Critique of Pure Reason (Transcendental Dialectic

Immanuel Kant's
Critique
trans. by Norman Kemp Smith
P 297
TRANSCENDENTAL LOGIC
SECOND DIVISION
TRANSCENDENTAL DIALECTIC
INTRODUCTION
I
TRANSCENDENTAL ILLUSION
WE have already entitled dialectic in general a logic of 
illusion. This does not mean a doctrine of probability; for 
probability is truth, known however on insufficient grounds, and
the knowledge of which, though thus imperfect, is not on that
account deceptive; and such doctrine, accordingly, is not to be
separated from the analytic part of logic. Still less justification
have we for regarding appearance and illusion as being identical.
For truth or illusion is not in the object, in so far as it is
intuited, but in the judgment about it, in so far as it is thought.
It is therefore correct to say that the senses do not err -- not
because they always judge rightly but because they do not
judge at all. Truth and error, therefore, and consequently also
illusion as leading to error, are only to be found in the judgment,
i.e. only in the relation of the object to our understanding.
In any knowledge which completely accords with the laws
of understanding there is no error. In a representation of the
senses -- as containing no judgment whatsoever -- there is also
no error. No natural force can of itself deviate from its own
laws. Thus neither the understanding by itself (uninfluenced
by another cause), nor the senses by themselves, would fall
into error. The former would not, since, if it acts only according
to its own laws, the effect (the judgment) must necessarily
be in conformity with these laws; conformity with the laws
P 298
of the understanding is the formal element in all truth. In
the senses there is no judgment whatsoever, neither a true
nor a false judgment. Now since we have no source of knowledge
besides these two, it follows that error is brought about
solely by the unobserved influence of sensibility on the 
understanding, through which it happens that the subjective grounds
of the judgment enter into union with the objective grounds
and make these latter deviate from their true function, -- just
as a body in motion would always of itself continue in a
straight line in the same direction, but if influenced by another
force acting in another direction starts off into curvilinear
motion. In order to distinguish the specific action of understanding
from the force which is intermixed with it, it is necessary
to regard the erroneous judgment as the diagonal between
two forces -- forces which determine the judgment in different
directions that enclose, as it were, an angle -- and to resolve
this composite action into the simple actions of the understanding
and of the sensibility. In the case of pure a priori
judgments this is a task which falls to be discharged by 
transcendental reflection, through which, as we have already shown,
every representation is assigned its place in the corresponding
faculty of knowledge, and by which the influence of the one
upon the other is therefore likewise distinguished.
We are not here concerned with empirical (e.g. optical)
illusion, which occurs in the empirical employment of rules of
understanding that are otherwise correct, and through which
the faculty of judgment is misled by the influence of imagination;
we are concerned only with transcendental illusion, which
exerts its influence on principles that are in no wise intended for
use in experience, in which case we should at least have had a
criterion of their correctness. In defiance of all the warnings of
criticism, it carries us altogether beyond the empirical employment
of categories and puts us off with a merely deceptive 
extension of pure understanding.
++ Sensibility, when subordinated to understanding, as the object
upon which the latter exercises its function, is the source of real
modes of knowledge. But the same sensibility, in so far as it 
influences the operation of understanding, and determines it to make
judgments, is the ground of error.
P 298
We shall entitle the principles whose
application is confined entirely within the limits of possible
P 299
experience, immanent; and those, on the other hand, which
profess to pass beyond these limits, transcendent. In the case of
these latter, I am not referring to the transcendental employment
or misemployment of the categories, which is merely an
error of the faculty of judgment when it is not duly curbed by
criticism, and therefore does not pay sufficient attention to the
bounds of the territory within which alone free play is allowed
to pure understanding. I mean actual principles which incite
us to tear down all those boundary-fences and to seize possession
of an entirely new domain which recognises no limits of
demarcation. Thus transcendental and transcendent are not
interchangeable terms. The principles of pure understanding,
which we have set out above, allow only of empirical and not
of transcendental employment, that is, employment extending
beyond the limits of experience. A principle, on the other
hand, which takes away these limits, or even commands us
actually to transgress them, is called transcendent. If our
criticism can succeed in disclosing the illusion in these alleged
principles, then those principles which are of merely empirical
employment may be called, in opposition to the others, 
immanent principles of pure understanding.
Logical illusion, which consists in the mere imitation
of the form of reason (the illusion of formal fallacies), arises
entirely from lack of attention to the logical rule. As soon
as attention is brought to bear on the case that is before us,
the illusion completely disappears. Transcendental illusion,
on the other hand, does not cease even after it has been 
detected and its invalidity clearly revealed by transcendental
criticism (e.g. the illusion in the proposition: the world must
have a beginning in time). The cause of this is that there are
fundamental rules and maxims for the employment of our
reason (subjectively regarded as a faculty of human knowledge),
and that these have all the appearance of being objective
principles. We therefore take the subjective necessity
of a connection of our concepts, which is to the advantage of
the understanding, for an objective necessity in the 
determination of things in themselves. This is an illusion which
can no more be prevented than we can prevent the sea
appearing higher at the horizon than at the shore, since we see
it through higher light rays; or to cite a still better example,
P 300
than the astronomer can prevent the moon from appearing
larger at its rising, although he is not deceived by this illusion.
The transcendental dialectic will therefore content itself
with exposing the illusion of transcendent judgments, and at
the same time taking precautions that we be not deceived by
it. That the illusion should, like logical illusion, actually 
disappear and cease to be an illusion, is something which 
transcendental dialectic can never be in a position to achieve. For
here we have to do with a natural and inevitable illusion,
which rests on subjective principles, and foists them upon us
as objective; whereas logical dialectic in its exposure of 
deceptive inferences has to do merely with an error in the following
out of principles, or with an illusion artificially created
in imitation of such inferences. There exists, then, a natural
and unavoidable dialectic of pure reason -- not one in which a
bungler might entangle himself through lack of knowledge,
or one which some sophist has artificially invented to confuse
thinking people, but one inseparable from human reason, and
which, even after its deceptiveness has been exposed, will not
cease to play tricks with reason and continually entrap it into
momentary aberrations ever and again calling for correction.
II
PURE REASON AS THE SEAT OF TRANSCENDENTAL
ILLUSION
A
Reason in general
All our knowledge starts with the senses, proceeds from
thence to understanding, and ends with reason, beyond which
there is no higher faculty to be found in us for elaborating the
matter of intuition and bringing it under the highest unity of
thought. Now that I have to give an explanation of this highest
faculty of knowledge, I find myself in some difficulty. Reason,
like understanding, can be employed in a merely formal, that
is, logical manner, wherein it abstracts from all content of
knowledge. But it is also capable of a real use, since it contains
within itself the source of certain concepts and principles,
P 301
which it does not borrow either from the senses or from the
understanding. The former faculty has long since been defined
by logicians as the faculty of making mediate inferences (in 
distinction from immediate inferences, consequentiis immediatis);
but the nature of the other faculty, which itself gives birth to 
concepts, is not to be understood from this definition. Now since
we are here presented with a division of reason into a logical
and a transcendental faculty, we are constrained to seek for a
higher concept of this source of knowledge which includes
both concepts as subordinate to itself. Following the analogy
of concepts of understanding, we may expect that the logical
concept will provide the key to the transcendental, and that
the table of the functions of the former will at once give us the
genealogical tree of the concepts of reason.
In the first part of our transcendental logic we treated the
understanding as being the faculty of rules; reason we shall
here distinguish from understanding by entitling it the faculty
of principles.
The term 'principle' is ambiguous, and commonly signifies
any knowledge which can be used as a principle,
although in itself, and as regards its proper origin, it is no
principle. Every universal proposition, even one derived from
experience, through induction, can serve as major premiss in
a syllogism; but it is not therefore itself a principle. The
mathematical axioms (e.g. that there can only be one straight
line between two points) are instances of universal a priori
knowledge, and are therefore rightly called principles, 
relatively to the cases which can be subsumed under them. But I
cannot therefore say that I apprehend this property of straight
lines in general and in itself, from principles; I apprehend it
only in pure intuition.
Knowledge from principles is, therefore, that knowledge
alone in which I apprehend the particular in the universal
through concepts. Thus every syllogism is a mode of deducing
knowledge from a principle. For the major premiss always
gives a concept through which everything that is subsumed
under the concept as under a condition is known from the 
concept according to a principle. Now since any universal 
knowledge can serve as major premiss in a syllogism, and since the
understanding presents us with universal a priori propositions
P 302
of this kind, they can also be called principles in respect of
their possible employment.
 But if we consider them in themselves in relation to their
origin, these fundamental propositions of pure understanding
are anything rather than knowledge based on concepts. For
they would not even be possible a priori, if we were not 
supported by pure intuition (in mathematics), or by conditions of
a possible experience in general. That everything that happens
has a cause cannot be inferred merely from the concept of
happening in general; on the contrary, it is this fundamental
proposition which shows how in regard to that which happens
we are in a position to obtain in experience any concept 
whatsoever that is really determinate.
The understanding can, then, never supply any synthetic
modes of knowledge derived from concepts; and it is such
modes of knowledge that are properly, without qualification,
to be entitled 'principles'. All universal propositions, however,
may be spoken of as 'principles' in a comparative sense.
It has long been wished -- and sometime perhaps (who
knows when! ) may be fulfilled -- that instead of the endless
multiplicity of civil laws we should be able to fall back on their
general principles. For it is in these alone that we can hope to
find the secret of what we are wont to call the simplifying of
legislation. In this domain, however, the laws are only 
limitations imposed upon our freedom in order that such freedom
may completely harmonise with itself; hence they are directed
to something which is entirely our own work, and of which we
ourselves, through these concepts, can be the cause. But that
objects in themselves, the very nature of things, should stand
under principles, and should be determined according to mere
concepts, is a demand which, if not impossible, is at least quite
contrary to common sense. But however that may be (it is
a question which we still have to discuss), it is now at least
evident that knowledge derived from principles which are
genuinely such is something quite different from knowledge
obtained merely through the understanding. The latter may,
indeed, also take the form of a principle and thus be prior to
some other knowledge, but in itself, in so far as it is 
P 303
synthetic, it does not depend on thought alone, nor contain in
itself a universal obtained from concepts.
Understanding may be regarded as a faculty which secures
the unity of appearances by means of rules, and reason as
being the faculty which secures the unity of the rules of 
understanding under principles. Accordingly, reason never applies
itself directly to experience or to any object, but to understanding,
in order to give to the manifold knowledge of the latter
an a priori unity by means of concepts, a unity which may be
called the unity of reason, and which is quite different in kind
from any unity that can be accomplished by the understanding.
This is the universal concept of the faculty of reason in so
far as it has been possible to make it clear in the total absence
of examples. These will be given in the course of our 
argument. 
B
The Logical Employment of Reason
A distinction is commonly made between what is immediately
known and what is merely inferred. That in a figure which is
bounded by three straight lines there are three angles, is known
immediately; but that the sum of these angles is equal to two right
angles, is merely inferred. Since we have constantly to make
use of inference, and so end by becoming completely accustomed
to it, we no longer take notice of this distinction, and
frequently, as in the so-called deceptions of the senses, treat as
being immediately perceived what has really only been inferred.
In every process of reasoning there is a fundamental 
proposition, and another, namely the conclusion, which is drawn
from it, and finally, the inference (logical sequence) by which
the truth of the latter is inseparably connected with the truth
of the former. If the inferred judgment is already so contained
in the earlier judgment that it may be derived from it without
the mediation of a third representation, the inference is called
immediate (consequentia immediata) -- I should prefer to entitle
it inference of the understanding. But if besides the knowledge
contained in the primary proposition still another judgment
P 304
is needed to yield the conclusion, it is to be entitled an
inference of the reason. In the proposition : "All men are
mortal", there are already contained the propositions : "some
men are mortal", "some mortal beings are men", "nothing
that is not mortal is a man"; and these are therefore immediate
conclusions from it. On the other hand, the proposition: "All
learned beings are mortal", is not contained in the fundamental
judgment (for the concept of learned beings does not
occur in it at all), and it can only be inferred from it by means
of a mediating judgment.
In every syllogism I first think a rule (the major premiss)
through the understanding. Secondly, I subsume something
known under the condition of the rule by means of judgment
(the minor premiss). Finally, what is thereby known I determine
through the predicate of the rule, and so a priori through
reason (the conclusion). The relation, therefore, which the
major premiss, as the rule, represents between what is known
and its condition is the ground of the different kinds of 
syllogism. Consequently, syllogisms, like judgments, are of three
kinds, according to the different ways in which, in the 
understanding, they express the relation of what is known; they
are either categorical, hypothetical, or disjunctive.
If, as generally happens, the judgment that forms the conclusion
is set as a problem -- to see whether it does not follow
from judgments already given, and through which a quite
different object is thought -- I look in the understanding for the
assertion of this conclusion, to discover whether it is not there
found to stand under certain conditions according to a universal
rule. If I find such a condition, and if the object of the
conclusion can be subsumed under the given condition, then
the conclusion is deduced from the rule, which is also valid for
other objects of knowledge. From this we see that in inference
reason endeavours to reduce the varied and manifold 
knowledge obtained through the understanding to the smallest
number of principles (universal conditions) and thereby to
achieve in it the highest possible unity.
P 305
C
The Pure Employment of Reason
Can we isolate reason, and is it, so regarded, an independent
source of concepts and judgments which spring from it
alone, and by means of which it relates to objects; or is it a
merely subordinate faculty, for imposing on given modes of
knowledge a certain form, called logical -- a faculty through
which what is known by means of the understanding is determined
in its interrelations, lower rules being brought under
higher (namely, those the condition of which includes in its
own sphere the condition of the lower), as far as this can be
done through [processes of] comparison? This is the question
with which we are now provisionally occupying ourselves. As
a matter of fact, multiplicity of rules and unity of principles is a
demand of reason, for the purpose of bringing the understanding
into thoroughgoing accordance with itself, just as the
understanding brings the manifold of intuition under concepts
and thereby connects the manifold. But such a principle does
not prescribe any law for objects, and does not contain any
general ground of the possibility of knowing or of determining
objects as such; it is merely a subjective law for the orderly
management of the possessions of our understanding, that by
comparison of its concepts it may reduce them to the smallest
possible number; it does not justify us in demanding from the
objects such uniformity as will minister to the convenience
and extension of our understanding; and we may not, therefore,
ascribe to the maxim any objective validity. In a word,
the question is, does reason in itself, that is, does pure reason,
contain a priori synthetic principles and rules, and in what
may these principles consist?
The formal and logical procedure of reason in syllogisms
gives us sufficient guidance as to the ground on which the
transcendental principle of pure reason in its synthetic 
knowledge will rest.
In the first place, reason in the syllogism does not concern
itself with intuitions, with a view to bringing them under rules
(as the understanding does with its categories), but with 
P 306
concepts and judgments. Accordingly, even if pure reason does
concern itself with objects, it has no immediate relation to
these and the intuition of them, but only to the understanding
and its judgments -- which deal at first hand with the senses
and their intuition for the purpose of determining their object.
The unity of reason is therefore not the unity of a possible
experience, but is essentially different from such unity, which
is that of understanding. That everything which happens has
a cause, is not a principle known and prescribed by reason.
That principle makes the unity of experience possible, and
borrows nothing from reason, which, apart from this relation
to possible experience, could never, from mere concepts, have
imposed any such synthetic unity.
Secondly, reason, in its logical employment, seeks to discover
the universal condition of its judgment (the conclusion),
and the syllogism is itself nothing but a judgment made by
means of the subsumption of its condition under a universal
rule (the major premiss). Now since this rule is itself subject
to the same requirement of reason, and the condition of the 
condition must therefore be sought (by means of a prosyllogism)
whenever practicable, obviously the principle peculiar to reason
in general, in its logical employment, is: -- to find for the 
conditioned knowledge obtained through the understanding the
unconditioned whereby its unity is brought to completion.
But this logical maxim can only become a principle of
pure reason through our assuming that if the conditioned is
given, the whole series of conditions, subordinated to one
another -- a series which is therefore itself unconditioned --
is likewise given, that is, is contained in the object and its
connection.
Such a principle of pure reason is obviously synthetic;
the conditioned is analytically related to some condition but
not to the unconditioned. From the principle there must also
follow various synthetic propositions, of which pure understanding
-- inasmuch as it has to deal only with objects of a
possible experience, the knowledge and synthesis of which is
always conditioned -- knows nothing. The unconditioned, if its
actuality be granted, is especially to be considered in respect
of all the determinations which distinguish it from whatever
P 307
is conditioned, and thereby must yield material for many
synthetic a priori propositions.
The principles arising from this supreme principle of
pure reason will, however, be transcendent in relation to all
appearances, i.e. there can never be any adequate empirical
employment of the principle. It will therefore be entirely
different from all principles of understanding, the employment
of which is wholly immanent, inasmuch as they have as their
theme only the possibility of experience. Take the principle,
that the series of conditions (whether in the synthesis of
appearances, or even in the thinking of things in general)
extends to the unconditioned. Does it, or does it not, have
objective applicability? What are its implications as regards
the empirical employment of understanding? Or is there
no such objectively valid principle of reason, but only
a logical precept, to advance towards completeness by an
ascent to ever higher conditions and so to give to our 
knowledge the greatest possible unity of reason? Can it be that
this requirement of reason has been wrongly treated in being
viewed as a transcendental principle of pure reason, and that
we have been overhasty in postulating such an unbounded
completeness of the series of conditions in the objects 
themselves? In that case, what other misunderstandings and 
delusions may have crept into the syllogisms, whose major 
premiss (perhaps rather an assumption than a postulate) is
derived from pure reason, and which proceed from experience
upwards to its conditions? To answer these questions will be
our task in the Transcendental Dialectic, which we shall now
endeavour to develop from its deeply concealed sources in
human reason. We shall divide the Dialectic into two chapters,
the first on the transcendent concepts of pure reason, the
second on its transcendent and dialectical syllogisms.
P 308
THE TRANSCENDENTAL DIALECTIC
BOOK I
THE CONCEPTS OF PURE REASON
WHATEVER we may have to decide as to the possibility of
the concepts derived from pure reason, it is at least true that
they are not to be obtained by mere reflection but only by
inference. Concepts of understanding are also thought a -
priori antecedently to experience and for the sake of experience,
but they contain nothing more than the unity of reflection
upon appearances, in so far as these appearances must
necessarily belong to a possible empirical consciousness.
Through them alone is knowledge and the determination of
an object possible. They first provide the material required
for making inferences, and they are not preceded by any
a priori concepts of objects from which they could be inferred.
On the other hand, their objective reality is founded solely
on the fact that, since they constitute the intellectual form of
all experience, it must always be possible to show their 
application in experience.
The title 'concept of reason' already gives a preliminary
indication that we are dealing with something which does
not allow of being confined within experience, since it 
concerns a knowledge of which any empirical knowledge (perhaps
even the whole of possible experience or of its empirical 
synthesis) is only a part. No actual experience has ever been 
completely adequate to it, yet to it every actual experience belongs.
Concepts of reason enable us to conceive, concepts of understanding
to understand -- ([as employed in reference to] perceptions).
If the concepts of reason contain the unconditioned, they
are concerned with something to which all experience is 
P 309
subordinate, but which is never itself an object of experience --
something to which reason leads in its inferences from 
experience, and in accordance with which it estimates and gauges
the degree of its empirical employment, but which is never
itself a member of the empirical synthesis. If, none the less,
these concepts possess objective validity, they may be called
conceptus ratiocinati (rightly inferred concepts); if, however,
they have no such validity, they have surreptitiously obtained
recognition through having at least an illusory appearance
of being inferences, and may be called conceptus ratiocinantes
(pseudo-rational concepts). But since this can be established
only in the chapter on the dialectical inferences of pure reason,
we are not yet in a position to deal with it. Meantime, just as
we have entitled the pure concepts of understanding categories,
so we shall give a new name to the concepts of pure
reason, calling them transcendental ideas. This title we shall
now explain and justify.
FIRST BOOK OF THE TRANSCENDENTAL
DIALECTIC
Section I
THE IDEAS IN GENERAL
Despite the great wealth of our languages, the thinker often
finds himself at a loss for the expression which exactly fits his
concept, and for want of which he is unable to be really intelligible
to others or even to himself. To coin new words is to advance
a claim to legislation in language that seldom succeeds;
and before we have recourse to this desperate expedient it is
advisable to look about in a dead and learned language, to see
whether the concept and its appropriate expression are not
already there provided. Even if the old-time usage of a term
should have become somewhat uncertain through the carelessness
of those who introduced it, it is always better to hold fast
to the meaning which distinctively belongs to it (even though
it remain doubtful whether it was originally used in precisely
this sense) than to defeat our purpose by making ourselves
unintelligible.
P 310
For this reason, if there be only a single word the established
meaning of which exactly agrees with a certain concept,
then, since it is of great importance that this concept be 
distinguished from related concepts, it is advisable to economise
in the use of the word and not to employ it, merely for the sake of
variety, as a synonym for some other expression, but carefully
to keep to its own proper meaning. Otherwise it may easily
happen that the expression ceasing to engage the attention in
one specific sense, and being lost in the multitude of other
words of very different meaning, the thought also is lost which
it alone could have preserved.
 Plato made use of the expression 'idea' in such a way as
quite evidently to have meant by it something which not only can
never be borrowed from the senses but far surpasses even the
concepts of understanding (with which Aristotle occupied himself),
inasmuch as in experience nothing is ever to be met with
that is coincident with it. For Plato ideas are archetypes
of the things themselves, and not, in the manner of the 
categories, merely keys to possible experiences. In his view they
have issued from highest reason, and from that source have
come to be shared in by human reason, which, however, is now
no longer in its original state, but is constrained laboriously to
recall, by a process of reminiscence (which is named 
philosophy), the old ideas, now very much obscured. I shall not
engage here in any literary enquiry into the meaning which
this illustrious philosopher attached to the expression. I need
only remark that it is by no means unusual, upon comparing
the thoughts which an author has expressed in regard to his
subject, whether in ordinary conversation or in writing, to find
that we understand him better than he has understood himself.
As he has not sufficiently determined his concept, he has 
sometimes spoken, or even thought, in opposition to his own 
intention. 
Plato very well realised that our faculty of knowledge feels
a much higher need than merely to spell out appearances 
according to a synthetic unity, in order to be able to read them
as experience. He knew that our reason naturally exalts itself
to forms of knowledge which so far transcend the bounds of
experience that no given empirical object can ever coincide
P 311
with them, but which must none the less be recognised as
having their own reality, and which are by no means mere
fictions of the brain.
Plato found the chief instances of his ideas in the field of
the practical, that is, in what rests upon freedom, which in its
turn rests upon modes of knowledge that are a peculiar product
of reason. Whoever would derive the concepts of virtue from
experience and make (as many have actually done) what at best
can only serve as an example in an imperfect kind of exposition,
into a pattern from which to derive knowledge, would
make of virtue something which changes according to time
and circumstance, an ambiguous monstrosity not admitting of
the formation of any rule. On the contrary, as we are well
aware, if anyone is held up as a pattern of virtue, the true
original with which we compare the alleged pattern and by
which alone we judge of its value is to be found only in our
minds. This original is the idea of virtue, in respect of which
the possible objects of experience may serve as examples
(proofs that what the concept of reason commands is in a 
certain degree practicable), but not as archetype. That no one
of us will ever act in a way which is adequate to what is contained
in the pure idea of virtue is far from proving this thought
to be in any respect chimerical. For it is only by means of this
idea that any judgment as to moral worth or its opposite is
possible; and it therefore serves as an indispensable foundation
for every approach to moral perfection -- however the
obstacles in human nature, to the degree of which there are no
assignable limits, may keep us far removed from its complete
achievement.
++ He also, indeed, extended his concept so as to cover speculative
knowledge, provided only the latter was pure and given completely
a priori. He even extended it to mathematics, although the
object of that science is to be found nowhere except in possible 
experience. In this I cannot follow him, any more than in his mystical
deduction of these ideas, or in the extravagances whereby he, so to
speak, hypostatised them -- although, as must be allowed, the exalted
language, which he employed in this sphere, is quite capable of a
milder interpretation that accords with the nature of things.
P 311
The Republic of Plato has become proverbial as a striking
example of a supposedly visionary perfection, such as can exist
P 312
only in the brain of the idle thinker; and Brucker has ridiculed
the philosopher for asserting that a prince can rule well only
in so far as he participates in the ideas. We should, however,
be better advised to follow up this thought, and, where the
great philosopher leaves us without help, to place it, through
fresh efforts, in a proper light, rather than to set it aside as 
useless on the very sorry and harmful pretext of impracticability.
A constitution allowing the greatest possible human freedom in
accordance with laws by which the freedom of each is made to
be consistent with that of all others -- I do not speak of the
greatest happiness, for this will follow of itself -- is at any rate
a necessary idea, which must be taken as fundamental not only
in first projecting a constitution but in all its laws. For at the
start we are required to abstract from the actually existing
hindrances, which, it may be, do not arise unavoidably out
of human nature, but rather are due to a quite remediable
cause, the neglect of the pure ideas in the making of the laws.
Nothing, indeed, can be more injurious, or more unworthy of
a philosopher, than the vulgar appeal to so-called adverse 
experience. Such experience would never have existed at all, if
at the proper time those institutions had been established in
accordance with ideas, and if ideas had not been displaced by
crude conceptions which, just because they have been derived
from experience, have nullified all good intentions. The more
legislation and government are brought into harmony with the
above idea, the rarer would punishments become, and it is therefore
quite rational to maintain, as Plato does, that in a perfect
state no punishments whatsoever would be required. This perfect
state may never, indeed, come into being; none the less
this does not affect the rightfulness of the idea, which, in order
to bring the legal organisation of mankind ever nearer to its
greatest possible perfection, advances this maximum as an
archetype. For what the highest degree may be at which
mankind may have to come to a stand, and how great a gulf
may still have to be left between the idea and its realisation,
are questions which no one can, or ought to, answer. For the
issue depends on freedom; and it is in the power of freedom
to pass beyond any and every specified limit.
P 313
But it is not only where human reason exhibits genuine
causality, and where ideas are operative causes (of actions and
their objects), namely, in the moral sphere, but also in regard
to nature itself, that Plato rightly discerns clear proofs of an
origin from ideas. A plant, an animal, the orderly arrangement
of the cosmos -- presumably therefore the entire natural world
-- clearly show that they are possible only according to ideas,
and that though no single creature in the conditions of its
individual existence coincides with the idea of what is most
perfect in its kind -- just as little as does any human being
with the idea of humanity, which he yet carries in his soul
as the archetype of his actions -- these ideas are none the
less completely determined in the Supreme Understanding,
each as an individual and each as unchangeable, and are
the original causes of things. But only the totality of things,
in their interconnection as constituting the universe, is 
completely adequate to the idea. If we set aside the 
exaggerations in Plato's methods of expression, the philosopher's
spiritual flight from the ectypal mode of reflecting upon the
physical world-order to the architectonic ordering of it 
according to ends, that is, according to ideas, is an enterprise
which calls for respect and imitation. It is, however, in regard
to the principles of morality, legislation, and religion, where
the experience, in this case of the good, is itself made possible
only by the ideas -- incomplete as their empirical expression
must always remain -- that Plato's teaching exhibits its quite
peculiar merits. When it fails to obtain recognition, this is due
to its having been judged in accordance with precisely those
empirical rules, the invalidity of which, regarded as principles,
it has itself demonstrated. For whereas, so far as nature is 
concerned, experience supplies the rules and is the source of
truth, in respect of the moral laws it is, alas, the mother of
illusion! Nothing is more reprehensible than to derive the laws
prescribing what ought to be done from what is done, or to
impose upon them the limits by which the latter is 
circumscribed. 
But though the following out of these considerations is
what gives to philosophy its peculiar dignity, we must meantime
occupy ourselves with a less resplendent, but still meritorious
P 314
task, namely, to level the ground, and to render it
sufficiently secure for moral edifices of these majestic 
dimensions. For this ground has been honeycombed by subterranean
workings which reason, in its confident but fruitless search
for hidden treasures, has carried out in all directions, and
which threaten the security of the superstructures. Our present
duty is to obtain insight into the transcendental employment
of pure reason, its principles and ideas, that we may be in a
position to determine and estimate its influence and true value.
Yet, before closing these introductory remarks, I beseech
those who have the interests of philosophy at heart (which is
more than is the case with most people) that, if they find
themselves convinced by these and the following considerations,
they be careful to preserve the expression 'idea' in
its original meaning, that it may not become one of those
expressions which are commonly used to indicate any and
every species of representation, in a happy-go-lucky confusion,
to the consequent detriment of science. There is no lack
of terms suitable for each kind of representation, that we
should thus needlessly encroach upon the province of any one
of them. Their serial arrangement is as follows. The genus
is representation in general (repraesentatio). Subordinate to
it stands representation with consciousness (perceptio). A
perception which relates solely to the subject as the modification
of its state is sensation (sensatio), an objective perception
is knowledge (cognitio). This is either intuition or concept
(intuitus vel conceptus). The former relates immediately to the
object and is single, the latter refers to it immediately by means
of a feature which several things may have in common. The
concept is either an empirical or a pure concept. The pure 
concept, in so far as it has its origin in the understanding alone
(not in the pure image of sensibility), is called a notion. A
concept formed from notions and transcending the possibility
of experience is an idea or concept of reason. Anyone who
has familiarised himself with these distinctions must find it
intolerable to hear the representation of the colour, red, called
an idea. It ought not even to be called a concept of 
understanding, a notion.
P 315
FIRST BOOK OF THE TRANSCENDENTAL
DIALECTIC
Section 2
THE TRANSCENDENTAL IDEAS
The Transcendental Analytic has shown us how the mere
logical form of our knowledge may in itself contain original
pure a priori concepts, which represent objects prior to all
experience, or, speaking more correctly, indicate the synthetic
unity which alone makes possible an empirical knowledge of
objects. The form of judgments (converted into a concept of
the synthesis of intuitions) yielded categories which direct all
employment of understanding in experience. Similarly, we
may presume that the form of syllogisms, when applied to
the synthetic unity of intuitions under the direction of the
categories, will contain the origin of special a priori concepts,
which we may call pure concepts of reason, or transcendental
ideas, and which will determine according to principles how
understanding is to be employed in dealing with experience
in its totality.
The function of reason in its inferences consists in the
universality of knowledge [which it yields] according to 
concepts, the syllogism being itself a judgment which is determined
a priori in the whole extent of its conditions. The proposition,
'Caius is mortal', I could indeed derive from experience
by means of the understanding alone. But I am in pursuit
of a concept (in this case, the concept 'man') that contains the
condition under which the predicate (general term for what
is asserted) of this judgment is given; and after I have 
subsumed the predicate under this condition taken in its whole
extension ('All men are mortal'), I proceed, in accordance
therewith, to determine the knowledge of my object ('Caius
is mortal').
Accordingly, in the conclusion of a syllogism we restrict a
P 316
predicate to a certain object, after having first thought it in
the major premiss in its whole extension under a given condition.
This complete quantity of the extension in relation to
such a condition is called universality (universalitas). In the
synthesis of intuitions we have corresponding to this the allness
(universitas) or totality of the conditions. The transcendental
concept of reason is, therefore, none other than the concept of
the totality of the conditions for any given conditioned. Now
since it is the unconditioned alone which makes possible the
totality of conditions, and, conversely, the totality of conditions
is always itself unconditioned, a pure concept of reason can
in general be explained by the concept of the unconditioned,
conceived as containing a ground of the synthesis of the
conditioned.
 The number of pure concepts of reason will be equal to the
number of kinds of relation which the understanding represents
to itself by means of the categories. We have therefore
to seek for an unconditioned, first, of the categorical synthesis
in a subject; secondly, of the hypothetical synthesis of the
members of a series; thirdly, of the disjunctive synthesis of the
parts in a system.
There is thus precisely the same number of kinds of syllogism,
each of which advances through prosyllogisms to the
unconditioned: first, to the subject which is never itself a 
predicate; secondly, to the presupposition which itself 
presupposes nothing further; thirdly, to such an aggregate of the
members of the division of a concept as requires nothing
further to complete the division. The pure concepts of reason
-- of totality in the synthesis of conditions -- are thus at least
necessary as setting us the task of extending the unity of
understanding, where possible, up to the unconditioned, and
are grounded in the nature of human reason. These 
transcendental concepts may, however, be without any suitable
corresponding employment in concreto, and may therefore
have no other utility than that of so directing the understanding
that, while it is extended to the uttermost, it is also at the
same time brought into complete consistency with itself.
 But while we are here speaking of the totality of conditions
and of the unconditioned, as being equivalent titles
for all concepts of reason, we again come upon an expression
P 317
with which we cannot dispense, and which yet, owing to an
ambiguity that attaches to it through long-standing misuse,
we also cannot with safety employ. The word 'absolute' is
one of the few words which in their original meaning were
adapted to a concept that no other word in the same language
exactly suits. Consequently its loss, or what amounts to the
same thing, looseness in its employment, must carry with it
the loss of the concept itself. And since, in this case, the 
concept is one to which reason devotes much of its attention,
it cannot be relinquished without greatly harming all 
transcendental philosophy. The word 'absolute' is now often used
merely to indicate that something is true of a thing considered
in itself, and therefore of its inward nature. In this sense the
absolutely possible would mean that which in itself (interne)
is possible -- which is, in fact, the least that can be said of an
object. On the other hand, the word is also sometimes used to
indicate that something is valid in all respects, without limitation,
e.g. absolute despotism, and in this sense the absolutely
possible would mean what is in every relation (in all respects)
possible -- which is the most that can be said of the possibility
of a thing. Now frequently we find these two meanings combined.
For example, what is internally impossible is impossible
in any relation, and therefore absolutely impossible. But in
most cases the two meanings are infinitely far apart, and I can
in no wise conclude that because something is in itself possible,
it is therefore also possible in every relation, and so absolutely
possible. Indeed, as I shall subsequently show, absolute 
necessity is by no means always dependent on inner necessity, and
must not, therefore, be treated as synonymous with it. If the
opposite of something is internally impossible, this opposite
is, of course, impossible in all respects, and the thing itself
is therefore absolutely necessary. But I cannot reverse the
reasoning so as to conclude that if something is absolutely
necessary its opposite is internally impossible, i.e. that the
absolute necessity of things is an inner necessity. For this
inner necessity is in certain cases a quite empty expression
to which we cannot attach any concept whatsoever, whereas
the concept of the necessity of a thing in all relations (to 
everything possible) involves certain quite special determinations.
P 318
Since the loss of a concept that is of great importance for
speculative science can never be a matter of indifference to the
philosopher, I trust that the fixing and careful preservation
of the expression, on which the concept depends, will 
likewise be not indifferent to him.
 It is, then, in this wider sense that I shall use the word
'absolute', opposing it to what is valid only comparatively,
that is, in some particular respect. For while the latter is 
restricted by conditions, the former is valid without restriction.
Now the transcendental concept of reason is directed
always solely towards absolute totality in the synthesis of 
conditions, and never terminates save in what is absolutely, that is,
in all relations, unconditioned. For pure reason leaves everything
to the understanding -- the understanding [alone] applying
immediately to the objects of intuition, or rather to their
synthesis in the imagination. Reason concerns itself exclusively
with absolute totality in the employment of the concepts of the
understanding, and endeavours to carry the synthetic unity.
which is thought in the category, up to the completely 
unconditioned. We may call this unity of appearances the unity of
reason, and that expressed by the category the unity of 
understanding. Reason accordingly occupies itself solely with the
employment of understanding, not indeed in so far as the latter
contains the ground of possible experience (for the concept of
the absolute totality of conditions is not applicable in any
experience, since no experience is unconditioned), but solely
in order to prescribe to the understanding its direction 
towards a certain unity of which it has itself no concept, and
in such manner as to unite all the acts of the understanding,
in respect of every object, into an absolute whole. The objective
employment of the pure concepts of reason is, therefore,
always transcendent, while that of the pure concepts of 
understanding must, in accordance with their nature, and inasmuch
as their application is solely to possible experience, be always
immanent.
I understand by idea a necessary concept of reason to
which no corresponding object can be given in sense-experience.
Thus the pure concepts of reason, now under consideration,
are transcendental ideas. They are concepts of pure
P 319
reason, in that they view all knowledge gained in experience as
being determined through an absolute totality of conditions.
They are not arbitrarily invented; they are imposed by the
very nature of reason itself, and therefore stand in necessary
relation to the whole employment of understanding. Finally,
they are transcendent and overstep the limits of all experience;
no object adequate to the transcendental idea can ever be
found within experience. If I speak of an idea, then as regards
its object, viewed as an object of pure understanding, I am
saying a great deal, but as regards its relation to the subject,
that is, in respect of its actuality under empirical conditions,
I am for the same reason saying very little, in that, as being
the concept of a maximum, it can never be correspondingly
given in concreto. Since in the merely speculative employment
of reason the latter [namely, to determine the actuality of the
idea under empirical conditions] is indeed our whole purpose,
and since the approximation to a concept, which yet is never
actually reached, puts us in no better position than if the concept
were entirely abortive, we say of such a concept -- it is only
an idea. The absolute whole of all appearances -- we might thus
say -- is only an idea; since we can never represent it in image,
it remains a problem to which there is no solution. But since,
on the other hand, in the practical employment of understanding,
our sole concern is with the carrying out of rules, the idea
of practical reason can always be given actually in concreto,
although only in part; it is, indeed, the indispensable condition
of all practical employment of reason. The practice of it is 
always limited and defective, but is not confined within 
determinable boundaries, and is therefore always under the influence
of the concept of an absolute completeness. The practical idea
is, therefore, always in the highest degree fruitful, and in its
relation to our actual activities is indispensably necessary.
Reason is here, indeed, exercising causality, as actually
bringing about that which its concept contains; and of such
wisdom we cannot, therefore, say disparagingly it is only an
idea. On the contrary, just because it is the idea of the necessary
unity of all possible ends, it must as an original, and at
least restrictive condition, serve as standard in all that bears
on the practical.
Although we must say of the transcendental concepts of
P 320
reason that they are only ideas, this is not by any means to be
taken as signifying that they are superfluous and void. For
even if they cannot determine any object, they may yet, in a
fundamental and unobserved fashion, be of service to the
understanding as a canon for its extended and consistent 
employment. The understanding does not thereby obtain more
knowledge of any object than it would have by means of its
own concepts, but for the acquiring of such knowledge it
receives better and more extensive guidance. Further -- what
we need here no more than mention -- concepts of reason may
perhaps make possible a transition from the concepts of
nature to the practical concepts, and in that way may give
support to the moral ideas themselves, bringing them into
connection with the speculative knowledge of reason. As to
all this, we must await explanation in the sequel.
In accordance with our plan we leave aside the practical
ideas, and consider reason only in its speculative, or rather,
restricting ourselves still further, only in its transcendental
employment. Here we must follow the path that we have
taken in the deduction of the categories; we must consider
the logical form of knowledge through reason, to see whether
perhaps reason may not thereby be likewise a source of concepts
which enable us to regard objects in themselves as determined
synthetically a priori, in relation to one or other of the
functions of reason.
 Reason, considered as the faculty of a certain logical form
of knowledge, is the faculty of inferring, i.e. judging mediately
(by the subsumption of the condition of a possible judgment
under the condition of a given judgment). The given
judgment is the universal rule (major premiss). The subsumption
of the condition of another possible judgment under the
condition of the rule is the minor premiss. The actual judgment
which applies the assertion of the rule to the subsumed
case is the conclusion. The rule states something universally,
subject to a certain condition. The condition of the rule is
found to be fulfilled in an actual case. What has been asserted
to be universally valid under that condition is therefore to be
regarded as valid also in the actual case, which involves that
condition. It is very evident, therefore, that reason arrives at
P 321
knowledge by means of acts of the understanding which constitute
a series of conditions. Thus if I arrive at the proposition
that all bodies are alterable, only by beginning with
the more remote knowledge (in which the concept of body
does not occur, but which nevertheless contains the condition
of that concept), namely, that everything composite is
alterable; if I then proceed from this to a proposition which
is less remote and stands under the condition of the last-
named proposition, namely, that bodies are composite; and if
from this I finally pass to a third proposition, which connects
the more remote knowledge (alterable) with the knowledge
actually before me, and so conclude that bodies are alterable
-- by this procedure I have arrived at knowledge (a conclusion)
by means of a series of conditions (the premisses).
Now every series the exponent of which is given (in categorical
or hypothetical judgment) can be continued; consequently
this same activity of reason leads to ratiocinatio polysyllogistica,
which is a series of inferences that can be prolonged
indefinitely on the side either of the conditions (per 
prosyllogismos) or of the conditioned (per episyllogismos).
But we soon become aware that the chain or series of 
prosyllogisms, that is, of inferred knowledge on the side of the
grounds or conditions of a given knowledge, in other words, of
the ascending series of syllogisms, must stand in a different
relation to the faculty of reason from that of the descending
series, that is, of the advance of reason in the direction of the
conditioned, by means of episyllogisms. For since in the former
case the knowledge (conclusio) is given only as conditioned, we
cannot arrive at it by means of reason otherwise than on the
assumption that all the members of the series on the side of the
conditions are given (totality in the series of the premisses);
only on this assumption is the judgment before us possible
a priori: whereas on the side of the conditioned, in respect of
consequences, we only think a series in process of becoming,
not one already presupposed or given in its completeness, and
therefore an advance that is merely potential. If, therefore,
knowledge be viewed as conditioned, reason is constrained to
regard the series of conditions in the ascending line as completed
and as given in their totality. But if the same knowledge
P 322
is viewed as a condition of yet other knowledge, and this 
knowledge as constituting a series of consequences in a descending
line, reason can be quite indifferent as to how far this
advance extends a parte posteriori, and whether a totality of
the series is possible at all. For it does not need such a series
in order to be able to draw its conclusion, this being already
sufficiently determined and secured by its grounds a parte -
priori. The series of premisses on the side of the conditions
may have a first member, as its highest condition, or it may
have no such member, in which case it is without limits a parte -
priori. But however this may be, and even admitting that we
can never succeed in comprehending a totality of conditions,
the series must none the less contain such a totality; and the
entire series must be unconditionally true if the conditioned,
which is regarded as a consequence resulting from it, is to be
counted as true. This is a requirement of reason, which 
announces its knowledge as being determined a priori and as
necessary, either in itself, in which case it needs no grounds,
or, if it be derivative, as a member of a series of grounds,
which itself, as a series, is unconditionally true.
 FIRST BOOK OF THE TRANSCENDENTAL
DIALECTIC
Section 3
SYSTEM OF THE TRANSCENDENTAL IDEAS
We are not at present concerned with logical dialectic,
which abstracts from all the content of knowledge and confines
itself to exposing the fallacies concealed in the form of
syllogisms, but with a transcendental dialectic which has to
contain, completely a priori, the origin of certain modes of
knowledge derived from pure reason as well as of certain
inferred concepts, the object of which can never be given 
empirically and which therefore lie entirely outside [the sphere
of ] the faculty of pure understanding. From the natural 
relation which the transcendental employment of our knowledge,
alike in inferences and in judgments, must bear to its logical
P 323
employment, we have gathered that there can be only three
kinds of dialectical inference, corresponding to the three
kinds of inference through which reason can arrive at knowledge
by means of principles, and that in all of these its business
is to ascend from the conditioned synthesis, to which
understanding always remains restricted, to the unconditioned,
which understanding can never reach.
The relations which are to be universally found in all our
representations are (1) relation to the subject; (2) relation to
objects, either as appearances or as objects of thought in
general. If we combine the subdivision with the main division,
all relation of representations, of which we can form either a
concept or an idea, is then threefold: (1) the relation to the
subject; (2) the relation to the manifold of the object in the
[field of] appearance; (3) the relation to all things in general.
Now all pure concepts in general are concerned with the
synthetic unity of representations, but [those of them which
are] concepts of pure reason (transcendental ideas) are concerned
with the unconditioned synthetic unity of all conditions
in general. All transcendental ideas can therefore be arranged in
three classes, the first containing the absolute (unconditioned)
unity of the thinking subject, the second the absolute unity of
the series of conditions of appearance, the third the absolute
unity of the condition of all objects of thought in general.
The thinking subject is the object of psychology, the sum-
total of all appearances (the world) is the object of cosmology,
and the thing which contains the highest condition of the
possibility of all that can be thought (the being of all beings)
the object of theology. Pure reason thus furnishes the idea
for a transcendental doctrine of the soul (psychologia 
rationalis), for a transcendental science of the world (cosmologia
rationalis), and, finally, for a transcendental knowledge of
God (theologia transzendentalis). The understanding is not in
a position to yield even the mere project of any one of these
sciences, not even though it be supported by the highest
logical employment of reason, that is, by all the conceivable
inferences through which we seek to advance from one
of its objects (appearance) to all others, up to the most remote
P 324
members of the empirical synthesis; each of these sciences is
an altogether pure and genuine product, or problem, of pure
reason.
In what precise modes the pure concepts of reason come
under these three headings of all transcendental ideas will be
fully explained in the next chapter. They follow the guiding-
thread of the categories. For pure reason never relates directly
to objects, but to the concepts which understanding frames
in regard to objects. Similarly it is only by the process of
completing our argument that it can be shown how reason,
simply by the synthetic employment of that very function of
which it makes use in categorical syllogisms, is necessarily
brought to the concept of the absolute unity of the thinking
subject, how the logical procedure used in hypothetical syllogisms
leads to the ideal of the completely unconditioned in a
series of given conditions, and finally how the mere form of
the disjunctive syllogism must necessarily involve the highest
concept of reason, that of a being of all beings -- a thought
which, at first sight, seems utterly paradoxical.
No objective deduction, such as we have been able to give
of the categories, is, strictly speaking, possible in the case of
these transcendental ideas. Just because they are only ideas
they have, in fact, no relation to any object that could be given
as coinciding with them. We can, indeed, undertake a subjective
derivation of them from the nature of our reason; and
this has been provided in the present chapter.
As is easily seen, what pure reason alone has in view is
the absolute totality of the synthesis on the side of the 
conditions (whether of inherence, of dependence, or of 
concurrence); it is not concerned with absolute completeness on the
side of the conditioned. For the former alone is required in
order to presuppose the whole series of the conditions, and
to present it a priori to the understanding. Once we are given
a complete (and unconditioned) condition, no concept of
reason is required for the continuation of the series; for every
step in the forward direction from the condition to the 
conditioned is carried through by the understanding itself. The
P 325
transcendental ideas thus serve only for ascending, in the
series of conditions, to the unconditioned, that is, to principles.
As regards the descending to the conditioned, reason does,
indeed, make a very extensive logical employment of the
laws of understanding, but no kind of transcendental employment;
and if we form an idea of the absolute totality of such
a synthesis (of the progressus), as, for instance, of the whole
series of all future alterations in the world, this is a creation of
the mind (ens rationis) which is only arbitrarily thought, and
not a necessary presupposition of reason. For the possibility
of the conditioned presupposes the totality of its conditions,
but not of its consequences. Such a concept is not, therefore,
one of the transcendental ideas; and it is with these alone that
we have here to deal.
Finally, we also discern that a certain connection and
unity is evident among the transcendental ideas themselves,
and that by means of them pure reason combines all its
modes of knowledge into a system. The advance from the
knowledge of oneself (the soul) to the knowledge of the world,
and by means of this to the original being, is so natural that it
seems to resemble the logical advance of reason from premisses
to conclusion.
++ Metaphysics has as the proper object
of its enquiries three ideas only: God, freedom, and immortality
-- so related that the second concept, when combined with the first,
should lead to the third as a necessary conclusion. Any other matters
with which this science may deal serve merely as a means of arriving
at these ideas and of establishing their reality. It does not need the
ideas for the purposes of natural science, but in order to pass beyond
nature. Insight into them would render theology and morals, and,
through the union of these two, likewise religion, and therewith the
highest ends of our existence, entirely and exclusively dependent on
the faculty of speculative reason. In a systematic representation of
the ideas, the order cited, the synthetic, would be the most suitable;
but in the investigation which must necessarily precede it the
analytic, or reverse order, is better adapted to the purpose of 
completing our great project, as enabling us to start from what is 
immediately given us in experience -- advancing from the doctrine of
the soul, to the doctrine of the world, and thence to the knowledge
of God.
P 325
Whether this is due to a concealed relationship
of the same kind as subsists between the logical and the 
transcendental procedure, is one of the questions that await answer
P 326
in the course of these enquiries. Indeed, we have already,
in a preliminary manner, obtained an answer to the question,
since in treating of the transcendental concepts of reason
which, in philosophical theory, are commonly confused with
others, and not properly distinguished even from concepts of
understanding, we have been able to rescue them from their
ambiguous position, to determine their origin, and at the
same time, in so doing, to fix their precise number (to which
we can never add), presenting them in a systematic connection,
and so marking out and enclosing a special field for pure
reason.