Critique of Pure Reason

by Immanuel Kant

  This shows that the philosophy of Leibniz and Wolf has given a totally wrong direction to all investigations into the nature and origin of our knowledge, by representing the difference between the sensible and the intelligible as logical only. That difference is in truth transcendental. It affects not the form only, as being more or less confused, but the origin and contents of our knowledge; so that by our sensibility we know the nature of things by themselves not confusedly only, but not at all. If we drop our subjective condition, the object, as represented with its qualities bestowed on it by sensuous intuition, is nowhere to be found, and cannot possibly be found; because its form, as phenomenal appearance, is determined by those very subjective conditions.

  It has been the custom to distinguish in phenomena that which is essentially inherent in their intuition and is recognised by every human being, from that which belongs to their intuition accidentally only, being valid not for sensibility in general, but only for a particular position and organisation of this or that sense. In that case the former kind of knowledge is said to represent the object by itself, the latter its appearance only. But that distinction is merely empirical. If, as generally happens, people are satisfied with that distinction, without again, as they ought, treating the first empirical intuition as purely phenomenal also, in which nothing can be found belonging to the thing by itself, our transcendental distinction is lost, and we believe that we know things by themselves, though in the world of sense, however far we may carry our investigation, we can never have anything before us but mere phenomena. To give an illustration. People might call the rainbow a mere phenomenal appearance during a sunny shower, but the rain itself the thing by itself. This would be quite right, physically speaking, and taking rain as something which, in our ordinary experience and under all possible relations to our senses, can be determined thus and thus only in our intuition. But if we take the empirical in general, and ask, without caring whether it is the same with every particular observer, whether it represents a thing by itself (not the drops of rain, for these are already, as phenomena, empirical objects), then the question as to the relation between the representation and the object becomes transcendental, and not only the drops are mere phenomena, but even their round shape, nay even the space in which they fall, are nothing by themselves, but only modifications or fundamental dispositions of our sensuous intuition, the transcendental object remaining unknown to us.

  The second important point in our transcendental æsthetic is, that it should not only gain favour as a plausible hypothesis, but assume as certain and undoubted a character as can be demanded of any theory which is to serve as an organum. In order to make this certainty self-evident we shall select a case which will make its validity palpable.

  Let us suppose that space and time are in themselves objective, and conditions of the possibility of things by themselves. Now there is with regard to both a large number of a priori apodictic and synthetical propositions, and particularly with regard to space, which for this reason we shall chiefly investigate here as an illustration. As the propositions of geometry are known synthetically a priori, and with apodictic certainty, I ask, whence do you take such propositions? and what does the understanding rely on in order to arrive at such absolutely necessary and universally valid truths? There is no other way but by concepts and intuitions, and both as given either a priori or a posteriori. The latter, namely empirical concepts, as well as the empirical intuition on which they are founded, cannot yield any synthetical propositions except such as are themselves also empirical only, that is, empirical propositions, which can never possess that necessity and absolute universality which are characteristic of all geometrical propositions. As to the other and only means of arriving at such knowledge through mere concepts or intuitions a priori, it must be clear that only analytical, but no synthetical knowledge can ever be derived from mere concepts. Take the proposition that two straight lines cannot enclose a space and cannot therefore form a figure, and try to deduce it from the concept of straight lines and the number two; or take the proposition that with three straight lines it is possible to form a figure, and try to deduce that from those concepts. All your labour will be lost, and in the end you will be obliged to have recourse to intuition, as is always done in geometry. You then give yourselves an object in intuition. But of what kind is it? Is it a pure intuition a priori or an empirical one? In the latter case, you would never arrive at a universally valid, still less at an apodictic proposition, because experience can never yield such. You must therefore take the object as given a priori in intuition, and found your synthetical proposition on that. If you did not possess in yourselves the power of a priori intuition, if that subjective condition were not at the same time, as to the form, the general condition a priori under which alone the object of that (external) intuition becomes possible, if, in fact, the object (the triangle) were something by itself without any reference to you as the subject, how could you say that what exists necessarily in your subjective conditions of constructing a triangle, belongs of necessity to the triangle itself? For you could not add something entirely new (the figure) to your concepts of three lines, something which should of necessity belong to the object, as that object is given before your knowledge of it, and not by it. If therefore space, and time also, were not pure forms of your intuition, which contains the a priori conditions under which alone things can become external objects to you, while, without that subjective condition, they are nothing, you could not predicate anything of external objects a priori and synthetically. It is therefore beyond the reach of doubt, and not possible only or probable, that space and time, as the necessary conditions of all experience, external and internal, are purely subjective conditions of our intuition, and that, with reference to them, all things are phenomena only, and not things thus existing by themselves in such or such wise. Hence, so far as their form is concerned, much may be predicated of them a priori, but nothing whatever of the things by themselves on which these phenomena may be grounded.9

  1 The Germans are the only people who at present (1781) use the word æsthetic for what others call criticism of taste. There is implied in that name a false hope, first conceived by the excellent analytical philosopher, Baumgarten, of bringing the critical judgment of the beautiful under rational principles, and to raise its rules to the rank of a science. But such endeavours are vain. For such rules or criteria are, according to their principal sources, empirical only, and can never serve as definite a priori rules for our judgment in matters of taste; on the contrary, our judgment is the real test of the truth of such rules. It would be advisable therefore to drop the name in that sense, and to apply it to a doctrine which is a real science, thus approaching more nearly to the language and meaning of the ancients with whom the division into was very famous (or to share that name in common with speculative philosophy, and thus to use æsthetic sometimes in a transcendental, sometimes in a psychological sense).

  2 No. 5 (No. 4) is differently worded in the Second Edition; see Supplement VIII.

  3 This passage to 'my object in what I have said' is differently worded in the Second Edition; see Supplement IX.

  4 In the Second Edition the title is, Metaphysical exposition of the concept of time, with reference to par. 5, Transcendental exposition of the concept of time.

  5 I retain the reading of the First Edition, vor derselben, nicht durch dieselbe. Von denselben, the reading of later editions, is wrong; the emendation of Rosenkranz, vor denselben, nicht durch dieselben, unnecessary. The Second Edition has likewise vor derselben.

  6 Here follows in the Second Edition, Supplement X.

  7 Read allein instead of alle.

  8 I can say indeed that my representations follow one another, but this means no more than that we are conscious of them as in a temporal succession, that is, according to the form of our own internal sense. Time, therefore, is nothing by itself, nor is it a determination inherent objectively in things.

  9 Here follows in the Se
cond Edition, Supplement XI.

  | Go to Table of Contents |

  Introduction

  The Idea Of A Transcendental Logic

  I

  Of Logic in General

  OUR knowledge springs from two fundamental sources of our soul; the first receives representations (receptivity of impressions), the second is the power of knowing an object by these representations (spontaneity of concepts). By the first an object is given us, by the second the object is thought, in relation to that representation which is a mere determination of the soul. Intuition therefore and concepts constitute the elements of all our knowledge, so that neither concepts without an intuition corresponding to them, nor intuition without concepts can yield any real knowledge.

  Both are either pure or empirical. They are empirical when sensation, presupposing the actual presence of the object, is contained in it. They are pure when no sensation is mixed up with the representation. The latter may be called the material of sensuous knowledge. Pure intuition therefore contains the form only by which something is seen, and pure conception the form only by which an object is thought. Pure intuitions and pure concepts only are possible a priori, empirical intuitions and empirical concepts a posteriori.

  We call sensibility the receptivity of our soul, or its power of receiving representations whenever it is in any wise affected, while the understanding, on the contrary, is with us the power of producing representations, or the spontaneity of knowledge. We are so constituted that our intuition must always be sensuous, and consist of the mode in which we are affected by objects. What enables us to think the objects of our sensuous intuition is the understanding. Neither of these qualities or faculties is preferable to the other. Without sensibility objects would not be given to us, without understanding they would not be thought by us. Thoughts without contents are empty, intuitions without concepts are blind. Therefore it is equally necessary to make our concepts sensuous, i.e. to add to them their object in intuition, as to make our intuitions intelligible, i.e. to bring them under concepts. These two powers or faculties cannot exchange their functions. The understanding cannot see, the senses cannot think. By their union only can knowledge be produced. But this is no reason for confounding the share which belongs to each in the production of knowledge. On the contrary, they should always be carefully separated and distinguished, and we have therefore divided the science of the rules of sensibility in general, i.e. æsthetic, from the science of the rules of the understanding in general, i.e. logic.

  Logic again can be taken in hand for two objects, either as logic of the general or of a particular use of the understanding. The former contains all necessary rules of thought without which the understanding cannot be used at all. It treats of the understanding without any regard to the different objects to which it may be directed. Logic of the particular use of the understanding contains rules how to think correctly on certain classes of objects. The former may be called Elementary Logic, the latter the Organum of this or that science. The latter is generally taught in the schools as a preparation for certain sciences, though, according to the real progress of the human understanding, it is the latest achievement, which does not become possible till the science itself is really made, and requires only a few touches for its correction and completion. For it is clear that the objects themselves must be very well known before it is possible to give rules according to which a science of them may be established.

  General logic is either pure or applied. In the former no account is taken of any empirical conditions under which our understanding acts, i.e. of the influence of the senses, the play of imagination, the laws of memory, the force of habit, the inclinations, and therefore the sources of prejudice also, nor of anything which supplies or seems to supply particular kinds of knowledge; for all this applies to the understanding under certain circumstances of its application only, and requires experience as a condition of knowledge. General but pure logic has to deal with principles a priori only, and is a canon of the understanding and of reason, though with reference to its formal application only, irrespective of any contents, whether empirical or transcendental. General logic is called applied, if it refers to the rules of the use of our understanding under the subjective empirical conditions laid down in psychology. It therefore contains empirical principles, yet it is general, because referring to the use of the understanding, whatever its objects may be. It is neither a canon of the understanding in general nor an organum of any particular science, but simply a catharticon of the ordinary understanding.

  In general logic, therefore, that part which is to constitute the science of pure reason must be entirely separated from that which forms applied, but for all that still general logic. The former alone is a real science, though short and dry, as a practical exposition of an elementary science of the understanding ought to be. In this logicians should never lose sight of two rules:—

  1. As general logic it takes no account of the contents of the knowledge of the understanding nor of the difference of its objects. It treats of nothing but the mere form of thought.

  2. As pure logic it has nothing to do with empirical principles, and borrows nothing from psychology (as some have imagined); psychology, therefore, has no influence whatever on the canon of the understanding. It proceeds by way of demonstration, and everything in it must be completely a priori.

  What I call applied logic (contrary to common usage according to which it contains certain exercises on the rules of pure logic) is a representation of the understanding and of the rules according to which it is necessarily applied in concrete, i.e. under the accidental conditions of the subject, which may hinder or help its application, and are all given empirically only. It treats of attention, its impediments and their consequences, the sources of error, the states of doubt, hesitation, and conviction, etc., and general and pure logic stands to it in the same relation as pure ethics, which treat only of the necessary moral laws of a free will, to applied ethics, which consider these laws as under the influence of sentiments, inclinations, and passions to which all human beings are more or less subject. This can never constitute a true and demonstrated science, because, like applied logic, it depends on empirical and psychological principles.

  II

  Of Transcendental Logic

  General logic, as we saw, takes no account of the contents of knowledge, i.e. of any relation between it and its objects, and considers the logical form only in the relation of cognitions to each other, that is, it treats of the form of thought in general. But as we found, when treating of Transcendental Æsthetic, that there are pure as well as empirical intuitions, it is possible that a similar distinction might appear between pure and empirical thinking. In this case we should have a logic in which the contents of knowledge are not entirely ignored, for such a logic which should contain the rules of pure thought only, would exclude only all knowledge of a merely empirical character. It would also treat of the origin of our knowledge of objects, so far as that origin cannot be attributed to the objects, while general logic is not at all concerned with the origin of our knowledge, but only considers representations (whether existing originally a priori in ourselves or empirically given to us), according to the laws followed by the understanding, when thinking and treating them in their relation to each other. It is confined therefore to the form imparted by the understanding to the representations, whatever may be their origin.

  And here I make a remark which should never be lost sight of, as it extends its influence on all that follows. Not every kind of knowledge a priori should be called transcendental (i.e. occupied with the possibility or the use of knowledge a priori), but that only by which we know that and how certain representations intuitional or conceptual) can be used or are possible a priori only. Neither space nor any a priori geometrical determination of it is a transcendental representation; but that knowledge only is rightly called transcendental which teaches us that these representations cannot be of empirical origin, and how they can yet
refer a priori to objects of experience. The application of space to objects in general would likewise be transcendental, but, if restricted to objects of sense, it is empirical. The distinction between transcendental and empirical belongs therefore to the critique of knowledge, and does not affect the relation of that knowledge to its objects.

  On the supposition therefore that there may be concepts, having an a priori reference to objects, not as pure or sensuous intuitions, but as acts of pure thought, being concepts in fact, but neither of empirical nor æsthetic origin, we form by anticipation an idea of a science of that knowledge which belongs to the pure understanding and reason, and by which we may think objects entirely a priori. Such a science, which has to determine the origin, the extent, and the objective validity of such knowledge, might be called Transcendental Logic, having to deal with the laws of the understanding and reason in so far only as they refer a priori to objects, and not, as general logic, in so far as they refer promiscuously to the empirical as well as to the pure knowledge of reason.