pothesis in support of the original one. We em-ploy the idea of the simple nature of the human

soul as the foundation of all the theories wemay form of its phenomena; but when we meetwith difficulties in our way, when we observein the soul phenomena similar to the changeswhich take place in matter, we require to call innew auxiliary hypotheses. These may, indeed,not be false, but we do not know them to betrue, because the only witness to their certitudeis the hypothesis which they themselves havebeen called in to explain.

We are not discussing the above-mentionedassertions regarding the immaterial unity of thesoul and the existence of a Supreme Being asdogmata, which certain philosophers profess todemonstrate a priori, but purely as hypotheses.In the former case, the dogmatist must takecare that his arguments possess the apodeicticcertainty of a demonstration. For the assertionthat the reality of such ideas is probable is asabsurd as a proof of the probability of a propo-sition in geometry. Pure abstract reason, apart

from all experience, can either cognize nothingat all; and hence the judgements it enounces arenever mere opinions, they are either apodeicticcertainties, or declarations that nothing can beknown on the subject. Opinions and probablejudgements on the nature of things can only beemployed to explain given phenomena, or theymay relate to the effect, in accordance with em-pirical laws, of an actually existing cause. Inother words, we must restrict the sphere of opi-nion to the world of experience and nature.Beyond this region opinion is mere invention;unless we are groping about for the truth on apath not yet fully known, and have some hopesof stumbling upon it by chance.

But, although hypotheses are inadmissible inanswers to the questions of pure speculativereason, they may be employed in the defence ofthese answers. That is to say, hypotheses areadmissible in polemic, but not in the sphere ofdogmatism. By the defence of statements of this

character, I do not mean an attempt at discove-ring new grounds for their support, but merelythe refutation of the arguments of opponents.All a priori synthetical propositions possess thepeculiarity that, although the philosopher whomaintains the reality of the ideas contained inthe proposition is not in possession of sufficientknowledge to establish the certainty of his sta-tements, his opponent is as little able to provethe truth of the opposite. This equality of fortu-ne does not allow the one party to be superiorto the other in the sphere of speculative cogni-tion; and it is this sphere, accordingly, that isthe proper arena of these endless speculativeconflicts. But we shall afterwards show that, inrelation to its practical exercise, Reason has theright of admitting what, in the field of purespeculation, she would not be justified in sup-posing, except upon perfectly sufficientgrounds; because all such suppositions destroythe necessary completeness of speculation—acondition which the practical reason, however,

does not consider to be requisite. In this sphere,therefore, Reason is mistress of a possession,her title to which she does not require to pro-ve—which, in fact, she could not do. The bur-den of proof accordingly rests upon the oppo-nent. But as he has just as little knowledge re-garding the subject discussed, and is as littleable to prove the non-existence of the object ofan idea, as the philosopher on the other side isto demonstrate its reality, it is evident that the-re is an advantage on the side of the philosop-her who maintains his proposition as a practi-cally necessary supposition (melior est conditiopossidentis). For he is at liberty to employ, inself-defence, the same weapons as his opponentmakes use of in attacking him; that is, he has aright to use hypotheses not for the purpose ofsupporting the arguments in favour of his ownpropositions, but to show that his opponentknows no more than himself regarding the sub-ject under 'discussion and cannot boast of anyspeculative advantage.

Hypotheses are, therefore, admissible in thesphere of pure reason only as weapons for self-defence, and not as supports to dogmatical as-sertions. But the opposing party we must al-ways seek for in ourselves. For speculative rea-son is, in the sphere of transcendentalism, dia-lectical in its own nature. The difficulties andobjections we have to fear lie in ourselves. Theyare like old but never superannuated claims;and we must seek them out, and settle themonce and for ever, if we are to expect a perma-nent peace. External tranquility is hollow andunreal. The root of these contradictions, whichlies in the nature of human reason, must bedestroyed; and this can only be done by givingit, in the first instance, freedom to grow, nay,by nourishing it, that it may send out shoots,and thus betray its own existence. It is our du-ty, therefore, to try to discover new objections,to put weapons in the bands of our opponent,and to grant him the most favourable positionin the arena that he can wish. We have nothing

to fear from these concessions; on the contrary,we may rather hope that we shall thus makeourselves master of a possession which no onewill ever venture to dispute.

The thinker requires, to be fully equipped, thehypotheses of pure reason, which, although butleaden weapons (for they have not been steeledin the armoury of experience), are as useful asany that can be employed by his opponents. If,accordingly, we have assumed, from a non-speculative point of view, the immaterial natu-re of the soul, and are met by the objection thatexperience seems to prove that the growth anddecay of our mental faculties are mere modifi-cations of the sensuous organism—we can wea-ken the force of this objection by the assump-tion that the body is nothing but the fundamen-tal phenomenon, to which, as a necessary con-dition, all sensibility, and consequently allthought, relates in the present state of our exis-tence; and that the separation of soul and body

forms the conclusion of the sensuous exerciseof our power of cognition and the beginning ofthe intellectual. The body would, in this view ofthe question, be regarded, not as the cause ofthought, but merely as its restrictive condition,as promotive of the sensuous and animal, butas a hindrance to the pure and spiritual life;and the dependence of the animal life on theconstitution of the body, would not prove thatthe whole life of man was also dependent onthe state of the organism. We might go stillfarther, and discover new objections, or carryout to their extreme consequences those whichhave already been adduced.

Generation, in the human race as well asamong the irrational animals, depends on somany accidents—of occasion, of proper suste-nance, of the laws enacted by the governmentof a country of vice even, that it is difficult tobelieve in the eternal existence of a being whoselife has begun under circumstances so mean

and trivial, and so entirely dependent upon ourown control. As regards the continuance of theexistence of the whole race, we need have nodifficulties, for accident in single cases is sub-ject to general laws; but, in the case of each in-dividual, it would seem as if we could hardlyexpect so wonderful an effect from causes soinsignificant. But, in answer to these objections,we may adduce the transcendental hypothesisthat all life is properly intelligible, and not sub-ject to changes of time, and that it neither beganin birth, nor will end in death. We may assumethat this life is nothing more than a sensuousrepresentation of pure spiritual life; that thewhole world of sense is but an image, hoveringbefore the faculty of cognition which we exerci-se in this sphere, and with no more objectivereality than a dream; and that if we could intui-te ourselves and other things as they really are,we should see ourselves in a world of spiritualnatures, our connection with which did not

begin at our birth and will not cease with thedestruction of the body. And so on.

We cannot be said to know what has been abo-ve asserted, nor do we seriously maintain thetruth of these assertions; and the notions the-rein indicated are not even ideas of reason, theyare purely fictitious conceptions. But this hy-pothetical procedure is in perfect conformitywith the laws of reason. Our opponent mista-kes the absence of empirical conditions for aproof of the complete impossibility of all thatwe have asserted; and we have to show himthat he has not exhausted the whole sphere ofpossibility and that he can as little compass thatsphere by the laws of experience and nature, aswe can lay a secure foundation for the opera-tions of reason beyond the region of experience.Such hypothetical defences against the preten-sions of an opponent must not be regarded asdeclarations of opinion. The philosopher aban-dons them, so soon as the opposite party re-

nounces its dogmatical conceit. To maintain asimply negative position in relation to proposi-tions which rest on an insecure foundation,well befits the moderation of a true philosop-her; but to uphold the objections urged againstan opponent as proofs of the opposite state-ment is a proceeding just as unwarrantable andarrogant as it is to attack the position of a phi-losopher who advances affirmative proposi-tions regarding such a subject.

It is evident, therefore, that hypotheses, in thespeculative sphere, are valid, not as indepen-dent propositions, but only relatively to opposi-te transcendent assumptions. For, to make theprinciples of possible experience conditions ofthe possibility of things in general is just astranscendent a procedure as to maintain theobjective reality of ideas which can be appliedto no objects except such as lie without the li-mits of possible experience. The judgementsenounced by pure reason must be necessary, or

they must not be enounced at all. Reason can-not trouble herself with opinions. But the hy-potheses we have been discussing are merelyproblematical judgements, which can neitherbe confuted nor proved; while, therefore, theyare not personal opinions, they are indispensa-ble as answers to objections which are liable tobe raised. But we must take care to confinethem to this function, and guard against anyassumption on their part of absolute validity, aproceeding which would involve reason ininextricable difficulties and contradictions.

SECTION IV. The Discipline of Pure Rea-son in Relation to Proofs.

It is a peculiarity, which distinguishes theproofs of transcendental synthetical proposi-tions from those of all other a priori syntheticalcognitions, that reason, in the case of the for-

mer, does not apply its conceptions directly toan object, but is first obliged to prove, a priori,the objective validity of these conceptions andthe possibility of their syntheses. This is notmerely a prudential rule, it is essential to thevery possibility of the proof of a transcendentalproposition. If I am required to pass, a priori,beyond the conception of an object, I find that itis utterly impossible without the guidance ofsomething which is not contained in the con-ception. In mathematics, it is a priori intuitionthat guides my synthesis; and, in this case, allour conclusions may be drawn immediatelyfrom pure intuition. In transcendental cogni-tion, so long as we are dealing only with con-ceptions of the understanding, we are guidedby possible experience. That is to say, a proof inthe sphere of transcendental cognition does notshow that the given conception (that of anevent, for example) leads directly to anotherconception (that of a cause)—for this would bea saltus which nothing can justify; but it shows

that experience itself, and consequently theobject of experience, is impossible without theconnection indicated by these conceptions. Itfollows that such a proof must demonstrate thepossibility of arriving, synthetically and a prio-ri, at a certain knowledge of things, which wasnot contained in our conceptions of thesethings. Unless we pay particular attention tothis requirement, our proofs, instead of pur-suing the straight path indicated by reason,follow the tortuous road of mere subjectiveassociation. The illusory conviction, which restsupon subjective causes of association, andwhich is considered as resulting from the per-ception of a real and objective natural affinity,is always open to doubt and suspicion. For thisreason, all the attempts which have been madeto prove the principle of sufficient reason, have,according to the universal admission of philo-sophers, been quite unsuccessful; and, beforethe appearance of transcendental criticism, itwas considered better, as this principle could

not be abandoned, to appeal boldly to thecommon sense of mankind (a proceedingwhich always proves that the problem, whichreason ought to solve, is one in which philo-sophers find great difficulties), rather than at-tempt to discover new dogmatical proofs.

But, if the proposition to be proved is a propo-sition of pure reason, and if I aim at passingbeyond my empirical conceptions by the aid ofmere ideas, it is necessary that the proof shouldfirst show that such a step in synthesis is possi-ble (which it is not), before it proceeds to provethe truth of the proposition itself. The so-calledproof of the simple nature of the soul from theunity of apperception, is a very plausible one.But it contains no answer to the objection, that,as the notion of absolute simplicity is not a con-ception which is directly applicable to a percep-tion, but is an idea which must be inferred—ifat all—from observation, it is by no means evi-dent how the mere fact of consciousness, which

is contained in all thought, although in so far asimple representation, can conduct me to theconsciousness and cognition of a thing which ispurely a thinking substance. When I representto my mind the power of my body as in mo-tion, my body in this thought is so far absoluteunity, and my representation of it is a simpleone; and hence I can indicate this representa-tion by the motion of a point, because I havemade abstraction of the size or volume of thebody. But I cannot hence infer that, given mere-ly the moving power of a body, the body maybe cogitated as simple substance, merely becau-se the representation in my mind takes no ac-count of its content in space, and is consequen-tly simple. The simple, in abstraction, is verydifferent from the objectively simple; and hencethe Ego, which is simple in the first sense, may,in the second sense, as indicating the soul itself,be a very complex conception, with a very va-rious content. Thus it is evident that in all sucharguments there lurks a paralogism. We guess

(for without some such surmise our suspicionwould not be excited in reference to a proof ofthis character) at the presence of the paralo-gism, by keeping ever before us a criterion ofthe possibility of those synthetical propositionswhich aim at proving more than experience canteach us. This criterion is obtained from theobservation that such proofs do not lead usdirectly from the subject of the proposition tobe proved to the required predicate, but find itnecessary to presuppose the possibility of ex-tending our cognition a priori by means ofideas. We must, accordingly, always use thegreatest caution; we require, before attemptingany proof, to consider how it is possible to ex-tend the sphere of cognition by the operationsof pure reason, and from what source we are toderive knowledge, which is not obtained fromthe analysis of conceptions, nor relates, by anti-cipation, to possible experience. We shall thusspare ourselves much severe and fruitless la-bour, by not expecting from reason what is be-

yond its power, or rather by subjecting it todiscipline, and teaching it to moderate its ve-hement desires for the extension of the sphereof cognition.

The first rule for our guidance is, therefore, notto attempt a transcendental proof, before wehave considered from what source we are toderive the principles upon which the proof is tobe based, and what right we have to expect thatour conclusions from these principles will beveracious. If they are principles of the unders-tanding, it is vain to expect that we should at-tain by their means to ideas of pure reason; forthese principles are valid only in regard to ob-jects of possible experience. If they are princi-ples of pure reason, our labour is alike in vain.For the principles of reason, if employed asobjective, are without exception dialectical andpossess no validity or truth, except as regulati-ve principles of the systematic employment ofreason in experience. But when such delusive

proof are presented to us, it is our duty to meetthem with the non liquet of a matured judge-ment; and, although we are unable to exposethe particular sophism upon which the proof isbased, we have a right to demand a deductionof the principles employed in it; and, if theseprinciples have their origin in pure reason alo-ne, such a deduction is absolutely impossible.And thus it is unnecessary that we shouldtrouble ourselves with the exposure and confu-tation of every sophistical illusion; we may, atonce, bring all dialectic, which is inexhaustiblein the production of fallacies, before the bar ofcritical reason, which tests the principles uponwhich all dialectical procedure is based. Thesecond peculiarity of transcendental proof isthat a transcendental proposition cannot restupon more than a single proof. If I am drawingconclusions, not from conceptions, but fromintuition corresponding to a conception, be itpure intuition, as in mathematics, or empirical,as in natural science, the intuition which forms

the basis of my inferences presents me withmaterials for many synthetical propositions,which I can connect in various modes, while, asit is allowable to proceed from different pointsin the intention, I can arrive by different pathsat the same proposition.

But every transcendental proposition sets outfrom a conception, and posits the syntheticalcondition of the possibility of an object accor-ding to this conception. There must, therefore,be but one ground of proof, because it is theconception alone which determines the object;and thus the proof cannot contain anythingmore than the determination of the object ac-cording to the conception. In our Transcenden-tal Analytic, for example, we inferred the prin-ciple: Every event has a cause, from the onlycondition of the objective possibility of our con-ception of an event. This is that an event cannotbe determined in time, and consequently can-not form a part of experience, unless it stands

under this dynamical law. This is the only pos-sible ground of proof; for our conception of anevent possesses objective validity, that is, is atrue conception, only because the law of causa-lity determines an object to which it can refer.Other arguments in support of this principlehave been attempted—such as that from thecontingent nature of a phenomenon; but whenthis argument is considered, we can discoverno criterion of contingency, except the fact ofan event—of something happening, that is tosay, the existence which is preceded by thenon-existence of an object, and thus we fallback on the very thing to be proved. If the pro-position: "Every thinking being is simple," is tobe proved, we keep to the conception of theego, which is simple, and to which all thoughthas a relation. The same is the case with thetranscendental proof of the existence of a Deity,which is based solely upon the harmony andreciprocal fitness of the conceptions of an ens

realissimum and a necessary being, and cannotbe attempted in any other manner.

This caution serves to simplify very much thecriticism of all propositions of reason. Whenreason employs conceptions alone, only oneproof of its thesis is possible, if any. When, the-refore, the dogmatist advances with ten argu-ments in favour of a proposition, we may besure that not one of them is conclusive. For if hepossessed one which proved the proposition hebrings forward to demonstration- as must al-ways be the case with the propositions of purereason- what need is there for any more? Hisintention can only be similar to that of the ad-vocate who had different arguments for diffe-rent judges; this availing himself of the weak-ness of those who examine his arguments, who,without going into any profound investigation,adopt the view of the case which seems mostprobable at first sight and decide according toit.

The third rule for the guidance of pure reasonin the conduct of a proof is that all transcenden-tal proofs must never be apagogic or indirect,but always ostensive or direct. The direct orostensive proof not only establishes the truth ofthe proposition to be proved, but exposes thegrounds of its truth; the apagogic, on the otherhand, may assure us of the truth of the proposi-tion, but it cannot enable us to comprehend thegrounds of its possibility. The latter is, accor-dingly, rather an auxiliary to an argument, thana strictly philosophical and rational mode ofprocedure. In one respect, however, they havean advantage over direct proofs, from the factthat the mode of arguing by contradiction,which they employ, renders our understandingof the question more clear, and approximatesthe proof to the certainty of an intuitional de-monstration.

The true reason why indirect proofs are emplo-yed in different sciences is this. When the

grounds upon which we seek to base a cogni-tion are too various or too profound, we trywhether or not we may not discover the truthof our cognition from its consequences. Themodus ponens of reasoning from the truth ofits inferences to the truth of a propositionwould be admissible if all the inferences thatcan be drawn from it are known to be true; forin this case there can be only one possibleground for these inferences, and that is the trueone. But this is a quite impracticable procedure,as it surpasses all our powers to discover all thepossible inferences that can be drawn from aproposition. But this mode of reasoning is em-ployed, under favour, when we wish to provethe truth of an hypothesis; in which case weadmit the truth of the conclusion- which issupported by analogy—that, if all the inferen-ces we have drawn and examined agree withthe proposition assumed, all other possible in-ferences will also agree with it. But, in this way,an hypothesis can never be established as a

demonstrated truth. The modus tollens of rea-soning from known inferences to the unknownproposition, is not only a rigorous, but a veryeasy mode of proof. For, if it can be shown thatbut one inference from a proposition is false,then the proposition must itself be false. Ins-tead, then, of examining, in an ostensive argu-ment, the whole series of the grounds on whichthe truth of a proposition rests, we need onlytake the opposite of this proposition, and if oneinference from it be false, then must the opposi-te be itself false; and, consequently, the propo-sition which we wished to prove must be true.

The apagogic method of proof is admissibleonly in those sciences where it is impossible tomistake a subjective representation for an ob-jective cognition. Where this is possible, it isplain that the opposite of a given propositionmay contradict merely the subjective conditionsof thought, and not the objective cognition; or itmay happen that both propositions contradict

each other only under a subjective condition,which is incorrectly considered to be objective,and, as the condition is itself false, both propo-sitions may be false, and it will, consequently,be impossible to conclude the truth of the onefrom the falseness of the other.

In mathematics such subreptions are impossi-ble; and it is in this science, accordingly, thatthe indirect mode of proof has its true place. Inthe science of nature, where all assertion is ba-sed upon empirical intuition, such subreptionsmay be guarded against by the repeated com-parison of observations; but this mode of proofis of little value in this sphere of knowledge.But the transcendental efforts of pure reasonare all made in the sphere of the subjective,which is the real medium of all dialectical illu-sion; and thus reason endeavours, in its pre-misses, to impose upon us subjective represen-tations for objective cognitions. In the transcen-dental sphere of pure reason, then, and in the

case of synthetical propositions, it is inadmissi-ble to support a statement by disproving thecounter-statement. For only two cases are pos-sible; either, the counter-statement is nothingbut the enouncement of the inconsistency of theopposite opinion with the subjective conditionsof reason, which does not affect the real case(for example, we cannot comprehend the un-conditioned necessity of the existence of abeing, and hence every speculative proof of theexistence of such a being must be opposed onsubjective grounds, while the possibility of thisbeing in itself cannot with justice be denied); or,both propositions, being dialectical in their na-ture, are based upon an impossible conception.In this latter case the rule applies: non entisnulla sunt predicata; that is to say, what weaffirm and what we deny, respecting such anobject, are equally untrue, and the apagogicmode of arriving at the truth is in this case im-possible. If, for example, we presuppose thatthe world of sense is given in itself in its totali-

ty, it is false, either that it is infinite, or that it isfinite and limited in space. Both are false, be-cause the hypothesis is false. For the notion ofphenomena (as mere representations) whichare given in themselves (as objects) is self-contradictory; and the infinitude of this imagi-nary whole would, indeed, be unconditioned,but would be inconsistent (as everything in thephenomenal world is conditioned) with theunconditioned determination and finitude ofquantities which is presupposed in our concep-tion.

The apagogic mode of proof is the true sourceof those illusions which have always had sostrong an attraction for the admirers of dogma-tical philosophy. It may be compared to achampion who maintains the honour andclaims of the party he has adopted by offeringbattle to all who doubt the validity of theseclaims and the purity of that honour; whilenothing can be proved in this way, except the

respective strength of the combatants, and theadvantage, in this respect, is always on the sideof the attacking party. Spectators, observingthat each party is alternately conqueror andconquered, are led to regard the subject of dis-pute as beyond the power of man to decideupon. But such an opinion cannot be justified;and it is sufficient to apply to these reasonersthe remark:

Non defensoribus istis Tempus eget.

Each must try to establish his assertions by atranscendental deduction of the grounds ofproof employed in his argument, and thus ena-ble us to see in what way the claims of reasonmay be supported. If an opponent bases hisassertions upon subjective grounds, he may berefuted with ease; not, however to the advanta-ge of the dogmatist, who likewise dependsupon subjective sources of cognition and is inlike manner driven into a corner by his oppo-

nent. But, if parties employ the direct methodof procedure, they will soon discover the diffi-culty, nay, the impossibility of proving theirassertions, and will be forced to appeal to pres-cription and precedence; or they will, by thehelp of criticism, discover with ease the dogma-tical illusions by which they had been mocked,and compel reason to renounce its exaggeratedpretensions to speculative insight and to confi-ne itself within the limits of its proper sphere—that of practical principles.

CHAPTER II. The Canon of Pure Reason.

It is a humiliating consideration for humanreason that it is incompetent to discover truthby means of pure speculation, but, on the con-trary, stands in need of discipline to check itsdeviations from the straight path and to exposethe illusions which it originates. But, on the

other hand, this consideration ought to elevateand to give it confidence, for this discipline isexercised by itself alone, and it is subject to thecensure of no other power. The bounds, mo-reover, which it is forced to set to its speculati-ve exercise, form likewise a check upon thefallacious pretensions of opponents; and thuswhat remains of its possessions, after theseexaggerated claims have been disallowed, issecure from attack or usurpation. The greatest,and perhaps the only, use of all philosophy ofpure reason is, accordingly, of a purely negati-ve character. It is not an organon for the exten-sion, but a discipline for the determination, ofthe limits of its exercise; and without layingclaim to the discovery of new truth, it has themodest merit of guarding against error.

At the same time, there must be some source ofpositive cognitions which belong to the domainof pure reason and which become the causes oferror only from our mistaking their true charac-

ter, while they form the goal towards whichreason continually strives. How else can weaccount for the inextinguishable desire in thehuman mind to find a firm footing in some re-gion beyond the limits of the world of expe-rience? It hopes to attain to the possession of aknowledge in which it has the deepest interest.It enters upon the path of pure speculation; butin vain. We have some reason, however, to ex-pect that, in the only other way that lies open toit—the path of practical reason—it may meetwith better success.

I understand by a canon a list of the a prioriprinciples of the proper employment of certainfaculties of cognition. Thus general logic, in itsanalytical department, is a formal canon for thefaculties of understanding and reason. In thesame way, Transcendental Analytic was seen tobe a canon of the pure understanding; for italone is competent to enounce true a priorisynthetical cognitions. But, when no proper

employment of a faculty of cognition is possi-ble, no canon can exist. But the synthetical cog-nition of pure speculative reason is, as has beenshown, completely impossible. There cannot,therefore, exist any canon for the speculativeexercise of this faculty—for its speculativeexercise is entirely dialectical; and, consequen-tly, transcendental logic, in this respect, is me-rely a discipline, and not a canon. If, then, thereis any proper mode of employing the faculty ofpure reason—in which case there must be acanon for this faculty—this canon will relate,not to the speculative, but to the practical use ofreason. This canon we now proceed to investi-gate.

SECTION I. Of the Ultimate End of the Pu-re Use of Reason.

There exists in the faculty of reason a naturaldesire to venture beyond the field of experien-ce, to attempt to reach the utmost bounds of allcognition by the help of ideas alone, and not torest satisfied until it has fulfilled its course andraised the sum of its cognitions into a self-subsistent systematic whole. Is the motive forthis endeavour to be found in its speculative, orin its practical interests alone?

Setting aside, at present, the results of the la-bours of pure reason in its speculative exercise,I shall merely inquire regarding the problemsthe solution of which forms its ultimate aim,whether reached or not, and in relation towhich all other aims are but partial and inter-mediate. These highest aims must, from thenature of reason, possess complete unity; ot-

herwise the highest interest of humanity couldnot be successfully promoted.

The transcendental speculation of reason rela-tes to three things: the freedom of the will, theimmortality of the soul, and the existence ofGod. The speculative interest which reason hasin those questions is very small; and, for itssake alone, we should not undertake the labourof transcendental investigation—a labour fullof toil and ceaseless struggle. We should be lothto undertake this labour, because the discove-ries we might make would not be of the sma-llest use in the sphere of concrete or physicalinvestigation. We may find out that the will isfree, but this knowledge only relates to the inte-lligible cause of our volition. As regards thephenomena or expressions of this will, that is,our actions, we are bound, in obedience to aninviolable maxim, without which reason cannotbe employed in the sphere of experience, toexplain these in the same way as we explain all

the other phenomena of nature, that is to say,according to its unchangeable laws. We mayhave discovered the spirituality and immortali-ty of the soul, but we cannot employ this know-ledge to explain the phenomena of this life, northe peculiar nature of the future, because ourconception of an incorporeal nature is purelynegative and does not add anything to ourknowledge, and the only inferences to bedrawn from it are purely fictitious. If, again, weprove the existence of a supreme intelligence,we should be able from it to make the confor-mity to aims existing in the arrangement of theworld comprehensible; but we should not bejustified in deducing from it any particulararrangement or disposition, or inferring anywhere it is not perceived. For it is a necessaryrule of the speculative use of reason that wemust not overlook natural causes, or refuse tolisten to the teaching of experience, for the sakeof deducing what we know and perceive fromsomething that transcends all our knowledge.

In one word, these three propositions are, forthe speculative reason, always transcendent,and cannot be employed as immanent princi-ples in relation to the objects of experience; theyare, consequently, of no use to us in this sphere,being but the valueless results of the severe butunprofitable efforts of reason.

If, then, the actual cognition of these three car-dinal propositions is perfectly useless, whileReason uses her utmost endeavours to induceus to admit them, it is plain that their real valueand importance relate to our practical, and notto our speculative interest.

I term all that is possible through free will,practical. But if the conditions of the exercise offree volition are empirical, reason can have on-ly a regulative, and not a constitutive, influenceupon it, and is serviceable merely for the intro-duction of unity into its empirical laws. In themoral philosophy of prudence, for example, thesole business of reason is to bring about a union

of all the ends, which are aimed at by our incli-nations, into one ultimate end—that of happi-ness—and to show the agreement whichshould exist among the means of attaining thatend. In this sphere, accordingly, reason cannotpresent to us any other than pragmatical lawsof free action, for our guidance towards theaims set up by the senses, and is incompetent togive us laws which are pure and determinedcompletely a priori. On the other hand, purepractical laws, the ends of which have beengiven by reason entirely a priori, and which arenot empirically conditioned, but are, on thecontrary, absolutely imperative in their nature,would be products of pure reason. Such are themoral laws; and these alone belong to the sphe-re of the practical exercise of reason, and admitof a canon.

All the powers of reason, in the sphere of whatmay be termed pure philosophy, are, in fact,directed to the three above-mentioned pro-

blems alone. These again have a still higherend—the answer to the question, what weought to do, if the will is free, if there is a Godand a future world. Now, as this problem rela-tes to our in reference to the highest aim ofhumanity, it is evident that the ultimate inten-tion of nature, in the constitution of our reason,has been directed to the moral alone.

We must take care, however, in turning ourattention to an object which is foreign* to thesphere of transcendental philosophy, not toinjure the unity of our system by digressions,nor, on the other hand, to fail in clearness, bysaying too little on the new subject of discus-sion. I hope to avoid both extremes, by keepingas close as possible to the transcendental, andexcluding all psychological, that is, empirical,elements.

[*Footnote: All practical conceptions relate toobjects of pleasure and pain, and consequen-tly—in an indirect manner, at least—to objects

of feeling. But as feeling is not a faculty of re-presentation, but lies out of the sphere of ourpowers of cognition, the elements of our jud-gements, in so far as they relate to pleasure orpain, that is, the elements of our practical jud-gements, do not belong to transcendental phi-losophy, which has to do with pure a prioricognitions alone.]

I have to remark, in the first place, that at pre-sent I treat of the conception of freedom in thepractical sense only, and set aside the corres-ponding transcendental conception, which can-not be employed as a ground of explanation inthe phenomenal world, but is itself a problemfor pure reason. A will is purely animal (arbi-trium brutum) when it is determined by sen-suous impulses or instincts only, that is, whenit is determined in a pathological manner. Awill, which can be determined independentlyof sensuous impulses, consequently by motivespresented by reason alone, is called a free will

(arbitrium liberum); and everything which isconnected with this free will, either as principleor consequence, is termed practical. The exis-tence of practical freedom can be proved fromexperience alone. For the human will is notdetermined by that alone which immediatelyaffects the senses; on the contrary, we have thepower, by calling up the notion of what is use-ful or hurtful in a more distant relation, ofovercoming the immediate impressions on oursensuous faculty of desire. But these considera-tions of what is desirable in relation to ourwhole state, that is, is in the end good and use-ful, are based entirely upon reason. This facul-ty, accordingly, enounces laws, which are im-perative or objective laws of freedom andwhich tell us what ought to take place, thusdistinguishing themselves from the laws ofnature, which relate to that which does takeplace. The laws of freedom or of free will arehence termed practical laws.

Whether reason is not itself, in the actual deli-very of these laws, determined in its turn byother influences, and whether the action which,in relation to sensuous impulses, we call free,may not, in relation to higher and more remoteoperative causes, really form a part of nature—these are questions which do not here concernus. They are purely speculative questions; andall we have to do, in the practical sphere, is toinquire into the rule of conduct which reasonhas to present. Experience demonstrates to usthe existence of practical freedom as one of thecauses which exist in nature, that is, it showsthe causal power of reason in the determinationof the will. The idea of transcendental freedom,on the contrary, requires that reason—in rela-tion to its causal power of commencing a seriesof phenomena—should be independent of allsensuous determining causes; and thus it seemsto be in opposition to the law of nature and toall possible experience. It therefore remains aproblem for the human mind. But this problem

does not concern reason in its practical use; andwe have, therefore, in a canon of pure reason,to do with only two questions, which relate tothe practical interest of pure reason: Is there aGod? and, Is there a future life? The question oftranscendental freedom is purely speculative,and we may therefore set it entirely aside whenwe come to treat of practical reason. Besides,we have already discussed this subject in theantinomy of pure reason.

SECTION II. Of the Ideal of the SummumBonum as a Determining Ground of the Ulti-mate End of Pure Reason.

Reason conducted us, in its speculative use,through the field of experience and, as it cannever find complete satisfaction in that sphere,from thence to speculative ideas—which,however, in the end brought us back again to

experience, and thus fulfilled the purpose ofreason, in a manner which, though useful, wasnot at all in accordance with our expectations. Itnow remains for us to consider whether purereason can be employed in a practical sphere,and whether it will here conduct us to thoseideas which attain the highest ends of pure rea-son, as we have just stated them. We shall thusascertain whether, from the point of view of itspractical interest, reason may not be able tosupply us with that which, on the speculativeside, it wholly denies us.

The whole interest of reason, speculative aswell as practical, is centred in the three follo-wing questions:

1. WHAT CAN I KNOW? 2. WHAT OUGHTI TO DO? 3. WHAT MAY I HOPE?

The first question is purely speculative. Wehave, as I flatter myself, exhausted all the re-plies of which it is susceptible, and have at last

found the reply with which reason must con-tent itself, and with which it ought to be con-tent, so long as it pays no regard to the practi-cal. But from the two great ends to the attain-ment of which all these efforts of pure reasonwere in fact directed, we remain just as far re-moved as if we had consulted our ease anddeclined the task at the outset. So far, then, asknowledge is concerned, thus much, at least, isestablished, that, in regard to those two pro-blems, it lies beyond our reach.

The second question is purely practical. Assuch it may indeed fall within the province ofpure reason, but still it is not transcendental,but moral, and consequently cannot in itselfform the subject of our criticism.

The third question: If I act as I ought to do,what may I then hope?—is at once practicaland theoretical. The practical forms a clue tothe answer of the theoretical, and—in its hig-hest form- speculative question. For all hoping

has happiness for its object and stands in preci-sely the same relation to the practical and thelaw of morality as knowing to the theoreticalcognition of things and the law of nature. Theformer arrives finally at the conclusion thatsomething is (which determines the ultimateend), because something ought to take place;the latter, that something is (which operates asthe highest cause), because something doestake place.

Happiness is the satisfaction of all our desires;extensive, in regard to their multiplicity; inten-sive, in regard to their degree; and protensive,in regard to their duration. The practical lawbased on the motive of happiness I term apragmatical law (or prudential rule); but thatlaw, assuming such to exist, which has no othermotive than the worthiness of being happy, Iterm a moral or ethical law. The first tells uswhat we have to do, if we wish to become pos-sessed of happiness; the second dictates how

we ought to act, in order to deserve happiness.The first is based upon empirical principles; forit is only by experience that I can learn eitherwhat inclinations exist which desire satisfac-tion, or what are the natural means of satisfyingthem. The second takes no account of our desi-res or the means of satisfying them, and re-gards only the freedom of a rational being, andthe necessary conditions under which alonethis freedom can harmonize with the distribu-tion of happiness according to principles. Thissecond law may therefore rest upon mere ideasof pure reason, and may be cognized a priori.

I assume that there are pure moral laws whichdetermine, entirely a priori (without regard toempirical motives, that is, to happiness), theconduct of a rational being, or in other words,to use which it makes of its freedom, and thatthese laws are absolutely imperative (not mere-ly hypothetically, on the supposition of otherempirical ends), and therefore in all respects

necessary. I am warranted in assuming this, notonly by the arguments of the most enlightenedmoralists, but by the moral judgement of everyman who will make the attempt to form a dis-tinct conception of such a law.

Pure reason, then, contains, not indeed in itsspeculative, but in its practical, or, more stric-tly, its moral use, principles of the possibility ofexperience, of such actions, namely, as, in ac-cordance with ethical precepts, might be metwith in the history of man. For since reasoncommands that such actions should take place,it must be possible for them to take place; andhence a particular kind of systematic unity—the moral—must be possible. We have found, itis true, that the systematic unity of nature couldnot be established according to speculativeprinciples of reason, because, while reason pos-sesses a causal power in relation to freedom, ithas none in relation to the whole sphere of na-ture; and, while moral principles of reason can

produce free actions, they cannot produce na-tural laws. It is, then, in its practical, but espe-cially in its moral use, that the principles ofpure reason possess objective reality.

I call the world a moral world, in so far as itmay be in accordance with all the ethicallaws—which, by virtue of the freedom of rea-sonable beings, it can be, and according to thenecessary laws of morality it ought to be. Butthis world must be conceived only as an intelli-gible world, inasmuch as abstraction is thereinmade of all conditions (ends), and even of allimpediments to morality (the weakness or pra-vity of human nature). So far, then, it is a mereidea- though still a practical idea—which mayhave, and ought to have, an influence on theworld of sense, so as to bring it as far as possi-ble into conformity with itself. The idea of amoral world has, therefore, objective reality,not as referring to an object of intelligible intui-tion—for of such an object we can form no con-

ception whatever—but to the world of sense—conceived, however, as an object of pure reasonin its practical use—and to a corpus mysticumof rational beings in it, in so far as the liberumarbitrium of the individual is placed, under andby virtue of moral laws, in complete systematicunity both with itself and with the freedom ofall others.

That is the answer to the first of the two ques-tions of pure reason which relate to its practicalinterest: Do that which will render thee worthyof happiness. The second question is this: If Iconduct myself so as not to be unworthy ofhappiness, may I hope thereby to obtain happi-ness? In order to arrive at the solution of thisquestion, we must inquire whether the princi-ples of pure reason, which prescribe a priori thelaw, necessarily also connect this hope with it.

I say, then, that just as the moral principles arenecessary according to reason in its practicaluse, so it is equally necessary according to rea-

son in its theoretical use to assume that everyone has ground to hope for happiness in themeasure in which he has made himself worthyof it in his conduct, and that therefore the sys-tem of morality is inseparably (though only inthe idea of pure reason) connected with that ofhappiness.

Now in an intelligible, that is, in the moralworld, in the conception of which we makeabstraction of all the impediments to morality(sensuous desires), such a system of happiness,connected with and proportioned to morality,may be conceived as necessary, because free-dom of volition—partly incited, and partly res-trained by moral laws—would be itself the cau-se of general happiness; and thus rationalbeings, under the guidance of such principles,would be themselves the authors both of theirown enduring welfare and that of others. Butsuch a system of self-rewarding morality isonly an idea, the carrying out of which depends

upon the condition that every one acts as heought; in other words, that all actions of reaso-nable beings be such as they would be if theysprung from a Supreme Will, comprehendingin, or under, itself all particular wills. But sincethe moral law is binding on each individual inthe use of his freedom of volition, even if othersshould not act in conformity with this law,neither the nature of things, nor the causality ofactions and their relation to morality, determi-ne how the consequences of these actions willbe related to happiness; and the necessary con-nection of the hope of happiness with the un-ceasing endeavour to become worthy of happi-ness, cannot be cognized by reason, if we takenature alone for our guide. This connection canbe hoped for only on the assumption that thecause of nature is a supreme reason, which go-verns according to moral laws.

I term the idea of an intelligence in which themorally most perfect will, united with supreme

blessedness, is the cause of all happiness in theworld, so far as happiness stands in strict rela-tion to morality (as the worthiness of beinghappy), the ideal of the supreme Good. It isonly, then, in the ideal of the supreme originalgood, that pure reason can find the ground ofthe practically necessary connection of bothelements of the highest derivative good, andaccordingly of an intelligible, that is, moralworld. Now since we are necessitated by rea-son to conceive ourselves as belonging to sucha world, while the senses present to us nothingbut a world of phenomena, we must assumethe former as a consequence of our conduct inthe world of sense (since the world of sensegives us no hint of it), and therefore as future inrelation to us. Thus God and a future life aretwo hypotheses which, according to the princi-ples of pure reason, are inseparable from theobligation which this reason imposes upon us.

Morality per se constitutes a system. But wecan form no system of happiness, except in sofar as it is dispensed in strict proportion to mo-rality. But this is only possible in the intelligibleworld, under a wise author and ruler. Such aruler, together with life in such a world, whichwe must look upon as future, reason finds itselfcompelled to assume; or it must regard the mo-ral laws as idle dreams, since the necessaryconsequence which this same reason connectswith them must, without this hypothesis, fall tothe ground. Hence also the moral laws are uni-versally regarded as commands, which theycould not be did they not connect a priori ade-quate consequences with their dictates, andthus carry with them promises and threats. Butthis, again, they could not do, did they not re-side in a necessary being, as the Supreme Good,which alone can render such a teleological uni-ty possible.

Leibnitz termed the world, when viewed inrelation to the rational beings which it contains,and the moral relations in which they stand toeach other, under the government of the Su-preme Good, the kingdom of Grace, and dis-tinguished it from the kingdom of Nature, inwhich these rational beings live, under morallaws, indeed, but expect no other consequencesfrom their actions than such as follow accor-ding to the course of nature in the world ofsense. To view ourselves, therefore, as in thekingdom of grace, in which all happinessawaits us, except in so far as we ourselves limitour participation in it by actions which renderus unworthy of happiness, is a practically ne-cessary idea of reason.

Practical laws, in so far as they are subjectivegrounds of actions, that is, subjective princi-ples, are termed maxims. The judgements ofmoral according to in its purity and ultimateresults are framed according ideas; the obser-

vance of its laws, according to according tomaxims.

The whole course of our life must be subject tomoral maxims; but this is impossible, unlesswith the moral law, which is a mere idea, rea-son connects an efficient cause which ordains toall conduct which is in conformity with themoral law an issue either in this or in anotherlife, which is in exact conformity with our hig-hest aims. Thus, without a God and without aworld, invisible to us now, but hoped for, theglorious ideas of morality are, indeed, objects ofapprobation and of admiration, but cannot bethe springs of purpose and action. For they donot satisfy all the aims which are natural toevery rational being, and which are determineda priori by pure reason itself, and necessary.

Happiness alone is, in the view of reason, farfrom being the complete good. Reason does notapprove of it (however much inclination maydesire it), except as united with desert. On the

other hand, morality alone, and with it, meredesert, is likewise far from being the completegood. To make it complete, he who conductshimself in a manner not unworthy of happi-ness, must be able to hope for the possession ofhappiness. Even reason, unbiased by privateends, or interested considerations, cannot judgeotherwise, if it puts itself in the place of a beingwhose business it is to dispense all happiness toothers. For in the practical idea both points areessentially combined, though in such a waythat participation in happiness is rendered pos-sible by the moral disposition, as its condition,and not conversely, the moral disposition bythe prospect of happiness. For a dispositionwhich should require the prospect of happinessas its necessary condition would not be moral,and hence also would not be worthy of comple-te happiness—a happiness which, in the viewof reason, recognizes no limitation but such asarises from our own immoral conduct.

Happiness, therefore, in exact proportion withthe morality of rational beings (whereby theyare made worthy of happiness), constitutesalone the supreme good of a world into whichwe absolutely must transport ourselves accor-ding to the commands of pure but practicalreason. This world is, it is true, only an intelli-gible world; for of such a systematic unity ofends as it requires, the world of sense gives usno hint. Its reality can be based on nothing elsebut the hypothesis of a supreme original good.In it independent reason, equipped with all thesufficiency of a supreme cause, founds, main-tains, and fulfils the universal order of things,with the most perfect teleological harmony,however much this order may be hidden fromus in the world of sense.

This moral theology has the peculiar advanta-ge, in contrast with speculative theology, ofleading inevitably to the conception of a sole,perfect, and rational First Cause, whereof spe-

culative theology does not give us any indica-tion on objective grounds, far less any convin-cing evidence. For we find neither in transcen-dental nor in natural theology, however farreason may lead us in these, any ground to wa-rrant us in assuming the existence of one onlyBeing, which stands at the head of all naturalcauses, and on which these are entirely depen-dent. On the other band, if we take our standon moral unity as a necessary law of the uni-verse, and from this point of view considerwhat is necessary to give this law adequateefficiency and, for us, obligatory force, we mustcome to the conclusion that there is one onlysupreme will, which comprehends all theselaws in itself. For how, under different wills,should we find complete unity of ends? Thiswill must be omnipotent, that all nature and itsrelation to morality in the world may be subjectto it; omniscient, that it may have knowledge ofthe most secret feelings and their moral worth;omnipresent, that it may be at hand to supply

every necessity to which the highest weal of theworld may give rise; eternal, that this harmonyof nature and liberty may never fail; and so on.

But this systematic unity of ends in this worldof intelligences- which, as mere nature, is onlya world of sense, but, as a system of freedom ofvolition, may be termed an intelligible, that is,moral world (regnum gratiae)—leads inevita-bly also to the teleological unity of all thingswhich constitute this great whole, according touniversal natural laws—just as the unity of theformer is according to universal and necessarymoral laws—and unites the practical with thespeculative reason. The world must be repre-sented as having originated from an idea, if it isto harmonize with that use of reason withoutwhich we cannot even consider ourselves asworthy of reason- namely, the moral use, whichrests entirely on the idea of the supreme good.Hence the investigation of nature receives ateleological direction, and becomes, in its wi-

dest extension, physico-theology. But this, ta-king its rise in moral order as a unity foundedon the essence of freedom, and not accidentallyinstituted by external commands, establishesthe teleological view of nature on groundswhich must be inseparably connected with theinternal possibility of things. This gives rise to atranscendental theology, which takes the idealof the highest ontological perfection as a prin-ciple of systematic unity; and this principleconnects all things according to universal andnecessary natural laws, because all things havetheir origin in the absolute necessity of the oneonly Primal Being.

What use can we make of our understanding,even in respect of experience, if we do not pro-pose ends to ourselves? But the highest endsare those of morality, and it is only pure reasonthat can give us the knowledge of these.Though supplied with these, and putting our-selves under their guidance, we can make no

teleological use of the knowledge of nature, asregards cognition, unless nature itself has esta-blished teleological unity. For without this uni-ty we should not even possess reason, becausewe should have no school for reason, and nocultivation through objects which afford thematerials for its conceptions. But teleologicalunity is a necessary unity, and founded on theessence of the individual will itself. Hence thiswill, which is the condition of the application ofthis unity in concreto, must be so likewise. Inthis way the transcendental enlargement of ourrational cognition would be, not the cause, butmerely the effect of the practical teleologywhich pure reason imposes upon us.

Hence, also, we find in the history of humanreason that, before the moral conceptions weresufficiently purified and determined, and befo-re men had attained to a perception of the sys-tematic unity of ends according to these con-ceptions and from necessary principles, the

knowledge of nature, and even a considerableamount of intellectual culture in many othersciences, could produce only rude and vagueconceptions of the Deity, sometimes even ad-mitting of an astonishing indifference with re-gard to this question altogether. But the moreenlarged treatment of moral ideas, which wasrendered necessary by the extreme pure morallaw of our religion, awakened the interest, andthereby quickened the perceptions of reason inrelation to this object. In this way, and withoutthe help either of an extended acquaintancewith nature, or of a reliable transcendental in-sight (for these have been wanting in all ages),a conception of the Divine Being was arrived at,which we now bold to be the correct one, notbecause speculative reason convinces us of itscorrectness, but because it accords with themoral principles of reason. Thus it is to purereason, but only in its practical use, that wemust ascribe the merit of having connectedwith our highest interest a cognition, of which

mere speculation was able only to form a con-jecture, but the validity of which it was unableto establish—and of having thereby rendered it,not indeed a demonstrated dogma, but a hy-pothesis absolutely necessary to the essentialends of reason.

But if practical reason has reached this eleva-tion, and has attained to the conception of asole Primal Being as the supreme good, it mustnot, therefore, imagine that it has transcendedthe empirical conditions of its application, andrisen to the immediate cognition of new objects;it must not presume to start from the concep-tion which it has gained, and to deduce from itthe moral laws themselves. For it was thesevery laws, the internal practical necessity ofwhich led us to the hypothesis of an indepen-dent cause, or of a wise ruler of the universe,who should give them effect. Hence we are notentitled to regard them as accidental and deri-ved from the mere will of the ruler, especially

as we have no conception of such a will, exceptas formed in accordance with these laws. So far,then, as practical reason has the right to con-duct us, we shall not look upon actions as bin-ding on us, because they are the commands ofGod, but we shall regard them as divine com-mands, because we are internally bound bythem. We shall study freedom under the teleo-logical unity which accords with principles ofreason; we shall look upon ourselves as actingin conformity with the divine will only in so faras we hold sacred the moral law which reasonteaches us from the nature of actions themsel-ves, and we shall believe that we can obey thatwill only by promoting the weal of the universein ourselves and in others. Moral theology is,therefore, only of immanent use. It teaches usto fulfil our destiny here in the world, by pla-cing ourselves in harmony with the generalsystem of ends, and warns us against the fana-ticism, nay, the crime of depriving reason of itslegislative authority in the moral conduct of

life, for the purpose of directly connecting thisauthority with the idea of the Supreme Being.For this would be, not an immanent, but atranscendent use of moral theology, and, likethe transcendent use of mere speculation,would inevitably pervert and frustrate the ul-timate ends of reason.

SECTION III. Of Opinion, Knowledge,and Belief.

The holding of a thing to be true is a phenome-non in our understanding which may rest onobjective grounds, but requires, also, subjectivecauses in the mind of the person judging. If ajudgement is valid for every rational being,then its ground is objectively sufficient, and it istermed a conviction. If, on the other hand, ithas its ground in the particular character of thesubject, it is termed a persuasion.

Persuasion is a mere illusion, the ground of thejudgement, which lies solely in the subject,being regarded as objective. Hence a judgementof this kind has only private validity—is onlyvalid for the individual who judges, and theholding of a thing to be true in this way cannotbe communicated. But truth depends uponagreement with the object, and consequentlythe judgements of all understandings, if true,must be in agreement with each other (consen-tientia uni tertio consentiunt inter se). Convic-tion may, therefore, be distinguished, from anexternal point of view, from persuasion, by thepossibility of communicating it and by showingits validity for the reason of every man; for inthis case the presumption, at least, arises thatthe agreement of all judgements with each ot-her, in spite of the different characters of indi-viduals, rests upon the common ground of theagreement of each with the object, and thus thecorrectness of the judgement is established.

Persuasion, accordingly, cannot be subjectivelydistinguished from conviction, that is, so longas the subject views its judgement simply as aphenomenon of its own mind. But if we inquirewhether the grounds of our judgement, whichare valid for us, produce the same effect on thereason of others as on our own, we have thenthe means, though only subjective means, not,indeed, of producing conviction, but of detec-ting the merely private validity of the judge-ment; in other words, of discovering that thereis in it the element of mere persuasion.

If we can, in addition to this, develop the sub-jective causes of the judgement, which we havetaken for its objective grounds, and thus ex-plain the deceptive judgement as a phenome-non in our mind, apart altogether from the ob-jective character of the object, we can then ex-pose the illusion and need be no longer decei-ved by it, although, if its subjective cause lies in

our nature, we cannot hope altogether to esca-pe its influence.

I can only maintain, that is, affirm as necessari-ly valid for every one, that which producesconviction. Persuasion I may keep for myself, ifit is agreeable to me; but I cannot, and oughtnot, to attempt to impose it as binding uponothers.

Holding for true, or the subjective validity of ajudgement in relation to conviction (which is, atthe same time, objectively valid), has the threefollowing degrees: opinion, belief, and know-ledge. Opinion is a consciously insufficient jud-gement, subjectively as well as objectively. Be-lief is subjectively sufficient, but is recognizedas being objectively insufficient. Knowledge isboth subjectively and objectively sufficient.Subjective sufficiency is termed conviction (formyself); objective sufficiency is termed certain-ty (for all). I need not dwell longer on the ex-planation of such simple conceptions.

I must never venture to be of opinion, withoutknowing something, at least, by which my jud-gement, in itself merely problematical, isbrought into connection with the truth—whichconnection, although not perfect, is still somet-hing more than an arbitrary fiction. Moreover,the law of such a connection must be certain.For if, in relation to this law, I have nothingmore than opinion, my judgement is but a playof the imagination, without the least relation totruth. In the judgements of pure reason, opi-nion has no place. For, as they do not rest onempirical grounds and as the sphere of purereason is that of necessary truth and a prioricognition, the principle of connection in it re-quires universality and necessity, and conse-quently perfect certainty—otherwise we shouldhave no guide to the truth at all. Hence it isabsurd to have an opinion in pure mathema-tics; we must know, or abstain from forming ajudgement altogether. The case is the same withthe maxims of morality. For we must not

hazard an action on the mere opinion that it isallowed, but we must know it to be so. In thetranscendental sphere of reason, on the otherhand, the term opinion is too weak, while theword knowledge is too strong. From the merelyspeculative point of view, therefore, we cannotform a judgement at all. For the subjectivegrounds of a judgement, such as produce belief,cannot be admitted in speculative inquiries,inasmuch as they cannot stand without empiri-cal support and are incapable of being commu-nicated to others in equal measure.

But it is only from the practical point of viewthat a theoretically insufficient judgement canbe termed belief. Now the practical reference iseither to skill or to morality; to the former,when the end proposed is arbitrary and acci-dental, to the latter, when it is absolutely neces-sary.

If we propose to ourselves any end whatever,the conditions of its attainment are hypothetica-

lly necessary. The necessity is subjectively, butstill only comparatively, sufficient, if I am ac-quainted with no other conditions under whichthe end can be attained. On the other hand, it issufficient, absolutely and for every one, if Iknow for certain that no one can be acquaintedwith any other conditions under which the at-tainment of the proposed end would be possi-ble. In the former case my supposition—myjudgement with regard to certain conditions—is a merely accidental belief; in the latter it is anecessary belief. The physician must pursuesome course in the case of a patient who is indanger, but is ignorant of the nature of the di-sease. He observes the symptoms, and conclu-des, according to the best of his judgement, thatit is a case of phthisis. His belief is, even in hisown judgement, only contingent: another manmight, perhaps come nearer the truth. Such abelief, contingent indeed, but still forming theground of the actual use of means for the at-

tainment of certain ends, I term Pragmaticalbelief.

The usual test, whether that which any onemaintains is merely his persuasion, or his sub-jective conviction at least, that is, his firm belief,is a bet. It frequently happens that a man deli-vers his opinions with so much boldness andassurance, that he appears to be under no ap-prehension as to the possibility of his being inerror. The offer of a bet startles him, and makeshim pause. Sometimes it turns out that his per-suasion may be valued at a ducat, but not atten. For he does not hesitate, perhaps, to ventu-re a ducat, but if it is proposed to stake ten, heimmediately becomes aware of the possibilityof his being mistaken—a possibility which hashitherto escaped his observation. If we imagineto ourselves that we have to stake the happi-ness of our whole life on the truth of any pro-position, our judgement drops its air oftriumph, we take the alarm, and discover the

actual strength of our belief. Thus pragmaticalbelief has degrees, varying in proportion to theinterests at stake.

Now, in cases where we cannot enter upon anycourse of action in reference to some object, andwhere, accordingly, our judgement is purelytheoretical, we can still represent to ourselves,in thought, the possibility of a course of action,for which we suppose that we have sufficientgrounds, if any means existed of ascertainingthe truth of the matter. Thus we find in purelytheoretical judgements an analogon of practicaljudgements, to which the word belief may pro-perly be applied, and which we may term doc-trinal belief. I should not hesitate to stake myall on the truth of the proposition- if there wereany possibility of bringing it to the test of expe-rience—that, at least, some one of the planets,which we see, is inhabited. Hence I say that Ihave not merely the opinion, but the strongbelief, on the correctness of which I would sta-

ke even many of the advantages of life, thatthere are inhabitants in other worlds.

Now we must admit that the doctrine of theexistence of God belongs to doctrinal belief.For, although in respect to the theoretical cogni-tion of the universe I do not require to form anytheory which necessarily involves this idea, asthe condition of my explanation of the pheno-mena which the universe presents, but, on thecontrary, am rather bound so to use my reasonas if everything were mere nature, still teleolo-gical unity is so important a condition of theapplication of my reason to nature, that it isimpossible for me to ignore it—especially since,in addition to these considerations, abundantexamples of it are supplied by experience. Butthe sole condition, so far as my knowledge ex-tends, under which this unity can be my guidein the investigation of nature, is the assumptionthat a supreme intelligence has ordered allthings according to the wisest ends. Conse-

quently, the hypothesis of a wise author of theuniverse is necessary for my guidance in theinvestigation of nature—is the condition underwhich alone I can fulfil an end which is contin-gent indeed, but by no means unimportant.Moreover, since the result of my attempts sofrequently confirms the utility of this assump-tion, and since nothing decisive can be adducedagainst it, it follows that it would be saying fartoo little to term my judgement, in this case, amere opinion, and that, even in this theoreticalconnection, I may assert that I firmly believe inGod. Still, if we use words strictly, this mustnot be called a practical, but a doctrinal belief,which the theology of nature (physico-theology) must also produce in my mind. In thewisdom of a Supreme Being, and in the short-ness of life, so inadequate to the developmentof the glorious powers of human nature, wemay find equally sufficient grounds for a doc-trinal belief in the future life of the human soul.

The expression of belief is, in such cases, anexpression of modesty from the objective pointof view, but, at the same time, of firm confiden-ce, from the subjective. If I should venture toterm this merely theoretical judgement even somuch as a hypothesis which I am entitled toassume; a more complete conception, with re-gard to another world and to the cause of theworld, might then be justly required of me thanI am, in reality, able to give. For, if I assumeanything, even as a mere hypothesis, I must, atleast, know so much of the properties of such abeing as will enable me, not to form the concep-tion, but to imagine the existence of it. But theword belief refers only to the guidance whichan idea gives me, and to its subjective influenceon the conduct of my reason, which forces meto hold it fast, though I may not be in a positionto give a speculative account of it.

But mere doctrinal belief is, to some extent,wanting in stability. We often quit our hold of

it, in consequence of the difficulties which oc-cur in speculation, though in the end we inevi-tably return to it again.

It is quite otherwise with moral belief. For inthis sphere action is absolutely necessary, thatis, I must act in obedience to the moral law inall points. The end is here incontrovertibly es-tablished, and there is only one condition pos-sible, according to the best of my perception,under which this end can harmonize with allother ends, and so have practical validity—namely, the existence of a God and of a futureworld. I know also, to a certainty, that no onecan be acquainted with any other conditionswhich conduct to the same unity of ends underthe moral law. But since the moral precept is, atthe same time, my maxim (as reason requiresthat it should be), I am irresistibly constrainedto believe in the existence of God and in a futu-re life; and I am sure that nothing can make mewaver in this belief, since I should thereby

overthrow my moral maxims, the renunciationof which would render me hateful in my owneyes.

Thus, while all the ambitious attempts of rea-son to penetrate beyond the limits of experienceend in disappointment, there is still enough leftto satisfy us in a practical point of view. Noone, it is true, will be able to boast that heknows that there is a God and a future life; for,if he knows this, he is just the man whom Ihave long wished to find. All knowledge, re-garding an object of mere reason, can be com-municated; and I should thus be enabled tohope that my own knowledge would receivethis wonderful extension, through the instru-mentality of his instruction. No, my convictionis not logical, but moral certainty; and since itrests on subjective grounds (of the moral sen-timent), I must not even say: It is morally cer-tain that there is a God, etc., but: I am morallycertain, that is, my belief in God and in another

world is so interwoven with my moral naturethat I am under as little apprehension of havingthe former torn from me as of losing the latter.

The only point in this argument that may ap-pear open to suspicion is that this rational be-lief presupposes the existence of moral senti-ments. If we give up this assumption, and takea man who is entirely indifferent with regard tomoral laws, the question which reason propo-ses, becomes then merely a problem for specu-lation and may, indeed, be supported by stronggrounds from analogy, but not by such as willcompel the most obstinate scepticism to giveway.* But in these questions no man is freefrom all interest. For though the want of goodsentiments may place him beyond the influenceof moral interests, still even in this case enoughmay be left to make him fear the existence ofGod and a future life. For he cannot pretend toany certainty of the non-existence of God andof a future life, unless- since it could only be

proved by mere reason, and therefore apodeic-tically—he is prepared to establish the impossi-bility of both, which certainly no reasonableman would undertake to do. This would be anegative belief, which could not, indeed, pro-duce morality and good sentiments, but stillcould produce an analogon of these, by opera-ting as a powerful restraint on the outbreak ofevil dispositions.

[*Footnote: The human mind (as, I believe, eve-ry rational being must of necessity do) takes anatural interest in morality, although this inter-est is not undivided, and may not be practicallyin preponderance. If you strengthen and in-crease it, you will find the reason becomedocile, more enlightened, and more capable ofuniting the speculative interest with thepractical. But if you do not take care at theoutset, or at least midway, to make men good,you will never force them into an honest belief.]

But, it will be said, is this all that pure reasoncan effect, in opening up prospects beyond thelimits of experience? Nothing more than twoarticles of belief? Common sense could havedone as much as this, without taking the philo-sophers to counsel in the matter!

I shall not here eulogize philosophy for the be-nefits which the laborious efforts of its criticismhave conferred on human reason- even gran-ting that its merit should turn out in the end tobe only negative—for on this point somethingmore will be said in the next section. But, I ask,do you require that that knowledge which con-cerns all men, should transcend the commonunderstanding, and should only be revealed toyou by philosophers? The very circumstancewhich has called forth your censure, is the bestconfirmation of the correctness of our previousassertions, since it discloses, what could nothave been foreseen, that Nature is not chargea-ble with any partial distribution of her gifts in

those matters which concern all men withoutdistinction and that, in respect to the essentialends of human nature, we cannot advancefurther with the help of the highest philosophy,than under the guidance which nature hasvouchsafed to the meanest understanding.

CHAPTER III. The Architectonic of PureReason.

By the term architectonic I mean the art of cons-tructing a system. Without systematic unity,our knowledge cannot become science; it willbe an aggregate, and not a system. Thus archi-tectonic is the doctrine of the scientific in cogni-tion, and therefore necessarily forms part of ourmethodology.

Reason cannot permit our knowledge to remainin an unconnected and rhapsodistic state, but

requires that the sum of our cognitions shouldconstitute a system. It is thus alone that theycan advance the ends of reason. By a system Imean the unity of various cognitions under oneidea. This idea is the conception—given by rea-son—of the form of a whole, in so far as theconception determines a priori not only thelimits of its content, but the place which each ofits parts is to occupy. The scientific idea con-tains, therefore, the end and the form of thewhole which is in accordance with that end.The unity of the end, to which all the parts ofthe system relate, and through which all have arelation to each other, communicates unity tothe whole system, so that the absence of anypart can be immediately detected from ourknowledge of the rest; and it determines a prio-ri the limits of the system, thus excluding allcontingent or arbitrary additions. The whole isthus an organism (articulatio), and not an ag-gregate (coacervatio); it may grow from within(per intussusceptionem), but it cannot increase

by external additions (per appositionem). It is,thus, like an animal body, the growth of whichdoes not add any limb, but, without changingtheir proportions, makes each in its spherestronger and more active.

We require, for the execution of the idea of asystem, a schema, that is, a content and anarrangement of parts determined a priori bythe principle which the aim of the system pres-cribes. A schema which is not projected in ac-cordance with an idea, that is, from the stand-point of the highest aim of reason, but merelyempirically, in accordance with accidental aimsand purposes (the number of which cannot bepredetermined), can give us nothing more thantechnical unity. But the schema which is origi-nated from an idea (in which case reason pre-sents us with aims a priori, and does not lookfor them to experience), forms the basis of ar-chitectonical unity. A science, in the properacceptation of that term, cannot be formed

technically, that is, from observation of the si-milarity existing between different objects, andthe purely contingent use we make of ourknowledge in concreto with reference to allkinds of arbitrary external aims; its constitutionmust be framed on architectonical principles,that is, its parts must be shown to possess anessential affinity, and be capable of being de-duced from one supreme and internal aim orend, which forms the condition of the possibili-ty of the scientific whole. The schema of ascience must give a priori the plan of it (mono-gramma), and the division of the whole intoparts, in conformity with the idea of the scien-ce; and it must also distinguish this whole fromall others, according to certain understoodprinciples.

No one will attempt to construct a science, un-less he have some idea to rest on as a properbasis. But, in the elaboration of the science, hefinds that the schema, nay, even the definition

which he at first gave of the science, rarely co-rresponds with his idea; for this idea lies, like agerm, in our reason, its parts undeveloped andhid even from microscopical observation. Forthis reason, we ought to explain and definesciences, not according to the description whichthe originator gives of them, but according tothe idea which we find based in reason itself,and which is suggested by the natural unity ofthe parts of the science already accumulated.For it will of ten be found that the originator ofa science and even his latest successors remainattached to an erroneous idea, which they can-not render clear to themselves, and that theythus fail in determining the true content, thearticulation or systematic unity, and the limitsof their science.

It is unfortunate that, only after having occu-pied ourselves for a long time in the collectionof materials, under the guidance of an ideawhich lies undeveloped in the mind, but not

according to any definite plan of arrange-ment—nay, only after we have spent muchtime and labour in the technical disposition ofour materials, does it become possible to viewthe idea of a science in a clear light, and to pro-ject, according to architectonical principles, aplan of the whole, in accordance with the aimsof reason. Systems seem, like certain worms, tobe formed by a kind of generatio aequivoca—by the mere confluence of conceptions, and togain completeness only with the progress oftime. But the schema or germ of all lies in rea-son; and thus is not only every system organi-zed according to its own idea, but all are unitedinto one grand system of human knowledge, ofwhich they form members. For this reason, it ispossible to frame an architectonic of all humancognition, the formation of which, at the pre-sent time, considering the immense materialscollected or to be found in the ruins of old sys-tems, would not indeed be very difficult. Ourpurpose at present is merely to sketch the plan

of the architectonic of all cognition given bypure reason; and we begin from the point whe-re the main root of human knowledge dividesinto two, one of which is reason. By reason Iunderstand here the whole higher faculty ofcognition, the rational being placed in contra-distinction to the empirical.

If I make complete abstraction of the content ofcognition, objectively considered, all cognitionis, from a subjective point of view, either histo-rical or rational. Historical cognition is cognitioex datis, rational, cognitio ex principiis. Whate-ver may be the original source of a cognition, itis, in relation to the person who possesses it,merely historical, if he knows only what hasbeen given him from another quarter, whetherthat knowledge was communicated by directexperience or by instruction. Thus the Personwho has learned a system of philosophy—saythe Wolfian—although he has a perfect know-ledge of all the principles, definitions, and ar-

guments in that philosophy, as well as of thedivisions that have been made of the system,possesses really no more than an historicalknowledge of the Wolfian system; he knowsonly what has been told him, his judgementsare only those which he has received from histeachers. Dispute the validity of a definition,and he is completely at a loss to find another.He has formed his mind on another's; but theimitative faculty is not the productive. Hisknowledge has not been drawn from reason;and although, objectively considered, it is ra-tional knowledge, subjectively, it is merely his-torical. He has learned this or that philosophyand is merely a plaster cast of a living man.Rational cognitions which are objective, that is,which have their source in reason, can be sotermed from a subjective point of view, onlywhen they have been drawn by the individualhimself from the sources of reason, that is, fromprinciples; and it is in this way alone that criti-

cism, or even the rejection of what has beenalready learned, can spring up in the mind.

All rational cognition is, again, based either onconceptions, or on the construction of concep-tions. The former is termed philosophical, thelatter mathematical. I have already shown theessential difference of these two methods ofcognition in the first chapter. A cognition maybe objectively philosophical and subjectivelyhistorical—as is the case with the majority ofscholars and those who cannot look beyond thelimits of their system, and who remain in a sta-te of pupilage all their lives. But it is remarka-ble that mathematical knowledge, when com-mitted to memory, is valid, from the subjectivepoint of view, as rational knowledge also, andthat the same distinction cannot be drawn hereas in the case of philosophical cognition. Thereason is that the only way of arriving at thisknowledge is through the essential principlesof reason, and thus it is always certain and in-

disputable; because reason is employed in con-creto—but at the same time a priori—that is, inpure and, therefore, infallible intuition; andthus all causes of illusion and error are exclu-ded. Of all the a priori sciences of reason, there-fore, mathematics alone can be learned. Philo-sophy—unless it be in an historical manner—cannot be learned; we can at most learn to phi-losophize.

Philosophy is the system of all philosophicalcognition. We must use this term in an objecti-ve sense, if we understand by it the archetypeof all attempts at philosophizing, and the stan-dard by which all subjective philosophies are tobe judged. In this sense, philosophy is merelythe idea of a possible science, which does notexist in concreto, but to which we endeavour invarious ways to approximate, until we havediscovered the right path to pursue—a pathovergrown by the errors and illusions of sen-se—and the image we have hitherto tried in

vain to shape has become a perfect copy of thegreat prototype. Until that time, we cannotlearn philosophy—it does not exist; if it does,where is it, who possesses it, and how shall weknow it? We can only learn to philosophize; inother words, we can only exercise our powersof reasoning in accordance with general princi-ples, retaining at the same time, the right ofinvestigating the sources of these principles, oftesting, and even of rejecting them.

Until then, our conception of philosophy is on-ly a scholastic conception—a conception, thatis, of a system of cognition which we are tryingto elaborate into a science; all that we at presentknow being the systematic unity of this cogni-tion, and consequently the logical completenessof the cognition for the desired end. But there isalso a cosmical conception (conceptus cosmi-cus) of philosophy, which has always formedthe true basis of this term, especially when phi-losophy was personified and presented to us in

the ideal of a philosopher. In this view philo-sophy is the science of the relation of all cogni-tion to the ultimate and essential aims ofhuman reason (teleologia rationis humanae),and the philosopher is not merely an artist—who occupies himself with conceptions—but alawgiver, legislating for human reason. In thissense of the word, it would be in the highestdegree arrogant to assume the title of philosop-her, and to pretend that we had reached theperfection of the prototype which lies in theidea alone.

The mathematician, the natural philosopher,and the logician—how far soever the first mayhave advanced in rational, and the two latter inphilosophical knowledge—are merely artists,engaged in the arrangement and formation ofconceptions; they cannot be termed philosop-hers. Above them all, there is the ideal teacher,who employs them as instruments for the ad-vancement of the essential aims of human rea-

son. Him alone can we call philosopher; but henowhere exists. But the idea of his legislativepower resides in the mind of every man, and italone teaches us what kind of systematic unityphilosophy demands in view of the ultimateaims of reason. This idea is, therefore, a cosmi-cal conception.*

[*Footnote: By a cosmical conception, I meanone in which all men necessarily take an inter-est; the aim of a science must accordingly bedetermined according to scholastic conceptions,if it is regarded merely as a means to certainarbitrarily proposed ends.]

In view of the complete systematic unity ofreason, there can only be one ultimate end of allthe operations of the mind. To this all otheraims are subordinate, and nothing more thanmeans for its attainment. This ultimate end isthe destination of man, and the philosophywhich relates to it is termed moral philosophy.The superior position occupied by moral philo-

sophy, above all other spheres for the opera-tions of reason, sufficiently indicates the reasonwhy the ancients always included the idea—and in an especial manner—of moralist in thatof philosopher. Even at the present day, we calla man who appears to have the power of self-government, even although his knowledge maybe very limited, by the name of philosopher.

The legislation of human reason, or philosophy,has two objects- nature and freedom—and thuscontains not only the laws of nature, but alsothose of ethics, at first in two separate systems,which, finally, merge into one grand philosop-hical system of cognition. The philosophy ofnature relates to that which is, that of ethics tothat which ought to be.

But all philosophy is either cognition on thebasis of pure reason, or the cognition of reasonon the basis of empirical principles. The formeris termed pure, the latter empirical philosophy.

The philosophy of pure reason is either pro-paedeutic, that is, an inquiry into the powers ofreason in regard to pure a priori cognition, andis termed critical philosophy; or it is, secondly,the system of pure reason—a science contai-ning the systematic presentation of the wholebody of philosophical knowledge, true as wellas illusory, given by pure reason—and is calledmetaphysic. This name may, however, be alsogiven to the whole system of pure philosophy,critical philosophy included, and may designa-te the investigation into the sources or possibili-ty of a priori cognition, as well as the presenta-tion of the a priori cognitions which form asystem of pure philosophy—excluding, at thesame time, all empirical and mathematical ele-ments.

Metaphysic is divided into that of the specula-tive and that of the practical use of pure reason,and is, accordingly, either the metaphysic ofnature, or the metaphysic of ethics. The former

contains all the pure rational principles—basedupon conceptions alone (and thus excludingmathematics)—of all theoretical cognition; thelatter, the principles which determine and ne-cessitate a priori all action. Now moral philo-sophy alone contains a code of laws—for theregulation of our actions—which are deducedfrom principles entirely a priori. Hence the me-taphysic of ethics is the only pure moral philo-sophy, as it is not based upon anthropologicalor other empirical considerations. The metap-hysic of speculative reason is what is common-ly called metaphysic in the more limited sense.But as pure moral philosophy properly forms apart of this system of cognition, we must allowit to retain the name of metaphysic, although itis not requisite that we should insist on so ter-ming it in our present discussion.

It is of the highest importance to separate thosecognitions which differ from others both inkind and in origin, and to take great care that

they are not confounded with those with whichthey are generally found connected. What thechemist does in the analysis of substances,what the mathematician in pure mathematics,is, in a still higher degree, the duty of the philo-sopher, that the value of each different kind ofcognition, and the part it takes in the operationsof the mind, may be clearly defined. Humanreason has never wanted a metaphysic of somekind, since it attained the power of thought, orrather of reflection; but it has never been able tokeep this sphere of thought and cognition purefrom all admixture of foreign elements. Theidea of a science of this kind is as old as specu-lation itself; and what mind does not specula-te—either in the scholastic or in the popularfashion? At the same time, it must be admittedthat even thinkers by profession have beenunable clearly to explain the distinction bet-ween the two elements of our cognition—theone completely a priori, the other a posteriori;and hence the proper definition of a peculiar

kind of cognition, and with it the just idea of ascience which has so long and so deeply enga-ged the attention of the human mind, has neverbeen established. When it was said: "Metaphy-sic is the science of the first principles of humancognition," this definition did not signalize apeculiarity in kind, but only a difference in de-gree; these first principles were thus declared tobe more general than others, but no criterion ofdistinction from empirical principles was given.Of these some are more general, and thereforehigher, than others; and—as we cannot distin-guish what is completely a priori from thatwhich is known to be a posteriori—where shallwe draw the line which is to separate the hig-her and so-called first principles, from the lo-wer and subordinate principles of cognition?What would be said if we were asked to be sa-tisfied with a division of the epochs of theworld into the earlier centuries and those fo-llowing them? "Does the fifth, or the tenth cen-tury belong to the earlier centuries?" it would

be asked. In the same way I ask: Does the con-ception of extension belong to metaphysics?You answer, "Yes." Well, that of body too?"Yes." And that of a fluid body? You stop, youare unprepared to admit this; for if you do, eve-rything will belong to metaphysics. From this itis evident that the mere degree of subordina-tion—of the particular to the general—cannotdetermine the limits of a science; and that, inthe present case, we must expect to find a diffe-rence in the conceptions of metaphysics both inkind and in origin. The fundamental idea ofmetaphysics was obscured on another side bythe fact that this kind of a priori cognition sho-wed a certain similarity in character with thescience of mathematics. Both have the propertyin common of possessing an a priori origin; but,in the one, our knowledge is based upon con-ceptions, in the other, on the construction ofconceptions. Thus a decided dissimilarity bet-ween philosophical and mathematical cogni-tion comes out—a dissimilarity which was al-

ways felt, but which could not be made distinctfor want of an insight into the criteria of thedifference. And thus it happened that, as philo-sophers themselves failed in the proper deve-lopment of the idea of their science, the elabora-tion of the science could not proceed with adefinite aim, or under trustworthy guidance.Thus, too, philosophers, ignorant of the paththey ought to pursue and always disputingwith each other regarding the discoverieswhich each asserted he had made, broughttheir science into disrepute with the rest of theworld, and finally, even among themselves.

All pure a priori cognition forms, therefore, inview of the peculiar faculty which originates it,a peculiar and distinct unity; and metaphysic isthe term applied to the philosophy which at-tempts to represent that cognition in this sys-tematic unity. The speculative part of metaphy-sic, which has especially appropriated this ap-pellation—that which we have called the me-

taphysic of nature—and which considers eve-rything, as it is (not as it ought to be), by meansof a priori conceptions, is divided in the follo-wing manner.

Metaphysic, in the more limited acceptation ofthe term, consists of two parts—transcendentalphilosophy and the physiology of pure reason.The former presents the system of all the con-ceptions and principles belonging to the un-derstanding and the reason, and which relate toobjects in general, but not to any particular gi-ven objects (Ontologia); the latter has nature forits subject-matter, that is, the sum of given ob-jects—whether given to the senses, or, if wewill, to some other kind of intuition—and isaccordingly physiology, although only rationa-lis. But the use of the faculty of reason in thisrational mode of regarding nature is either phy-sical or hyperphysical, or, more properly spea-king, immanent or transcendent. The formerrelates to nature, in so far as our knowledge

regarding it may be applied in experience (inconcreto); the latter to that connection of theobjects of experience, which transcends all ex-perience. Transcendent physiology has, again,an internal and an external connection with itsobject, both, however, transcending possibleexperience; the former is the physiology of na-ture as a whole, or transcendental cognition ofthe world, the latter of the connection of thewhole of nature with a being above nature, ortranscendental cognition of God.

Immanent physiology, on the contrary, consi-ders nature as the sum of all sensuous objects,consequently, as it is presented to us—but stillaccording to a priori conditions, for it is underthese alone that nature can be presented to ourminds at all. The objects of immanent physiolo-gy are of two kinds: 1. Those of the externalsenses, or corporeal nature; 2. The object of theinternal sense, the soul, or, in accordance withour fundamental conceptions of it, thinking

nature. The metaphysics of corporeal nature iscalled physics; but, as it must contain only theprinciples of an a priori cognition of nature, wemust term it rational physics. The metaphysicsof thinking nature is called psychology, and forthe same reason is to be regarded as merely therational cognition of the soul.

Thus the whole system of metaphysics consistsof four principal parts: 1. Ontology; 2. RationalPhysiology; 3. Rational cosmology; and 4. Ra-tional theology. The second part—that of therational doctrine of nature—may be subdividedinto two, physica rationalis* and psychologiarationalis.

[*Footnote: It must not be supposed that I meanby this appellation what is generally calledphysica general is, and which is rather mat-hematics than a philosophy of nature. For themetaphysic of nature is completely differentfrom mathematics, nor is it so rich in results,although it is of great importance as a critical

test of the application of pure understanding-cognition to nature. For want of its guidance,even mathematicians, adopting certain com-mon notions- which are, in fact, metaphysical—have unconsciously crowded their theories ofnature with hypotheses, the fallacy of whichbecomes evident upon the application of theprinciples of this metaphysic, without detri-ment, however, to the employment of mat-hematics in this sphere of cognition.]

The fundamental idea of a philosophy of purereason of necessity dictates this division; it is,therefore, architectonical—in accordance withthe highest aims of reason, and not merely tech-nical, or according to certain accidentally-observed similarities existing between the dif-ferent parts of the whole science. For this rea-son, also, is the division immutable and of le-gislative authority. But the reader may observein it a few points to which he ought to demur,

and which may weaken his conviction of itstruth and legitimacy.

In the first place, how can I desire an a prioricognition or metaphysic of objects, in so far asthey are given a posteriori? and how is it possi-ble to cognize the nature of things according toa priori principles, and to attain to a rationalphysiology? The answer is this. We take fromexperience nothing more than is requisite topresent us with an object (in general) of theexternal or of the internal sense; in the formercase, by the mere conception of matter (impe-netrable and inanimate extension), in the latter,by the conception of a thinking being—given inthe internal empirical representation, I think.As to the rest, we must not employ in our me-taphysic of these objects any empirical princi-ples (which add to the content of our concep-tions by means of experience), for the purposeof forming by their help any judgements res-pecting these objects.

Secondly, what place shall we assign to empiri-cal psychology, which has always been consi-dered a part of metaphysics, and from which inour time such important philosophical resultshave been expected, after the hope of construc-ting an a priori system of knowledge had beenabandoned? I answer: It must be placed by theside of empirical physics or physics proper;that is, must be regarded as forming a part ofapplied philosophy, the a priori principles ofwhich are contained in pure philosophy, whichis therefore connected, although it must not beconfounded, with psychology. Empirical psy-chology must therefore be banished from thesphere of metaphysics, and is indeed excludedby the very idea of that science. In conformity,however, with scholastic usage, we must per-mit it to occupy a place in metaphysics—butonly as an appendix to it. We adopt this coursefrom motives of economy; as psychology is notas yet full enough to occupy our attention as anindependent study, while it is, at the same time,

of too great importance to be entirely excludedor placed where it has still less affinity than ithas with the subject of metaphysics. It is astranger who has been long a guest; and wemake it welcome to stay, until it can take up amore suitable abode in a complete system ofanthropology—the pendant to empirical phy-sics.

The above is the general idea of metaphysics,which, as more was expected from it than couldbe looked for with justice, and as these pleasantexpectations were unfortunately never realized,fell into general disrepute. Our Critique musthave fully convinced the reader that, althoughmetaphysics cannot form the foundation ofreligion, it must always be one of its most im-portant bulwarks, and that human reason,which naturally pursues a dialectical course,cannot do without this science, which checks itstendencies towards dialectic and, by elevatingreason to a scientific and clear self-knowledge,

prevents the ravages which a lawless speculati-ve reason would infallibly commit in the sphereof morals as well as in that of religion. We maybe sure, therefore, whatever contempt may bethrown upon metaphysics by those who judgea science not by its own nature, but accordingto the accidental effects it may have produced,that it can never be completely abandoned, thatwe must always return to it as to a beloved onewho has been for a time estranged, because thequestions with which it is engaged relate to thehighest aims of humanity, and reason mustalways labour either to attain to settled viewsin regard to these, or to destroy those whichothers have already established.

Metaphysic, therefore—that of nature, as wellas that of ethics, but in an especial manner thecriticism which forms the propaedeutic to allthe operations of reason—forms properly thatdepartment of knowledge which may be ter-med, in the truest sense of the word, philosop-

hy. The path which it pursues is that of science,which, when it has once been discovered, isnever lost, and never misleads. Mathematics,natural science, the common experience of men,have a high value as means, for the most part,to accidental ends—but at last also, to thosewhich are necessary and essential to the exis-tence of humanity. But to guide them to thishigh goal, they require the aid of rational cogni-tion on the basis of pure conceptions, which, beit termed as it may, is properly nothing butmetaphysics.

For the same reason, metaphysics forms likewi-se the completion of the culture of human rea-son. In this respect, it is indispensable, settingaside altogether the influence which it exerts asa science. For its subject-matter is the elementsand highest maxims of reason, which form thebasis of the possibility of some sciences and ofthe use of all. That, as a purely speculativescience, it is more useful in preventing error

than in the extension of knowledge, does notdetract from its value; on the contrary, the su-preme office of censor which it occupies assu-res to it the highest authority and importance.This office it administers for the purpose ofsecuring order, harmony, and well-being toscience, and of directing its noble and fruitfullabours to the highest possible aim—the happi-ness of all mankind.

CHAPTER IV. The History of Pure Reason.

This title is placed here merely for the purposeof designating a division of the system of purereason of which I do not intend to treat at pre-sent. I shall content myself with casting a cur-sory glance, from a purely transcendental pointof view—that of the nature of pure reason—onthe labours of philosophers up to the presenttime. They have aimed at erecting an edifice of

philosophy; but to my eye this edifice appearsto be in a very ruinous condition.

It is very remarkable, although naturally itcould not have been otherwise, that, in the in-fancy of philosophy, the study of the nature ofGod and the constitution of a future worldformed the commencement, rather than theconclusion, as we should have it, of the specu-lative efforts of the human mind. However ru-de the religious conceptions generated by theremains of the old manners and customs of aless cultivated time, the intelligent classes werenot thereby prevented from devoting themsel-ves to free inquiry into the existence and natureof God; and they easily saw that there could beno surer way of pleasing the invisible ruler ofthe world, and of attaining to happiness inanother world at least, than a good and honestcourse of life in this. Thus theology and moralsformed the two chief motives, or rather thepoints of attraction in all abstract inquiries. But

it was the former that especially occupied theattention of speculative reason, and which af-terwards became so celebrated under the nameof metaphysics.

I shall not at present indicate the periods oftime at which the greatest changes in metaphy-sics took place, but shall merely give a hastysketch of the different ideas which occasionedthe most important revolutions in this sphere ofthought. There are three different ends in rela-tion to which these revolutions have taken pla-ce.

1. In relation to the object of the cognition ofreason, philosophers may be divided into sen-sualists and intellectualists. Epicurus may beregarded as the head of the former, Plato of thelatter. The distinction here signalized, subtle asit is, dates from the earliest times, and was longmaintained. The former asserted that realityresides in sensuous objects alone, and that eve-rything else is merely imaginary; the latter, that

the senses are the parents of illusion and thattruth is to be found in the understanding alone.The former did not deny to the conceptions ofthe understanding a certain kind of reality; butwith them it was merely logical, with the othersit was mystical. The former admitted intellec-tual conceptions, but declared that sensuousobjects alone possessed real existence. The lat-ter maintained that all real objects were intelli-gible, and believed that the pure understandingpossessed a faculty of intuition apart from sen-se, which, in their opinion, served only to con-fuse the ideas of the understanding.

2. In relation to the origin of the pure cognitionsof reason, we find one school maintaining thatthey are derived entirely from experience, andanother that they have their origin in reasonalone. Aristotle may be regarded as the bead ofthe empiricists, and Plato of the noologists.Locke, the follower of Aristotle in modern ti-mes, and Leibnitz of Plato (although he cannot

be said to have imitated him in his mysticism),have not been able to bring this question to asettled conclusion. The procedure of Epicurusin his sensual system, in which he always re-stricted his conclusions to the sphere of experi-ence, was much more consequent than that ofAristotle and Locke. The latter especially, afterhaving derived all the conceptions and princi-ples of the mind from experience, goes so far, inthe employment of these conceptions and prin-ciples, as to maintain that we can prove theexistence of God and the existence of God andthe immortality of them objects lying beyondthe soul—both of them of possible experience—with the same force of demonstration as anymathematical proposition.

3. In relation to method. Method is procedureaccording to principles. We may divide themethods at present employed in the field ofinquiry into the naturalistic and the scientific.The naturalist of pure reason lays it down as

his principle that common reason, without theaid of science—which he calls sound reason, orcommon sense—can give a more satisfactoryanswer to the most important questions of me-taphysics than speculation is able to do. Hemust maintain, therefore, that we can deter-mine the content and circumference of themoon more certainly by the naked eye, than bythe aid of mathematical reasoning. But this sys-tem is mere misology reduced to principles;and, what is the most absurd thing in this doc-trine, the neglect of all scientific means is pa-raded as a peculiar method of extending ourcognition. As regards those who are naturalistsbecause they know no better, they are certainlynot to be blamed. They follow common sense,without parading their ignorance as a methodwhich is to teach us the wonderful secret, howwe are to find the truth which lies at the bottomof the well of Democritus.

Quod sapio satis est mihi, non ego curo Essequod Arcesilas aerumnosique Solones. PERSIUS — Satirae, iii. 78-79.

is their motto, under which they may lead apleasant and praiseworthy life, without trou-bling themselves with science or troubling sci-ence with them.

As regards those who wish to pursue a scien-tific method, they have now the choice of fol-lowing either the dogmatical or the sceptical,while they are bound never to desert the sys-tematic mode of procedure. When I mention, inrelation to the former, the celebrated Wolf, andas regards the latter, David Hume, I may leave,in accordance with my present intention, allothers unnamed. The critical path alone is stillopen. If my reader has been kind and patientenough to accompany me on this hitherto un-travelled route, he can now judge whether, if heand others will contribute their exertions to-

wards making this narrow footpath a high roadof thought, that which many centuries havefailed to accomplish may not be executed be-fore the close of the present—namely, to bringReason to perfect contentment in regard to thatwhich has always, but without permanent re-sults, occupied her powers and engaged herardent desire for knowledge.