rendered not only possible, but necessary, through
the perfect unity of this kind of knowledge, all
derived from pure concepts, without any influence
from experience, or from special intuitions leading
to a definite kind of experience, that might serve
to enlarge and increase it. Tecum habita et noris
quam sit tibi curta supellex (Persius, Sat. iv. 52).

Such a system of pure (speculative) reason I
hope myself to produce under the title of ' Meta-
physic of Nature.' It will not be half so large, yet
infinitely richer than this Critique of Pure Reason,



XXX PREFACE.

which has, first of all, to discover its source, nay,
the conditions of its possibility, in fact, to clear
and level a soil quite overgrown with weeds. Here
I expect from my readers the patience and im-
partiality of a judge, there the goodwill and aid
of a fellow-worker. For however completely all
the principles of the system have been propounded
in my Critique, the completeness of the whole
system requires also that no derivative concepts
should be omitted, such as cannot be found out
by an estimate a priori, but have to be discovered
step by step. There the synthesis of concepts has
been exhausted, here it will be requisite to do
the same for their analysis, a task which is easy
and an amusement rather than a labour.



-



TABLE OF CONTENTS TO THE
FIRST EDITION. 1



Introduction



PAGES



I. ELEMENTS OF TRANSCENDENTALISM.

PARTI. Transcendental JEsthetic . . 17 (19)

Section I. Of Space . . . . .20 (22)

Section II. Of Time 27 (30)

PART II. Transcendental Logic . . -44 (50)

Division I. Transcendental Analytic in two

books, with their chapters and sections . 56 (64)

Division II. Transcendental Dialectic in two

books, with their chapters and sections . 254 (293)



II. METHOD OF TRANSCENDENTALISM.

Chapter I. The Discipline of Pure Reason . 607 (708)

Chapter II. The Canon of Pure Reason . 682 (795)

Chapter III. The Architectonic of Pure Reason 714 (832)

Chapter IV. The History of Pure Reason . 731 (852)

1 Instead of this simple Table of Contents later editions have a much fuller
one (Supplement III) which, as Rosenkranz observes, obscures rather than
illustrates the articulation of the book.




INTRODUCTION. [p- ]

I. V

THE IDEA OP TRANSCENDENTAL PHILOSOPHY.

[Experience 1 is no doubt the first product of our
understanding, while employed in fashioning the
raw material of our sensations. It is therefore our
first instruction, and in its progress so rich in new
lessons that the chain of all future generations will
never be in want of new information that may be
gathered on that field. Nevertheless, experience is
by no means the only field to which our under-
standing can be confined. Experience tells us what
is, but not that it must be necessarily as it is, and
not otherwise. It therefore never gives us any really
general truths, and our reason, which is particularly
anxious for that class of knowledge, is roused by
it rather than satisfied. General truths, which [p. 2]
at the same time bear the character of an inward
necessity, must be independent of experience, clear
and certain by themselves. They are therefore called
knowledge a priori, while what is simply taken from

1 The beginning of this Introduction down to ' But what is still
more extraordinary/ is left out in the Second Edition. Instead of it
Supplement IV.

VOL. II. B



2 INTRODUCTION.

experience is said to be, in ordinary parlance, known
a posteriori or empirically only.

Now it appears, and this is extremely curious, that
even with our experience different kinds of know-
ledge are mixed up, which must have their origin
a priori, and which perhaps serve only to produce a
certain connection between our sensuous representa-
tions. For even if we remove from experience
everything that belongs to the senses, there remain
nevertheless certain original concepts, and certain
judgments derived from them, which must have had
their origin entirely a priori, and independent of all
experience, because it is owing to them that we are
able, or imagine we are able, to predicate more of the
objects of our senses than can be learnt from mere
experience, and that our propositions contain real
generality and strict necessity, such as mere em-
pirical knowledge can never supply.]

But l what is still more extraordinary is this, that
certain kinds of knowledge leave the field of all [p. 3]
possible experience, and seem to enlarge the sphere
of our judgments beyond the limits of experience
by means of concepts to which experience can never
supply any corresponding objects.

And it is in this very kind of knowledge which
transcends the world of the senses, and where ex-
perience can neither guide nor correct us, that reason
prosecutes its investigations, which by their import-
ance we consider far more excellent and by their

1 The Second Edition gives here a new heading: III, Philosophy
requires a science to determine a priori the possibility, the prin-
ciples, and the extent of all knowledge.



INTRODUCTION. 3

tendency far more elevated than anything the under-
standing can find in the sphere of phenomena. Nay,
we risk rather anything, even at the peril of error,
than that we should surrender such investigations,
either on the ground of their uncertainty, or from
any feeling of indifference or contempt l .

Now it might seem natural that, after we have
left the solid ground of experience, we should not
at once proceed to erect an edifice with knowledge
which we possess without knowing whence it came,
and trust to principles the origin of which is
unknown, without having made sure of the safety
of the foundations by means of careful examination.
It would seem natural, I say, that we should first
of all have asked the question how the mere under-
standing could arrive at all this knowledge a priori,
and what extent, what truth, and what value [p. 4]
it could possess. If we take natural to mean what
is just and reasonable, then indeed nothing could be
more natural. But if we understand by natural
what takes place ordinarily, then, on the contrary,
nothing is more natural and more intelligible than
that this examination should have been neglected
for so long a time. For one part of this knowledge,
namely, the mathematical, has always been in pos-
session of perfect trustworthiness ; and thus produces

1 The Second Edition adds here : ' These inevitable problems of
pure reason itself are, God, Freedom, and Immortality.' The
science which with all its apparatus is really intended for the
solution of these problems, is called Metaphysic. Its procedure is
at first dogmatic, i. e. unchecked by a previous examination of
what reason can and cannot do, before it engages confidently in so
arduous an undertaking.

B 2



4 INTRODUCTION.

a favourable presumption with regard to other parts
also, although these may be of a totally different
nature. Besides, once beyond the precincts of ex-
perience, and we are certain that experience can
never contradict us, while the charm of enlarging
our knowledge is so great that nothing will stop our
progress until we encounter a clear contradiction.
This can be avoided if only we are cautious in our
imaginations, which nevertheless remain what they
are, imaginations only. How far we can advance
independent of all experience in a priori knowledge
is shown by the brilliant example of mathematics.
It is true they deal with objects and knowledge
so far only as they can be represented in intuition.
But this is easily overlooked, because that intuition
itself may be given a priori, and be difficult to distin-
guish from a pure concept. Thus inspirited [p. 5]
by a splendid proof of the power of reason, the
desire of enlarging our knowledge sees no limits.
The light dove, piercing in her easy flight the air and
perceiving its resistance, imagines that flight would
be easier still in empty space. It was thus that
Plato left the world of sense, as opposing so many
hindrances to our understanding, and ventured beyond
on the wings of his ideas into the empty space of pure
understanding. He did not perceive that he was
making no progress by these endeavours, because he
had no resistance as a fulcrum on which to rest or to
apply his powers, in order to cause the understand-
ing to advance. It is indeed a very common fate of
human reason first of all to finish its speculative
edifice as soon as possible, and then only to inquire



INTRODUCTION. 5

whether the foundation be sure. Then all sorts of
excuses are made in order to assure us as to its
solidity, or to decline altogether such a late and
dangerous inquiry. The reason why during the
time of building we feel free from all anxiety and
suspicion and believe in the apparent solidity of
our foundation, is this : A great, perhaps the
greatest portion of what our reason finds to do
consists in the analysis of our concepts of objects.
This gives us a great deal of knowledge which,
though it consists in no more than in simplifi- [p. 6]
cations and explanations of what is comprehended in
our concepts (though in a confused manner), is yet con-
sidered as equal, at least in form, to new knowledge.
It only separates and arranges our concepts, it does
not enlarge them in matter or contents. As by this
process we gain a kind of real knowledge a priori,
which progresses safely and usefully, it happens
that our reason, without being aware of it, appro-
priates under that pretence propositions of a totally
different character, adding to given concepts new and
strange ones a priori, without knowing whence they
come, nay without even thinking of such a question.
I shall therefore at the very outset treat of the
distinction between these two kinds of knowledge.

Of the distinction between analytical and synthetical
Judgments.

In all judgments in which there is a relation
between subject and predicate (I speak of affirmative
judgments only, the application to negative ones
being easy), that relation can be of two kinds. Either



6 INTRODUCTION.

the predicate B belongs to the subject A as some-
thing contained (though covertly) in the concept A ;
or B lies outside the sphere of the concept A, though
somehow connected with it. In the former case I
call the judgment analytical, in the latter synthetical.
Analytical judgments (affirmative) are therefore [p. 7]
those in which the connection of the predicate with
the subject is conceived through identity, while others
in which that connection is conceived without identity,
may be called synthetical. The former might be
called illustrating, the latter expanding judgments,
because in the former nothing is added by the pre-
dicate to the concept of the subject, but the concept
is only divided into its constituent concepts which
were always conceived as existing within it, though
confusedly; while the latter add to the concept of
the subject a predicate not conceived as existing
within it, and not to be extracted from it by any
process of mere analysis. If I say, for instance, All
bodies are extended, this is an analytical judgment.
I need not go beyond the concept connected with
the name of body, in order to find that extension is
connected with it. I have only to analyse that
concept and become conscious of the manifold elements
always contained in it, in order to find that predicate.
This is therefore an analytical judgment. But if I
say, All bodies are heavy, the predicate is something
quite different from what I think as the mere concept
of body. The addition of such a predicate gives us
a synthetical judgment.

[It becomes clear from this \

1 These two paragraphs to ' In synthetical judgments a prion



INTRODUCTION. 7

[i. That our knowledge is in no way extended
by analytical judgments, but that all they [p. 8]
effect is to put the concepts which we possess into
better order and render them more intelligible.

2. That in synthetical judgments I must have
besides the concept of the subject something else (x)
on which the understanding relies in order to know
that a predicate, not contained in the concept, never-
theless belongs to it.

In empirical judgments this causes no difficulty, be-
cause this x is here simply the complete experience of
an object which. I conceive by the concept A, that
concept forming one part only of my experience.
For though I do not include the predicate of gravity
in the general concept of body, that concept never-
theless indicates the complete experience through
one of its parts, so that I may add other parts also
of the same experience, all belonging to that concept.
I may first, by an analytical process, realise the
concept of body through the predicates of extension,
impermeability, form, &c, all of which are contained
in it. Afterwards I expand my knowledge, and
looking back to the experience from which my
concept of body was abstracted, I find gravity always
connected with the before -mentioned predicates.
Experience therefore is the x which lies beyond the
concept A, and on which rests the possibility of a
synthesis of the predicate of gravity B with the
concept A.]

In synthetical judgments a priori, however, [p. 9]

however/ are left out in the Second Edition, and replaced by
Supplement V.



8 INTRODUCTION.

that help is entirely wanting. If I want to go
beyond the concept A in order to find another
concept B, connected with it, where is there any-
thing on which I may rest and through which a
synthesis might become possible, considering that I
cannot have the advantage of looking about in the
field of experience 1 ? Take the proposition that all
which happens has its cause. In the concept of
somethipg that happens I no doubt conceive of some-
thing existing preceded by time, and from this
certain analytical judgments may be deduced. But
the concept of cause is entirely outside that concept,
and indicates something different from that which
happens, and is by no means contained in that re-
presentation. How can I venture then to predicate of
that which happens something totally different from
it, and to represent the concept of cause, though not
contained in it, as belonging to it, and belonging to
it by necessity? What is here the unknown x, on
which the understanding may rest in order to find
beyond the concept A a foreign predicate B, which
nevertheless is believed to be connected with it ? It
cannot be experience, because the proposition that
all which happens has its cause represents this second
predicate as added to the subject not only with
greater generality than experience can ever supply,
but also with a character of necessity, and therefore
purely a priori, and based on concepts. All our
speculative knowledge a priori aims at and rests on
such synthetical, i.e. expanding propositions, [p. 10]
for the analytical are no doubt very important and
necessary, yet only in order to arrive at that clearness



INTRODUCTION. 9

of concepts which is requisite for a safe and wide
synthesis, serving as a really new addition to what
we possess already.

[We 1 have here a certain mystery 2 before us,
which must be cleared up before any advance into the
unlimited field of a pure knowledge of the under-
standing can become safe and trustworthy. We
must discover on the largest scale the ground of the
possibility of synthetical judgments a priori; we
must understand the conditions which render every
class of them possible, and endeavour not only to
indicate in a sketchy outline, but to define in its
fulness and practical completeness, the whole of
that knowledge, which forms a class by itself,
systematically arranged according to its original
sources, its divisions, its extent and its limits. So
much for the present with regard to the peculiar
character of synthetical judgments.]

It will now be seen how there can be a special
science serving as a critique of pure reason, [p. 1 1]
[Every kind of knowledge is called pure, if not mixed
with anything heterogeneous. But more particularly
is that knowledge called absolutely pure, which is
not mixed up with any experience or sensation, and
is therefore possible entirely a priori] Eeason is the
faculty which supplies the principles of knowledge

1 This paragraph left out in the Second Edition, and replaced by
Supplement VI.

2 If any of the ancients had ever thought of asking this question,
this alone would have formed a powerful barrier against all systems
of pure reason to the present day, and would have saved many vain
attempts undertaken blindly and without a true knowledge of the
subject in hand.



IO INTRODUCTION.

a priori. Pure reason therefore is that faculty
which supplies the principles of knowing anything
entirely a priori. An Organum of pure reason ought
to comprehend all the principles by which pure
knowledge a priori can be acquired and fully estab-
lished. A complete application of such an Organum
would give us a System of Pure Reason. But as
that would be a difficult task, and as at present
it is still doubtful whether such an expansion of
our knowledge is here possible, we may look on a
mere criticism of pure reason, its sources and limits,
as a kind of preparation for a complete system of
pure reason. It should be called a critique, not a
doctrine, of pure reason. Its usefulness would be
negative only, serving for a purging rather than for
an expansion of our reason, and, what after all is a
considerable gain, guarding reason against errors.

I call all knowledge transcendental which is occu-
pied not so much with objects, as with our a priori
concepts of objects 1 . A system of such [p. 12]
concepts might be called Transcendental Philosophy.
But for the present this is again too great an under-
taking. We should have to treat therein completely
both of analytical knowledge, and of synthetical
knowledge a priori, which is more than we intend to
do, being satisfied to carry on the analysis so far
only as is indispensably necessary in order to under-
stand in their whole extent the principles of synthesis
a priori, which alone concern us. This investigation
which should be called a transcendental critique, but

1 'As with our manner of knowing objects, so far as this is
meant to be possible a priori.' Second Edition.



INTRODUCTION. I I

not a systematic doctrine, is all we are occupied with
at present. It is not meant to extend our knowledge,
but only to rectify it, and to become the test of the
value of all a priori knowledge. Such a critique
therefore is a preparation for a New Organum, or,
if that should not be possible, for a Canon at least,
according to which hereafter a complete system of
a philosophy of pure reason, whether it serve for an
expansion or merely for a limitation of it, may be
carried out, both analytically and synthetically.
That such a system is possible, nay that it need
not be so comprehensive as to prevent the hope of its
completion, may be gathered from the fact that it would
have to deal, not with the nature of things, which
is endless, but with the understanding which [p. 13]
judges of the nature of things, and this again so far
only as its knowledge a priori is concerned. What-
ever the understanding possesses, as it has not to be
looked for without, can hardly escape our notice,
nor is there any reason to suppose that it will prove
too extensive for a complete inventory, and for such
a valuation as shall assign to it its true merits or*
demerits l .

II.

DIVISION OF TRANSCENDENTAL PHILOSOPHY.

Transcendental Philosophy is with us an idea
(of a science) only, for which the critique of pure
reason should trace, according to fixed principles, an

1 Here follows Supplement VII in Second Edition.



1 2 INTRODUCTION.

architectonic plan, guaranteeing the completeness
and certainty of all parts of which the building
consists. (It is a system of all principles of pure
reason.) The reason why we do not call such a
critique a transcendental philosophy in itself is simply
this, that in order to be a complete system, it
ought to contain likewise a complete analysis of
the whole of human knowledge a priori. It is true
that our critique must produce a complete list of
all the fundamental concepts which constitute pure
knowledge. But it need not give a detailed analysis
of these concepts, nor a complete list of all derivative
concepts. Such an analysis would be out of [p. 14]
place, because it is not beset with the doubts and
difficulties which are inherent in synthesis, and which
alone necessitate a critique of pure reason. Nor
would it answer our purpose to take the responsi-
bility of the completeness of such an analysis and
derivation. This completeness of analysis, however,
and of derivation from such a priori concepts as
we shall have to deal with presently, may easily be
supplied, if only they have first been laid down as
perfect principles of synthesis, and nothing is wanting
to them in that respect.

All that constitutes transcendental philosophy be-
longs to the critique of pure reason, nay it is the com-
plete idea of transcendental philosophy, but not yet
the whole of that philosophy itself, because it carries
the analysis so far only as is requisite for a complete
examination of synthetical knowledge a priori.

The most important consideration in the arrange-
ment of such a science is that no concepts should be



INTRODUCTION. 1 3

admitted which contain anything empirical, and that
the a priori knowledge shall be perfectly pure.
Therefore, although the highest principles of morality
and their fundamental concepts are a priori [p. 15]
knowledge, they do not belong to transcendental
philosophy, because the concepts of pleasure and
pain, desire, inclination, free-will, etc., which are all
of empirical origin, must here be presupposed. Tran-
scendental philosophy is the wisdom of pure specula-
tive reason. Everything practical, so far as it contains
motives, has reference to sentiments, and these be-
long to empirical sources of knowledge.

If we wish to carry out a proper division of our
science systematically, it must contain first a doc-
trine of the elements, secondly, a doctrine of the
method of pure reason. Each of these principal
divisions will have its subdivisions, the grounds of
which cannot however be explained here. So much
only seems necessary for previous information, that
there are two stems of human knowledge, which per-
haps may spring from a common root, unknown to
us, viz. sensibility and the understanding, objects being
given by the former and thought by the latter. If
our sensibility should contain a -priori representations,
constituting conditions under which alone objects
can be given, it would belong to transcendental philo-
sophy, and the doctrine of this transcendental [p. 16]
sense-perception would necessarily form the first part
of the doctrine of elements, because the conditions
under which alone objects of human knowledge can
be given must precede those under which they are
thought.



CKITIQUE OF PUEE BEASON.



I.

THE ELEMENTS OF TEANSCENDENTALISM.



THE

ELEMENTS OF TEANSCENDENTALISM. [p.**]

FIRST PART.
TRANSCENDENTAL ^ESTHETIC.

Whatever the process and the means may be by
which knowledge reaches its objects, there is one
that reaches them directly, and forms the ultimate
material of all thought, viz. intuition (Anschau-
ung). This is possible only when the object is
given, and the object can be given only (to human
beings at least) through a certain affection of the
mind (Gemuth).

This faculty (receptivity) of receiving representa-
tions (Vorstellungen), according to the manner in
which we are affected by objects, is called sensibility
(Sinnlichkeit).

Objects therefore are given to us through our
sensibility. Sensibility alone supplies us with in-
tuitions- (Anschauungen). These intuitions become
thought through the understanding (Verstand), and
hence arise conceptions (Begriffe). All thought
therefore must, directly or indirectly, go back to in-
tuitions (Anschauungen), i. e. to our sensibility, be-
cause in no other way can objects be given to us.

The effect produced by an object upon the faculty
of representation (Vorstellungsfahigkeit), so [p. 20.]

vol. 11. c



1 8 TRANSCENDENTAL ^ESTHETIC.

far as we are affected by it, is called sensation
(Ernpfindung). An intuition (Anschauung) of an
object, by means of sensation, is called empirical.
The undefined object of such an empirical intuition
is called phenomenon (Erscheinung).

In a phenomenon I call that which corresponds to
the sensation its matter ; but that which causes the