Immanuel Kant's

trans. by Norman Kemp Smith

Table of Contents Previous Section Next Section Search Engine

P 532
APPENDIX TO THE TRANSCENDENTAL DIALECTIC
THE REGULATIVE EMPLOYMENT OF THE IDEAS OF PURE
 REASON
The outcome of all dialectical attempts of pure reason
does not merely confirm what we have already proved in the
Transcendental Analytic, namely, that all those conclusions of
ours which profess to lead us beyond the field of possible 
experience are deceptive and without foundation; it likewise
teaches us this further lesson, that human reason has a natural
tendency to transgress these limits, and that transcendental
ideas are just as natural to it as the categories are to understanding
-- though with this difference, that while the categories
lead to truth, that is, to the conformity of our concepts with
the object, the ideas produce what, though a mere illusion,
is none the less irresistible, and the harmful influence of
which we can barely succeed in neutralising even by means
of the severest criticism.
Everything that has its basis in the nature of our powers
must be appropriate to, and consistent with, their right 
employment -- if only we can guard against a certain 
misunderstanding and so can discover the proper direction of these
powers. We are entitled, therefore, to suppose that 
transcendental ideas have their own good, proper, and therefore
immanent use, although, when their meaning is misunderstood,
and they are taken for concepts of real things, they
become transcendent in their application and for that very
reason can be delusive. For it is not the idea in itself, but its
use only, that can be either transcendent or immanent (that
is, either range beyond all possible experience or find 
employment within its limits), according as it is applied to an
object which is supposed to correspond to it, or is directed
solely to the use of understanding in general, in respect of
those objects that fall to be dealt with by the understanding.
All errors of subreption are to be ascribed to a defect
of judgment, never to understanding or to reason.
Reason is never in immediate relation to an object, but
P 533
only to the understanding; and it is only through the 
understanding that it has its own [specific] empirical employment.
It does not, therefore, create concepts (of objects) but only
orders them, and gives them that unity which they can have
only if they be employed in their widest possible application,
that is, with a view to obtaining totality in the various series.
The understanding does not concern itself with this totality,
but only with that connection through which, in accordance
with concepts, such series of conditions come into being.
Reason has, therefore, as its sole object, the understanding
and its effective application. Just as the understanding unifies
the manifold in the object by means of concepts, so reason
unifies the manifold of concepts by means of ideas, positing
a certain collective unity as the goal of the activities of the
understanding, which otherwise are concerned solely with
distributive unity.
I accordingly maintain that transcendental ideas never
allow of any constitutive employment. When regarded in
that mistaken manner, and therefore as supplying concepts
of certain objects, they are but pseudo-rational, merely 
dialectical concepts. On the other hand, they have an excellent,
and indeed indispensably necessary, regulative employment,
namely, that of directing the understanding towards a certain
goal upon which the routes marked out by all its rules converge,
as upon their point of intersection. This point is indeed
a mere idea, a focus imaginarius, from which, since it lies
quite outside the bounds of possible experience, the concepts
of the understanding do not in reality proceed; none the less
it serves to give to these concepts the greatest [possible] unity
combined with the greatest [possible] extension. Hence arises
the illusion that the lines have their source in a real object
lying outside the field of empirically possible knowledge -- just
as objects reflected in a mirror are seen as behind it. 
Nevertheless this illusion (which need not, however, be allowed
to deceive us) is indispensably necessary if we are to direct
the understanding beyond every given experience (as part of
the sum of possible experience), and thereby to secure its
greatest possible extension, just as, in the case of mirror-
vision, the illusion involved is indispensably necessary if,
P 534
besides the objects which lie before our eyes, we are also to
see those which lie at a distance behind our back.
If we consider in its whole range the knowledge obtained
for us by the understanding, we find that what is peculiarly
distinctive of reason in its attitude to this body of knowledge,
is that it prescribes and seeks to achieve its systematisation,
that is, to exhibit the connection of its parts in conformity
with a single principle. This unity of reason always presupposes
an idea, namely, that of the form of a whole of knowledge
-- a whole which is prior to the determinate knowledge
of the parts and which contains the conditions that determine
a priori for every part its position and relation to
the other parts. This idea accordingly postulates a complete
unity in the knowledge obtained by the understanding, by
which this knowledge is to be not a mere contingent aggregate,
but a system connected according to necessary laws. We may
not say that this idea is a concept of the object, but only of the
thoroughgoing unity of such concepts, in so far as that unity
serves as a rule for the understanding. These concepts of
reason are not derived from nature; on the contrary, we 
interrogate nature in accordance with these ideas, and consider
our knowledge as defective so long as it is not adequate to
them. By general admission, pure earth, pure water, pure air,
etc. , are not to be found. We require, however, the concepts of
them (though, in so far as their complete purity is concerned,
they have their origin solely in reason) in order properly to
determine the share which each of these natural causes has in
producing appearances. Thus in order to explain the chemical
interactions of bodies in accordance with the idea of a 
mechanism, every kind of matter is reduced to earths (qua mere
weight), to salts and inflammable substances (qua force), and
to water and air as vehicles (machines, as it were, by which
the first two produce their effects). The modes of expression
usually employed are, indeed, somewhat different; but the
influence of reason on the classifications of the natural
scientist is still easily detected.
If reason is a faculty of deducing the particular from the
universal, and if the universal is already certain in itself and
given, only judgment is required to execute the process of
P 535
subsumption, and the particular is thereby determined in a
necessary manner. This I shall entitle the apodeictic use of
reason. If, however, the universal is admitted as problematic
only, and is a mere idea, the particular is certain, but
the universality of the rule of which it is a consequence
is still a problem. Several particular instances, which are
one and all certain, are scrutinised in view of the rule, to
see whether they follow from it. If it then appears that all
particular instances which can be cited follow from the rule,
we argue to its universality, and from this again to all particular
instances, even to those which are not themselves given.
This I shall entitle the hypothetical employment of reason.
The hypothetical employment of reason, based upon ideas
viewed as problematic concepts, is not, properly speaking,
constitutive, that is, it is not of such a character that, judging
in all strictness, we can regard it as proving the truth of the
universal rule which we have adopted as hypothesis. For how
are we to know all the possible consequences which, as actually
following from the adopted principle, prove its universality?
The hypothetical employment of reason is regulative only; its
sole aim is, so far as may be possible, to bring unity into the
body of our detailed knowledge, and thereby to approximate
the rule to universality.
 The hypothetical employment of reason has, therefore, as
its aim the systematic unity of the knowledge of understanding,
and this unity is the criterion of the truth of its rules. The
systematic unity (as a mere idea) is, however, only a projected
unity, to be regarded not as given in itself, but as a problem
only. This unity aids us in discovering a principle for the
understanding in its manifold and special modes of employment,
directing its attention to cases which are not given, and
thus rendering it more coherent.
But the only conclusion which we are justified in drawing
from these considerations is that the systematic unity of the
manifold knowledge of understanding, as prescribed by reason,
is a logical principle. Its function is to assist the understanding
by means of ideas, in those cases in which the understanding
cannot by itself establish rules, and at the same time to give
P 536
to the numerous and diverse rules of the understanding unity
or system under a single principle, and thus to secure 
coherence in every possible way. But to say that the constitution
of the objects or the nature of the understanding which
knows them as such, is in itself determined to systematic
unity, and that we can in a certain measure postulate this
unity a priori, without reference to any such special interest
of reason, and that we are therefore in a position to maintain
that knowledge of the understanding in all its possible modes
(including empirical knowledge) has the unity required by
reason, and stands under common principles from which all
its various modes can, in spite of their diversity, be deduced
-- that would be to assert a transcendental principle of reason,
and would make the systematic unity necessary, not only
subjectively and logically, as method, but objectively also.
We may illustrate this by an instance of the employment
of reason. Among the various kinds of unity which conform
to the concepts of the understanding, is that of the causality
of a substance, which is called power. The various appearances
of one and the same substance show at first sight so
great a diversity, that at the start we have to assume just as
many different powers as there are different effects. For 
instance in the human mind we have sensation, consciousness,
imagination, memory, wit, power of discrimination,
pleasure, desire, etc. Now there is a logical maxim which
requires that we should reduce, so far as may be possible, this
seeming diversity, by comparing these with one another and
detecting their hidden identity. We have to enquire whether
imagination combined with consciousness may not be the same
thing as memory, wit, power of discrimination, and perhaps
even identical with understanding and reason. Though logic
is not capable of deciding whether a fundamental power
actually exists, the idea of such a power is the problem 
involved in a systematic representation of the multiplicity of
powers. The logical principle of reason calls upon us to bring
about such unity as completely as possible; and the more the
appearances of this and that power are found to be identical
with one another, the more probable it becomes that they are
simply different manifestations of one and the same power,
P 537
which may be entitled, relatively to the more specific powers,
the fundamental power. The same is done with the other
powers.
The relatively fundamental powers must in turn be compared
with one another, with a view to discovering their harmony,
and so to bring them nearer to a single radical, that
is, absolutely fundamental, power. But this unity of reason is
purely hypothetical. We do not assert that such a power must
necessarily be met with, but that we must seek it in the
interests of reason, that is, of establishing certain principles
for the manifold rules which experience may supply to us.
We must endeavour, wherever possible, to bring in this way
systematic unity into our knowledge.
On passing, however, to the transcendental employment
of understanding, we find that this idea of a fundamental
power is not treated merely as a problem for the hypothetical
use of reason, but claims to have objective reality, as postulating
the systematic unity of the various powers of a substance,
and as giving expression to an apodeictic principle of reason.
For without having made any attempt to show the harmony
of these various powers, nay, even after all attempts to do so
have failed, we yet presuppose that such a unity does actually
exist, and this not only, as in the case cited, on account of the
unity of the substance, but also in those cases in which, as with
matter in general, we encounter powers which, though to a
certain extent homogeneous, are likewise diverse. In all such
cases reason presupposes the systematic unity of the various
powers, on the ground that special natural laws fall under more
general laws, and that parsimony in principles is not only an
economical requirement of reason, but is one of nature's own
laws.
It is, indeed, difficult to understand how there can be a
logical principle by which reason prescribes the unity of rules,
unless we also presuppose a transcendental principle whereby
such a systematic unity is a priori assumed to be necessarily
inherent in the objects. For with what right can reason, in its
logical employment, call upon us to treat the multiplicity of
powers exhibited in nature as simply a disguised unity and
to derive this unity, so far as may be possible, from a fundamental
power -- how can reason do this, if it be free to admit
P 538
as likewise possible that all powers may be heterogeneous, and
that such systematic unity of derivation may not be in 
conformity with nature? Reason would then run counter to its own
vocation, proposing as its aim an idea quite inconsistent with
the constitution of nature. Nor can we say that reason, while
proceeding in accordance with its own principles, has arrived at
knowledge of this unity through observation of the accidental
constitution of nature. The law of reason which requires us to
seek for this unity, is a necessary law, since without it we should
have no reason at all, and without reason no coherent 
employment of the understanding, and in the absence of this no
sufficient criterion of empirical truth. In order, therefore, to
secure an empirical criterion we have no option save to presuppose
the systematic unity of nature as objectively valid and
necessary.
Although philosophers have not always acknowledged this
transcendental principle, even to themselves, or indeed been
conscious of employing it, we none the less find it covertly 
implied, in remarkable fashion, in the principles upon which they
proceed. That the manifold respects in which individual things
differ do not exclude identity of species, that the various species
must be regarded merely as different determinations of a few
genera, and these, in turn, of still higher genera, and so on; in
short, that we must seek for a certain systematic unity of all
possible empirical concepts, in so far as they can be deduced
from higher and more general concepts -- this is a logical
principle, a rule of the Schools, without which there could
be no employment of reason. For we can conclude from the
universal to the particular, only in so far as universal 
properties are ascribed to things as being the foundation upon
which the particular properties rest.
That such unity is to be found in nature, is presupposed by
philosophers in the well-known scholastic maxim, that rudiments
or principles must not be unnecessarily multiplied (entia
praeter necessitatem non esse multiplicanda). This maxim declares
that things by their very nature supply material for the
unity of reason, and that the seemingly infinite variety need
not hinder us from assuming that behind this variety there is
a unity of fundamental properties -- properties from which the
P 539
diversity can be derived through repeated determination. This
unity, although it is a mere idea, has been at all times so eagerly
sought, that there has been need to moderate the desire for it,
not to encourage it. A great advance was made when chemists
succeeded in reducing all salts to two main genera, acids and
alkalies; and they endeavour to show that even this difference
is merely a variety, or diverse manifestation, of one and the
same fundamental material. Chemists have sought, step by
step, to reduce the different kinds of earths (the material of
stones and even of metals) to three, and at last to two; but, not
content with this, they are unable to banish the thought that
behind these varieties there is but one genus, nay, that there
may even be a common principle for the earths and the salts.
It might be supposed that this is merely an economical contrivance
whereby reason seeks to save itself all possible trouble,
a hypothetical attempt, which, if it succeeds, will, through the
unity thus attained, impart probability to the presumed principle
of explanation. But such a selfish purpose can very easily
be distinguished from the idea. For in conformity with the
idea everyone presupposes that this unity of reason accords
with nature itself, and that reason -- although indeed unable
to determine the limits of this unity -- does not here beg but
command.
If among the appearances which present themselves to us,
there were so great a variety -- I do not say in form, for in that
respect the appearances might resemble one another; but in
content, that is, in the manifoldness of the existing entities --
that even the acutest human understanding could never by
comparison of them detect the slightest similarity (a possibility
which is quite conceivable), the logical law of genera
would have no sort of standing; we should not even have the
concept of a genus, or indeed any other universal concept; and
the understanding itself, which has to do solely with such 
concepts, would be non-existent. If, therefore, the logical principle
of genera is to be applied to nature (by which I here understand
those objects only which are given to us), it presupposes
a transcendental principle. And in accordance with this latter
principle, homogeneity is necessarily presupposed in the manifold
of possible experience (although we are not in a position
to determine in a priori fashion its degree); for in the absence
P 540
of homogeneity, no empirical concepts, and therefore no 
experience, would be possible.
The logical principle of genera, which postulates identity,
is balanced by another principle, namely, that of species,
which calls for manifoldness and diversity in things, 
notwithstanding their agreement as coming under the same genus,
and which prescribes to the understanding that it attend to the
diversity no less than to the identity. This principle (of 
discriminative observation, that is, of the faculty of distinction) sets
a limit to possible indiscretion in the former principle (of the
faculty of wit); and reason thus exhibits a twofold, self-conflicting
interest, on the one hand interest in extent (universality)
in respect of genera, and on the other hand in content 
(determinateness) in respect of the multiplicity of the species. In
the one case the understanding thinks more under its concepts,
in the other more in them. This twofold interest manifests itself
also among students of nature in the diversity of their ways
of thinking. Those who are more especially speculative are,
we may almost say, hostile to heterogeneity, and are always on
the watch for the unity of the genus; those, on the other hand,
who are more especially empirical, are constantly endeavouring
to differentiate nature in such manifold fashion as almost
to extinguish the hope of ever being able to determine its 
appearances in accordance with universal principles.
This latter mode of thought is evidently based upon a logical
principle which aims at the systematic completeness of all
knowledge -- prescribing that, in beginning with the genus, we
descend to the manifold which may be contained thereunder,
in such fashion as to secure extension for the system, just as in
the alternative procedure, that of ascending to the genus, we
endeavour to secure the unity of the system. For if we limit
our attention to the sphere of the concept which marks out a
genus, we can no more determine how far it is possible to proceed
in the [logical] division of it, than we can judge merely
from the space which a body occupies how far it is possible to
proceed in the [physical] division of its parts. Consequently,
P 541
every genus requires diversity of species, and these in turn
diversity of subspecies; and since no one of these subspecies is
ever itself without a sphere (extent as conceptus communis),
reason, in being carried to completion, demands that no
species be regarded as being in itself the lowest. For since the
species is always a concept, containing only what is common
to different things, it is not completely determined. It cannot,
therefore, be directly related to an individual, and other 
concepts, that is, subspecies, must always be contained under it.
This law of specification can be formulated as being the 
principle: entium varietates non temere esse minuendas.
But it is easily seen that this logical law would be without
meaning and application if it did not rest upon a transcendental
law of specification, which does not indeed demand an actual
infinity of differences in the things which can be objects to us
-- the logical principle, as affirming only the indeterminateness
of the logical sphere in respect of possible division, gives no
occasion for any such assertion -- but which none the less 
imposes upon the understanding the obligation of seeking under
every discoverable species for subspecies, and under every 
difference for yet smaller differences. For if there were no lower
concepts, there could not be higher concepts. Now the 
understanding can have knowledge only through concepts, and
therefore, however far it carries the process of division, never
through mere intuition, but always again through lower
concepts. The knowledge of appearances in their complete
determination, which is possible only through the understanding,
demands an endless progress in the specification of
our concepts, and an advance to yet other remaining differences,
from which we have made abstraction in the concept of
the species, and still more so in that of the genus.
This law of specification cannot be derived from experience,
which can never open to our view any such extensive
prospects. Empirical specification soon comes to a stop in the
distinction of the manifold, if it be not guided by the 
antecedent transcendental law of specification, which, as a principle
of reason, leads us to seek always for further differences,
and to suspect their existence even when the senses are unable
to disclose them. That absorbent earths are of different kinds
(chalk and muriatic earths), is a discovery that was possible
P 542
only under the guidance of an antecedent rule of reason -- reason
proceeding on the assumption that nature is so richly diversified
that we may presume the presence of such differences,
and therefore prescribing to the understanding the task of
searching for them. Indeed it is only on the assumption of
differences in nature, just as it is also only under the condition
that its objects exhibit homogeneity, that we can have any
faculty of understanding whatsoever. For the diversity of that
which is comprehended under a concept is precisely what gives
occasion for the employment of the concept and the exercise
of the understanding.
Reason thus prepares the field for the understanding: (1)
through a principle of the homogeneity of the manifold under
higher genera; (2) through a principle of the variety of the
homogeneous under lower species; and (3) in order to complete
the systematic unity, a further law, that of the affinity of all
concepts -- a law which prescribes that we proceed from each
species to every other by gradual increase of the diversity.
These we may entitle the principles of homogeneity, specification,
and continuity of forms. The last named arises from
union of the other two, inasmuch as only through the processes
of ascending to the higher genera and of descending to
the lower species do we obtain the idea of systematic connection
in its completeness. For all the manifold differences are
then related to one another, inasmuch as they one and all
spring from one highest genus, through all degrees of a more
and more widely extended determination.
The systematic unity, prescribed by the three logical
principles, can be illustrated in the following manner. Every
concept may be regarded as a point which, as the station for
an observer, has its own horizon, that is, a variety of things
which can be represented, and, as it were, surveyed from that
standpoint. This horizon must be capable of containing an
infinite number of points, each of which has its own narrower
horizon; that is, every species contains subspecies, according
to the principle of specification, and the logical horizon 
consists exclusively of smaller horizons (subspecies), never of
points which possess no extent (individuals). But for different
horizons, that is, genera, each of which is determined by its
own concept, there can be a common horizon, in reference to
P 543
which, as from a common centre, they can all be surveyed; and
from this higher genus we can proceed until we arrive at the
highest of all genera, and so at the universal and true horizon,
which is determined from the standpoint of the highest concept,
and which comprehends under itself all manifoldness --
genera, species, and subspecies.
We are carried to this highest standpoint by the law of
homogeneity, and to all lower standpoints, and their greatest
possible variety, by the law of specification. And since there
is thus no void in the whole sphere of all possible concepts,
and since nothing can be met with outside this sphere,
there arises from the presupposition of this universal horizon
and of its complete division, the principle: non datur vacuum
formarum, that is, that there are not different, original, first
genera, which are isolated from one another, separated, as it
were, by an empty intervening space; but that all the manifold
genera are simply divisions of one single highest and universal
genus. From this principle there follows, as its immediate 
consequence: datur continuum formarum, that is, that all 
differences of species border upon one another, admitting of no
transition from one to another per saltum, but only through
all the smaller degrees of difference that mediate between
them. In short, there are no species or subspecies which (in
the view of reason) are the nearest possible to each other; still
other intermediate species are always possible, the difference
of which from each of the former is always smaller than the
difference between these.
The first law thus keeps us from resting satisfied with an
excessive number of different original genera, and bids us pay
due regard to homogeneity; the second, in turn, imposes a
check upon this tendency towards unity, and insists that before
we proceed to apply a universal concept to individuals we
distinguish subspecies within it. The third law combines these
two laws by prescribing that even amidst the utmost 
manifoldness we observe homogeneity in the gradual transition
from one species to another, and thus recognise a relationship
of the different branches, as all springing from the same stem.
This logical law of the continuum specierum (formarum
logicarum) presupposes, however, a transcendental law (lex
P 544
continui in natura), without which the former law would only
lead the understanding astray, causing it to follow a path
which is perhaps quite contrary to that prescribed by nature
itself. This law must therefore rest upon pure transcendental,
not on empirical, grounds. For if it rested on empirical
grounds, it would come later than the systems, whereas in
actual fact it has itself given rise to all that is systematic in
our knowledge of nature. The formulation of these laws is
not due to any secret design of making an experiment, by
putting them forward as merely tentative suggestions. Such
anticipations, when confirmed, yield strong evidence in 
support of the view that the hypothetically conceived unity is
well-grounded; and such evidence has therefore in this respect
a certain utility. But it is evident that the laws contemplate
the parsimony of fundamental causes, the manifoldness
of effects, and the consequent affinity of the parts of nature
as being in themselves in accordance both with reason and
with nature. Hence these principles carry their recommendation
directly in themselves, and not merely as 
methodological devices.
But it is easily seen that this continuity of forms is a mere
idea, to which no congruent object can be discovered in experience.
For in the first place, the species in nature are actually
divided, and must therefore constitute a quantum discretum.
Were the advance in the tracing of their affinity continuous,
there would be a true infinity of intermediate members between
any two given species, which is impossible. And further,
in the second place, we could not make any determinate empirical
use of this law, since it instructs us only in quite general
terms that we are to seek for grades of affinity, and yields no
criterion whatsoever as to how far, and in what manner, we
are to prosecute the search for them.
 If we place these principles of systematic unity in the order
appropriate to their empirical employment, they will stand
thus: manifoldness, affinity, unity, each being taken, as an
idea, in the highest degree of its completeness. Reason 
presupposes the knowledge which is obtained by the understanding
and which stands in immediate relation to experience, and
P 545
seeks for the unity of this knowledge in accordance with ideas
which go far beyond all possible experience. The affinity of
the manifold (as, notwithstanding its diversity, coming under
a principle of unity) refers indeed to things, but still more to
their properties and powers. Thus, for instance, if at first our 
imperfect experience leads us to regard the orbits of the planets
as circular, and if we subsequently detect deviations therefrom,
we trace the deviations to that which can change the circle,
in accordance with a fixed law, through all the infinite 
intermediate degrees, into one of these divergent orbits. That is to
say, we assume that the movements of the planets which are
not circular will more or less approximate to the properties of a
circle; and thus we come upon the idea of an ellipse. Since the
comets do not, so far as observation reaches, return in any such
courses, their paths exhibit still greater deviations. What we
then do is to suppose that they proceed in a parabolic course,
which is akin to the ellipse, and which in all our observation
is indistinguishable from an ellipse that has its major axis 
indefinitely extended. Thus, under the guidance of these 
principles, we discover a unity in the generic forms of the orbits,
and thereby a unity in the cause of all the laws of planetary
motion, namely, gravitation. And we then extend our conquests
still further, endeavouring to explain by the same principle
all variations and seeming departures from these rules;
finally, we even go on to make additions such as experience
can never confirm, namely, to conceive, in accordance with
the rules of affinity, hyperbolic paths of comets, in the course
of which these bodies entirely leave our solar system, and
passing from sun to sun, unite the most distant parts of the
universe -- a universe which, though for us unlimited, is
throughout held together by one and the same moving force.
The remarkable feature of these principles, and what in
them alone concerns us, is that they seem to be transcendental,
and that although they contain mere ideas for the guidance of
the empirical employment of reason -- ideas which reason
follows only as it were asymptotically, i.e. ever more closely
without ever reaching them -- they yet possess, as synthetic
a priori propositions, objective but indeterminate validity, and
serve as rules for possible experience. They can also be 
employed with great advantage in the elaboration of experience,
P 546
as heuristic principles. A transcendental deduction of them
cannot, however, be effected; in the case of ideas, as we have
shown above, such a deduction is never possible.
In the Transcendental Analytic we have distinguished the
dynamical principles of the understanding, as merely regulative
principles of intuition, from the mathematical, which, as
regards intuition, are constitutive. None the less these dynamical
laws are constitutive in respect of experience, since they
render the concepts, without which there can be no experience,
possible a priori. But principles of pure reason can
never be constitutive in respect of empirical concepts; for since
no schema of sensibility corresponding to them can ever be
given, they can never have an object in concreto. If, then, we
disallow such empirical employment of them, as constitutive
principles, how are we to secure for them a regulative 
employment, and therewith some sort of objective validity, and
what can we mean by such regulative employment?
The understanding is an object for reason, just as sensibility
is for the understanding. It is the business of reason to
render the unity of all possible empirical acts of the understanding
systematic; just as it is of the understanding to connect
the manifold of the appearances by means of concepts,
and to bring it under empirical laws. But the acts of the 
understanding are, without the schemata of sensibility, undetermined;
just as the unity of reason is in itself undetermined, as
regards the conditions under which, and the extent to which,
the understanding ought to combine its concepts in systematic
fashion. But although we are unable to find in intuition a
schema for the complete systematic unity of all concepts of the
understanding, an analogon of such a schema must necessarily
allow of being given. This analogon is the idea of the maximum
in the division and unification of the knowledge of the
understanding under one principle. For what is greatest and
absolutely complete can be determinately thought, all 
restricting conditions, which give rise to an indeterminate
manifoldness, being left aside. Thus the idea of reason is
an analogon of a schema of sensibility; but with this difference,
that the application of the concepts of the understanding
to the schema of reason does not yield knowledge of the object
itself (as is the case in the application of categories to their
P 547
sensible schemata) but only a rule or principle for the 
systematic unity of all employment of the understanding. Now
since every principle which prescribes a priori to the 
understanding thoroughgoing unity in its employment, also holds,
although only indirectly, of the object of experience, the
principles of pure reason must also have objective reality
in respect of that object, not, however, in order to determine
anything in it, but only in order to indicate the procedure
whereby the empirical and determinate employment of the
understanding can be brought into complete harmony with
itself. This is acheived by bringing its employment, so far as
may be possible, into connection with the principle of thoroughgoing
unity, and by determining its procedure in the light of
this principle.
I entitle all subjective principles which are derived, not
from the constitution of an object but from the interest of
reason in respect of a certain possible perfection of the
knowledge of the object, maxims of reason. There are therefore
maxims of speculative reason, which rest entirely on its
speculative interest, although they may seem to be objective
principles.
When merely regulative principles are treated as constitutive,
and are therefore employed as objective principles, they
may come into conflict with one another. But when they
are treated merely as maxims, there is no real conflict, but
merely those differences in the interest of reason that give rise
to differing modes of thought. In actual fact, reason has only
one single interest, and the conflict of its maxims is only a
difference in, and a mutual limitation of, the methods 
whereby this interest endeavours to obtain satisfaction.
Thus one thinker may be more particularly interested in
manifoldness (in accordance with the principle of specification),
another thinker in unity (in accordance with the principle
of aggregation). Each believes that his judgment has
been arrived at through insight into the object, whereas it really
rests entirely on the greater or lesser attachment to one of the
two principles. And since neither of these principles is based
on objective grounds, but solely on the interest of reason, the
P 548
title 'principles' is not strictly applicable; they may more 
fittingly be entitled 'maxims'. When we observe intelligent people
disputing in regard to the characteristic properties of men,
animals, or plants -- even of bodies in the mineral realm -- some
assuming, for instance, that there are certain special hereditary
characteristics in each nation, certain well-defined inherited
differences in families, races, etc. , whereas others are bent upon
maintaining that in all such cases nature has made precisely
the same provision for all, and that it is solely to external
accidental conditions that the differences are due, we have
only to consider what sort of an object it is about which they
are making these assertions, to realise that it lies too deeply
hidden to allow of their speaking from insight into its nature.
The dispute is due simply to the twofold interest of reason,
the one party setting its heart upon, or at least adopting, the
one interest, and the other party the other. The differences
between the maxims of manifoldness and of unity in nature
thus easily allow of reconciliation. So long, however, as the
maxims are taken as yielding objective insight, and until a
way has been discovered of adjusting their conflicting claims,
and of satisfying reason in that regard, they will not only
give rise to disputes but will be a positive hindrance, and
cause long delays in the discovery of truth.
Similar observations are relevant in regard to the assertion
or denial of the widely discussed law of the continuous gradation
of created beings, which was propounded by Leibniz, and
admirably supported by Bonnet. It is simply the following
out of that principle of affinity which rests on the interest of
reason. For observation and insight into the constitution of
nature could never justify us in the objective assertion of the
law. The steps of this ladder, as they are presented to us in
experience, stand much too far apart; and what may seem to
us small differences are usually in nature itself such wide gaps,
that from any such observations we can come to no decision
in regard to nature's ultimate design -- especially if we bear in
mind that in so great a multiplicity of things there can never
be much difficulty in finding similarities and approximations.
On the other hand, the method of looking for order in nature
P 549
in accordance with such a principle, and the maxim which
prescribes that we regard such order -- leaving, however, 
undetermined where and how far -- as grounded in nature as
such, is certainly a legitimate and excellent regulative principle
of reason. In this regulative capacity it goes far beyond
what experience or observation can verify; and though not
itself determining anything, yet serves to mark out the path
towards systematic unity.
THE FINAL PURPOSE OF THE NATURAL DIALECTIC
 OF HUMAN REASON
The ideas of pure reason can never be dialectical in 
themselves; any deceptive illusion to which they give occasion
must be due solely to their misemployment. For they arise
from the very nature of our reason; and it is impossible that
this highest tribunal of all the rights and claims of speculation
should itself be the source of deceptions and illusions. 
Presumably, therefore, the ideas have their own good and 
appropriate vocation as determined by the natural disposition of
our reason. The mob of sophists, however, raise against reason
the usual cry of absurdities and contradictions, and though
unable to penetrate to its innermost designs, they none the less
inveigh against its prescriptions. Yet it is to the beneficent 
influences exercised by reason that they owe the possibility of
their own self-assertiveness, and indeed that very culture
which enables them to blame and to condemn what reason
requires of them.
We cannot employ an a priori concept with any certainty
without having first given a transcendental deduction of it.
The ideas of pure reason do not, indeed, admit of the kind of
deduction that is possible in the case of the categories. But if
they are to have the least objective validity, no matter how
indeterminate that validity may be, and are not to be mere
empty thought-entities (entia rationis ratiocinantis), a 
deduction of them must be possible, however greatly (as we admit)
it may differ from that which we have been able to give of the
categories. This will complete the critical work of pure reason,
and is what we now propose to undertake.
P 550
There is a great difference between something being given
to my reason as an object absolutely, or merely as an object in
the idea. In the former case our concepts are employed to determine
the object; in the latter case there is in fact only a schema
which no object, not even a hypothetical one, is directly
given, and which only enables us to represent to ourselves
other objects in an indirect manner, namely in their systematic
unity, by means of their relation to this idea. Thus I say that
the concept of a highest intelligence is a mere idea, that is
to say, its objective reality is not to be taken as consisting in
its referring directly to an object (for in that sense we should
not be able to justify its objective validity). It is only a schema
constructed in accordance with the conditions of the greatest
possible unity of reason -- the schema of the concept of a thing
in general, which serves only to secure the greatest possible 
systematic unity in the empirical employment of our reason. We
then, as it were, derive the object of experience from the 
supposed object of this idea, viewed as the ground or cause of the
object of experience. We declare, for instance, that the things
of the world must be viewed as if they received their existence
from a highest intelligence. The idea is thus really only a 
heuristic, not an ostensive concept. It does not show us how an
object is constituted, but how, under its guidance, we should
seek to determine the constitution and connection of the objects
of experience. If, then, it can be shown that the three transcendental
ideas (the psychological, the cosmological, and the theological),
although they do not directly relate to, or determine,
any object corresponding to them, none the less, as rules of the
empirical employment of reason, lead us to systematic unity,
under the presupposition of such an object in the idea; and
that they thus contribute to the extension of empirical 
knowledge, without ever being in a position to run counter to it,
we may conclude that it is a necessary maxim of reason to
proceed always in accordance with such ideas. This, indeed,
is the transcendental deduction of all ideas of speculative
reason, not as constitutive principles for the extension of our
knowledge to more objects than experience can give, but as
regulative principles of the systematic unity of the manifold
of empirical knowledge in general, whereby this empirical
P 551
knowledge is more adequately secured within its own Limits
and more effectively improved than would be possible, in the
absence of such ideas, through the employment merely of the
principles of the understanding.
I shall endeavour to make this clearer. In conformity with
these ideas as principles we shall, first, in psychology, under
the guidance of inner experience, connect all the appearances,
all the actions and receptivity of our mind, as if the mind were
a simple substance which persists with personal identity (in
this life at least), while its states, to which those of the body
belong only as outer conditions, are in continual change.
Secondly, in cosmology, we must follow up the conditions of
both inner and outer natural appearances, in an enquiry which
is to be regarded as never allowing of completion, just as if
the series of appearances were in itself endless, without any
first or supreme member. We need not, in so doing, deny that,
outside all appearances, there are purely intelligible grounds
of the appearances; but as we have no knowledge of these
whatsoever, we must never attempt to make use of them in our
explanations of nature. Thirdly, and finally, in the domain
of theology, we must view everything that can belong to the
context of possible experience as if this experience formed
an absolute but at the same time completely dependent and
sensibly conditioned unity, and yet also at the same time as if
the sum of all appearances (the sensible world itself) had a
single, highest and all-sufficient ground beyond itself, namely,
a self-subsistent, original, creative reason. For it is in the light
of this idea of a creative reason that we so guide the empirical
employment of our reason as to secure its greatest possible
extension -- that is, by viewing all objects as if they drew their
origin from such an archetype. In other words, we ought not
to derive the inner appearances of the soul from a simple
thinking substance but from one another, in accordance with
the idea of a simple being; we ought not to derive the order
and systematic unity of the world from a supreme intelligence,
but to obtain from the idea of a supremely wise cause the rule
according to which reason in connecting empirical causes and
effects in the world may be employed to best advantage, and in
such manner as to secure satisfaction of its own demands.
Now there is nothing whatsoever to hinder us from as-
P 552
suming these ideas to be also objective, that is, from 
hypostatising them -- except in the case of the cosmological ideas,
where reason, in so proceeding, falls into antinomy. The
psychological and theological ideas contain no antinomy,
and involve no contradiction. How, then, can anyone dispute
their [possible] objective reality? He who denies their possibility
must do so with just as little knowledge [of this possibility]
as we can have in affirming it. It is not, however, a
sufficient ground for assuming anything, that there is no
positive hindrance to our so doing; we are not justified in
introducing thought-entities which transcend all our concepts,
though without contradicting them, as being real and
determinate objects, merely on the authority of a speculative
reason that is bent upon completing the tasks which it has
set itself. They ought not to be assumed as existing in
themselves, but only as having the reality of a schema -- the
schema of the regulative principle of the systematic unity of
all knowledge of nature. They should be regarded only as
analoga of real things, not as in themselves real things. We
remove from the object of the idea the conditions which limit
the concept provided by our understanding, but which also
alone make it possible for us to have a determinate 
concept of anything. What we then think is a something of
which, as it is in itself, we have no concept whatsoever, but
which we none the less represent to ourselves as standing to
the sum of appearances in a relation analogous to that in
which appearances stand to one another.
If, in this manner, we assume such ideal beings, we do not
really extend our knowledge beyond the objects of possible
experience; we extend only the empirical unity of such 
experience, by means of the systematic unity for which the schema
is provided by the idea -- an idea which has therefore no claim
to be a constitutive, but only a regulative principle. For
to allow that we posit a thing, a something, a real being,
corresponding to the idea, is not to say that we profess
to extend our knowledge of things by means of transcendental
concepts. For this being is posited only in the idea and
not in itself; and therefore only as expressing the systematic
P 553
unity which is to serve as a rule for the empirical employment
of reason. It decides nothing in regard to the ground of
this unity or as to what may be the inner character of the being
on which as cause the unity depends.
Thus the transcendental, and the only determinate, concept
which the purely speculative reason gives us of God is, in
the strictest sense, deistic; that is, reason does not determine
the objective validity of such a concept, but yields only the
idea of something which is the ground of the highest and
necessary unity of all empirical reality. This something we
cannot think otherwise than on the analogy of a real substance
that, in conformity with laws of reason, is the cause
of all things. This, indeed, is how we must think it, in
so far as we venture to think it as a special object, and do
not rather remain satisfied with the mere idea of the 
regulative principle of reason, leaving aside the completion of
all conditions of thought as being too surpassingly great
for the human understanding. The latter procedure is, however,
inconsistent with the pursuit of that complete systematic
unity in our knowledge to which reason at least sets
no limits.
This, then, is how matters stand: if we assume a divine
being, we have indeed no concept whatsoever either of the
inner possibility of its supreme perfection or of the necessity
of its existence; but, on the other hand, we are in a position
to give a satisfactory answer to all those questions which
relate to the contingent, and to afford reason the most complete
satisfaction in respect to that highest unity after which
it is seeking in its empirical employment. The fact, however,
that we are unable to satisfy reason in respect to the assumption
itself, shows that it is the speculative interest of reason,
not any insight, which justifies it in thus starting from a point
that lies so far above its sphere; and in endeavouring, by this
device, to survey its objects as constituting a complete whole.
We here come upon a distinction bearing on the procedure
of thought in dealing with one and the same assumption, a
distinction which is somewhat subtle, but of great importance
in transcendental philosophy. I may have sufficient ground to
assume something, in a relative sense (suppositio relativa), and
yet have no right to assume it absolutely (suppositio absoluta).
P 554
This distinction has to be reckoned with in the case of a
merely regulative principle. We recognise the necessity of the
principle, but have no knowledge of the source of its necessity;
and in assuming that it has a supreme ground, we do so
solely in order to think its universality more determinately.
Thus, for instance, when I think as existing a being that
corresponds to a mere idea, indeed to a transcendental idea,
I have no right to assume any such thing as in itself existing,
since no concepts through which I am able to think any
object as determined suffice for such a purpose -- the conditions
which are required for the objective validity of my concepts
being excluded by the idea itself. The concepts of reality,
substance, causality, even that of necessity in existence, apart
from their use in making possible the empirical knowledge of
an object, have no meaning whatsoever, such as might serve
to determine any object. They can be employed, therefore, to
explain the possibility of things in the world of sense, but not
to explain the possibility of the universe itself. Such a ground
of explanation would have to be outside the world, and could
not therefore be an object of a possible experience. None the
less, though I cannot assume such an inconceivable being [as
existing] in itself, I may yet assume it as the object of a mere
idea, relatively to the world of sense. For if the greatest
possible empirical employment of my reason rests upon an
idea (that of systematically complete unity, which I shall
presently be defining more precisely), an idea which, although
it can never itself be adequately exhibited in experience,
is yet indispensably necessary in order that we may
approximate to the highest possible degree of empirical unity,
I shall not only be entitled, but shall also be constrained, to
realise this idea, that is, to posit for it a real object. But I may
posit it only as a something which I do not at all know in
itself, and to which, as a ground of that systematic unity, I
ascribe, in relation to this unity, such properties as are 
analogous to the concepts employed by the understanding in the
empirical sphere. Accordingly, in analogy with realities in
the world, that is, with substances, with causality and with
necessity, I think a being which possesses all this in the
highest perfection; and since this idea depends merely on
my reason, I can think this being as self-subsistent reason,
P 555
which through ideas of the greatest harmony and unity is
the cause of the universe. I thus omit all conditions which
might limit the idea, solely in order, under countenance of
such an original ground, to make possible systematic unity
of the manifold in the universe, and thereby the greatest
possible empirical employment of reason. This I do by 
representing all connections as if they were the ordinances of a
supreme reason, of which our reason is but a faint copy. I then
proceed to think this supreme being exclusively through 
concepts which, properly, are applicable only in the world of
sense. But since I make none but a relative use of the 
transcendental assumption, namely, as giving the substratum of
the greatest possible unity of experience, I am quite in order in
thinking a being which I distinguish from the world of sense,
through properties which belong solely to that world. For I
do not seek, nor am I justified in seeking, to know this object
of my idea according to what it may be in itself. There are no
concepts available for any such purpose; even the concepts of
reality, substance, causality, nay, even that of necessity in
existence, lose all meaning, and are empty titles for [possible]
concepts, themselves entirely without content, when we thus
venture with them outside the field of the senses. I think to
myself merely the relation of a being, in itself completely 
unknown to me, to the greatest possible systematic unity of the
universe, solely for the purpose of using it as a schema of the
regulative principle of the greatest possible empirical 
employment of my reason.
If it be the transcendental object of our idea that we have
in view, it is obvious that we cannot thus, in terms of the
concepts of reality, substance, causality, etc. , presuppose its
reality in itself, since these concepts have not the least application
to anything that is entirely distinct from the world of sense.
The supposition which reason makes of a supreme being, as
the highest cause, is, therefore relative only; it is devised solely
for the sake of systematic unity in the world of sense, and is a
mere something in idea, of which, as it may be in itself, we
have no concept. This explains why, in relation to what is
given to the senses as existing, we require the idea of a 
primordial being necessary in itself, and yet can never form the
slightest concept of it or of its absolute necessity.
P 556
We are now in a position to have a clear view of the outcome
of the whole Transcendental Dialectic, and accurately to define
the final purpose of the ideas of pure reason, which become
dialectical only through heedlessness and misapprehension.
Pure reason is in fact occupied with nothing but itself. It can
have no other vocation. For what is given to it does not consist
in objects that have to be brought to the unity of the empirical
concept, but in those modes of knowledge supplied by the
understanding that require to be brought to the unity of the
concept of reason -- that is, to unity of connection in conformity
with a principle. The unity of reason is the unity of system;
and this systematic unity does not serve objectively as a principle
that extends the application of reason to objects, but subjectively
as a maxim that extends its application to all possible
empirical knowledge of objects. Nevertheless, since the 
systematic connection which reason can give to the empirical 
employment of the understanding not only furthers its extension,
but also guarantees its correctness, the principle of such 
systematic unity is so far also objective, but in an indeterminate
manner (principium vagum). It is not a constitutive principle
that enables us to determine anything in respect of its direct
object, but only a merely regulative principle and maxim, to
further and strengthen in infinitum (indeterminately) the
empirical employment of reason -- never in any way proceeding
counter to the laws of its empirical employment, and yet
at the same time opening out new paths which are not within
the cognisance of the understanding.
 But reason cannot think this systematic unity otherwise
than by giving to the idea of this unity an object; and since
experience can never give an example of complete systematic
unity, the object which we have to assign to the idea is not
such as experience can ever supply. This object, as thus 
entertained by reason (ens rationis ratiocinatae), is a mere idea;
it is not assumed as a something that is real absolutely and
in itself, but is postulated only problematically (since we
cannot reach it through any of the concepts of the 
understanding) in order that we may view all connection of the
things of the world of sense as if they had their ground in such
a being. In thus proceeding, our sole purpose is to secure
that systematic unity which is indispensable to reason, and
P 557
which while furthering in every way the empirical knowledge
obtainable by the understanding can never interfere to hinder
or obstruct it.
We misapprehend the meaning of this idea if we regard
it as the assertion or even as the assumption of a real thing,
to which we may proceed to ascribe the ground of the systematic
order of the world. On the contrary, what this ground
which eludes our concepts may be in its own inherent constitution
is left entirely undetermined; the idea is posited only
as being the point of view from which alone that unity, which
is so essential to reason and so beneficial to the understanding,
can be further extended. In short, this transcendental
thing is only the schema of the regulative principle by which
reason, so far as lies in its power, extends systematic unity
over the whole field of experience.
The first object of such an idea is the 'I' itself, viewed
simply as thinking nature or soul. If I am to investigate the
properties with which a thinking being is in itself endowed, I
must interrogate experience. For I cannot even apply any one
of the categories to this object, except in so far as the schema
of the category is given in sensible intuition. But I never thereby
attain to a systematic unity of all appearances of inner sense.
Instead, then, of the empirical concept (of that which the soul
actually is), which cannot carry us far, reason takes the concept
of the empirical unity of all thought; and by thinking this unity
as unconditioned and original, it forms from it a concept of
reason, that is, the idea of a simple substance, which, unchangeable
in itself (personally identical), stands in association with
other real things outside it; in a word, the idea of a simple self-
subsisting intelligence. Yet in so doing it has nothing in view
save principles of systematic unity in the explanation of the
appearances of the soul. It is endeavouring to represent all
determinations as existing in a single subject, all powers, so
far as possible, as derived from a single fundamental power, all
change as belonging to the states of one and the same permanent
being, and all appearances in space as completely different
from the actions of thought. The simplicity and other
properties of substance are intended to be only the schema of
this regulative principle, and are not presupposed as being the
actual ground of the properties of the soul. For these may rest
P 558
on altogether different grounds, of which we can know nothing.
The soul in itself could not be known through these assumed
predicates, not even if we regarded them as absolutely valid
in respect of it. For they constitute a mere idea which cannot
be represented in concreto. Nothing but advantage can result
from the psychological idea thus conceived, if only we take
heed that it is not viewed as more than a mere idea, and that
it is therefore taken as valid only relatively to the systematic
employment of reason in determining the appearances of our
soul. For no empirical laws of bodily appearance, which are
of a totally different kind, will then intervene in the 
explanation of what belongs exclusively to inner sense. No windy
hypotheses of generation, extinction, and palingenesis of souls
will be permitted. The consideration of this object of inner
sense will thus be kept completely pure and will not be 
confused by the introduction of heterogeneous properties. Also,
reason's investigations will be directed to reducing the grounds
of explanation in this field, so far as may be possible, to a
single principle. All this will be best attained through such a
schema, viewed as if it were a real being; indeed it is 
attainable in no other way. The psychological idea can signify
nothing but the schema of a regulative concept. For were
I to enquire whether the soul in itself is of spiritual nature,
the question would have no meaning. In employing such a
concept I not only abstract from corporeal nature, but from
nature in general, that is, from all predicates of any possible
experience, and therefore from all conditions requisite for
thinking an object for such a concept; yet only as related to
an object can the concept be said to have a meaning.
The second regulative idea of merely speculative reason
is the concept of the world in general. For nature is properly
the only given object in regard to which reason requires 
regulative principles. This nature is twofold, either thinking or
corporeal. To think the latter, so far as regards its inner
possibility, that is, to determine the application of the 
categories to it, we need no idea, that is, no representation which
transcends experience. Nor, indeed, is any idea possible in this
connection, since in dealing with corporeal nature we are
guided solely by sensible intuition. The case is different from
that of the fundamental psychological concept ('I'), which
P 559
contains a priori a certain form of thought, namely, the unity
of thought. There therefore remains for pure reason nothing
but nature in general, and the completeness of the conditions
in nature in accordance with some principle. The absolute
totality of the series of these conditions, in the derivation of
their members, is an idea which can never be completely
realised in the empirical employment of reason, but which
yet serves as a rule that prescribes how we ought to proceed
in dealing with such series, namely, that in explaining appearances,
whether in their regressive or in their ascending order,
we ought to treat the series as if it were in itself infinite, that
is, as if it proceeded in indefinitum. When, on the other hand,
reason is itself regarded as the determining cause, as in [the
sphere of] freedom, that is to say, in the case of practical 
principles, we have to proceed as if we had before us an object, not
of the senses, but of the pure understanding. In this practical
sphere the conditions are no longer in the series of appearances;
they can be posited outside the series, and the series of
states can therefore be regarded as if it had an absolute 
beginning, through an intelligible cause. All this shows that the
cosmological ideas are nothing but simply regulative principles,
and are very far from positing, in the manner of constitutive
principles, an actual totality of such series. The fuller
treatment of this subject will be found in the chapter on the
antinomy of pure reason.
The third idea of pure reason, which contains a merely
relative supposition of a being that is the sole and sufficient
cause of all cosmological series, is the idea of God. We have
not the slightest ground to assume in an absolute manner (to
suppose in itself) the object of this idea; for what can enable
us to believe in or assert a being of the highest perfection and
one absolutely necessary by its very nature, merely on the basis
of its concept, or if we did how could we justify our procedure?
It is only by way of its relation to the world that we can attempt
to establish the necessity of this supposition; and it then becomes
evident that the idea of such a being, like all speculative ideas,
seeks only to formulate the command of reason, that all 
connection in the world be viewed in accordance with the principles
of a systematic unity -- as if all such connection had its
source in one single all-embracing being, as the supreme and
P 560
all-sufficient cause. It is thus evident that reason has here no
other purpose than to prescribe its own formal rule for the
extension of its empirical employment, and not any extension
beyond all limits of empirical employment. Consequently it is
evident that this idea does not, in any concealed fashion, 
involve any principle that claims, in its application to possible
experience, to be constitutive in character.
This highest formal unity, which rests solely on concepts
of reason, is the purposive unity of things. The speculative
interest of reason makes it necessary to regard all order in the
world as if it had originated in the purpose of a supreme
reason. Such a principle opens out to our reason, as applied
in the field of experience, altogether new views as to how the
things of the world may be connected according to teleological
laws, and so enables it to arrive at their greatest systematic
unity. The assumption of a supreme intelligence, as the one
and only cause of the universe, though in the idea alone, can
therefore always benefit reason and can never injure it. Thus
if, in studying the shape of the earth (which is round, but somewhat
flattened), of the mountains, seas, etc. , we assume it to be
the outcome of wise purposes on the part of an Author of the
world, we are enabled to make in this way a number of 
discoveries. And provided we restrict ourselves to a merely 
regulative use of this principle, even error cannot do us any serious
harm. For the worst that can happen would be that where we
expected a teleological connection (nexus finalis), we find only
a mechanical or physical connection (nexus effectivus). In such
a case, we merely fail to find the additional unity; we do not
destroy the unity upon which reason insists in its empirical employment.
++ The advantage arising from the spherical shape of the earth
is well known. But few are aware that its spheroidal flattening alone
prevents the continental elevations, or even the smaller hills, thrown
up perhaps by earthquakes, from continuously, and indeed quite
appreciably in a comparatively short time, altering the position of
the axis of the earth. The protuberance of the earth at the equator
forms so vast a mountain that the impetus of all the other mountains
can never produce any observable effect in changing the position
of the earth's axis. And yet, wise as this arrangement is, we feel
no scruples in explaining it from the equilibrium of the formerly
fluid mass of the earth.
P 561
But even a disappointment of this sort cannot
affect the teleological law itself, in its general bearing. For
although an anatomist can be convicted of error when he
assigns to some member of an animal body an end which
it can be clearly shown not to subserve, it is yet quite 
impossible to prove in any given case that an arrangement
of nature, be it what it may, subserves no end whatsoever.
Accordingly, medical physiology extends its very limited 
empirical knowledge of the ends served by the articulation of an
organic body, by resorting to a principle for which pure reason
has alone been responsible; and it carries this principle so far as
to assume confidently, and with general approval, that everything
in an animal has its use, and subserves some good purpose.
If this assumption be treated as constitutive it goes much
further than observation has thus far been able to justify; and
we must therefore conclude that it is nothing more than a
regulative principle of reason, to aid us in securing the highest
possible systematic unity, by means of the idea of the purposive
causality of the supreme cause of the world -- as if this
being, as supreme intelligence, acting in accordance with a
supremely wise purpose, were the cause of all things.
If, however, we overlook this restriction of the idea to a
merely regulative use, reason is led away into mistaken paths.
For it then leaves the ground of experience, which alone can
contain the signs that mark out its proper course, and ventures
out beyond it to the incomprehensible and unsearchable,
rising to dizzy heights where it finds itself entirely cut off
from all possible action in conformity with experience.
The first error which arises from our using the idea of a
supreme being in a manner contrary to the nature of an idea,
that is, constitutively, and not regulatively only, is the error of
ignava ratio.
++ This was the title given by the ancient dialecticians to a
sophistical argument, which ran thus: If it is your fate to recover
from this illness, you will recover, whether you employ a physician
or not. Cicero states that this mode of argument has been so named,
because, if we conformed to it, reason would be left without any use
in life. On the same ground I apply the name also to the sophistical
argument of pure reason.
P 561
We may so entitle every principle which makes
P 562
us regard our investigation into nature, on any subject, as
absolutely complete, disposing reason to cease from further
enquiry, as if it had entirely succeeded in the task which it had
set itself. Thus the psychological idea, when it is employed as
a constitutive principle to explain the appearances of our soul,
and thereby to extend our knowledge of the self beyond the
limits of experience (its state after death), does indeed simplify
the task of reason; but it interferes with, and entirely ruins,
our use of reason in dealing with nature under the guidance
of our experiences. The dogmatic spiritualist explains the
abiding and unchanging unity of a person throughout all
change of state, by the unity of the thinking substance, of
which, as he believes, he has immediate perception in the 'I';
or he explains the interest which we take in what can happen
only after our death, by means of our consciousness of the 
immaterial nature of the thinking subject; and so forth. He thus
dispenses with all empirical investigation of the cause of these
inner appearances, so far as that cause is to be found in physical
grounds of explanation; and to his own great convenience,
though at the sacrifice of all real insight, he professes, in 
reliance upon the assumed authority of a transcendent reason, to
have the right to ignore those sources of knowledge which are
immanent in experience. These detrimental consequences are
even more obvious in the dogmatic treatment of our idea of a
supreme intelligence, and in the theological system of nature
(physico-theology) which is falsely based upon it. For in
this field of enquiry, if instead of looking for causes in the
universal laws of material mechanism, we appeal directly to
the unsearchable decree of supreme wisdom, all those ends
which are exhibited in nature, together with the many ends
which are only ascribed by us to nature, make our investigation
of the causes a very easy task, and so enable us to
regard the labour of reason as completed, when, as a matter
of fact, we have merely dispensed with its employment -- an
employment which is wholly dependent for guidance upon the
order of nature and the series of its alterations, in accordance
with the universal laws which they are found to exhibit. This
error can be avoided, if we consider from the teleological point
of view not merely certain parts of nature, such as the distribution
P 563
of land, its structure, the constitution and location of
the mountains, or only the organisation of the vegetable and
animal kingdoms, but make this systematic unity of nature
completely universal, in relation to the idea of a supreme 
intelligence. For we then treat nature as resting upon a 
purposiveness, in accordance with universal laws, from which no
special arrangement is exempt, however difficult it may be to
establish this in any given case. We then have a regulative
principle of the systematic unity of teleological connection --
a connection which we do not, however, predetermine. What
we may presume to do is to follow out the physico-mechanical
connection in accordance with universal laws in the hope of
discovering what the teleological connection actually is. In this
way alone can the principle of purposive unity aid always in
extending the employment of reason in reference to experience
without being in any instance prejudicial to it.
The second error arising from the misapprehension of the
above principle of systematic unity is that of perversa ratio
(husteron proteron). The idea of systematic unity should be
used only as a regulative principle to guide us in seeking for
such unity in the connection of things, according to universal
laws of nature; and we ought, therefore, to believe that we
have approximated to completeness in the employment of the
principle only in proportion as we are in a position to verify
such unity in empirical fashion -- a completeness which is
never, of course, attainable. Instead of this the reverse 
procedure is adopted. The reality of a principle of purposive
unity is not only presupposed but hypostatised; and since the
concept of a supreme intelligence is in itself completely 
beyond our powers of comprehension, we proceed to determine
it in an anthropomorphic manner, and so to impose ends
upon nature, forcibly and dictatorially, instead of pursuing
the more reasonable course of searching for them by the path
of physical investigation. And thus teleology, which is 
intended to aid us merely in completing the unity of nature in
accordance with universal laws, not only tends to abrogate
such unity, but also prevents reason from carrying out its own
professed purpose, that of proving from nature, in conformity
with these laws, the existence of a supreme intelligent cause.
P 564
For if the most complete purposiveness cannot be presupposed
a priori in nature, that is, as belonging to its essence, how can
we be required to search for it, and through all its gradations
to approximate to the supreme perfection of an Author of all
things, a perfection that, as absolutely necessary, must be
knowable a priori?  The regulative principle prescribes that
systematic unity as a unity in nature, which is not known
merely empirically but is presupposed a priori (although in
an indeterminate manner), be presupposed absolutely, and
consequently as following from the essence of things. If,
however, I begin with a supreme purposive being as the
ground of all things, the unity of nature is really surrendered,
as being quite foreign and accidental to the nature of things,
and as not capable of being known from its own universal laws.
There then arises a vicious circle; we are assuming just that
very point which is mainly in dispute.
To take the regulative principle of the systematic unity of
nature as being a constitutive principle, and to hypostatise, and
presuppose as a cause, that which serves, merely in idea, as the
ground of the consistent employment of reason, is simply to
confound reason. The investigation of nature takes its own
independent course, keeping to the chain of natural causes
in conformity with their universal laws. It does indeed, in so
doing, proceed in accordance with the idea of an Author of the
universe, but not in order to deduce therefrom the purposiveness
for which it is ever on the watch, but in order to obtain
knowledge of the existence of such an Author from this 
purposiveness. And by seeking this purposiveness in the essence
of the things of nature, and so far as may be possible in the
essence of things in general, it seeks to know the existence of
this supreme being as absolutely necessary. Whether this latter
enterprise succeed or not, the idea remains always true in itself,
and justified in its use, provided it be restricted to the 
conditions of a merely regulative principle.
Complete purposive unity constitutes what is, in the absolute
sense, perfection. If we do not find this unity in the
essence of the things which go to constitute the entire object of
experience, that is, of all our objectively valid knowledge, and
therefore do not find it in the universal and necessary laws of
nature, how can we profess to infer directly from this unity the
P 565
idea of a supreme and absolutely necessary perfection of an
original being, as the source of all causality? The greatest 
possible systematic unity, and consequently also purposive unity, is
the training school for the use of reason, and is indeed the very
foundation of the possibility of its greatest possible employment.
The idea of such unity is, therefore, inseparably bound
up with the very nature of our reason. This same idea is on
that account legislative for us; and it is therefore very natural
that we should assume a corresponding legislative reason
(intellectus archetypus), from which, as the object of our reason,
all systematic unity of nature is to be derived.
In discussing the antinomy of pure reason we have stated
that the questions propounded by pure reason must in every
case admit of an answer, and that in their regard it is not 
permissible to plead the limits of our knowledge (a plea which
in many questions that concern nature is as unavoidable as
it is relevant). For we are not here asking questions in regard
to the nature of things, but only such questions as arise from
the very nature of reason, and which concern solely its own
inner constitution. We are now in a position to confirm this
assertion -- which at first sight may have appeared rash -- so
far as regards the two questions in which pure reason is most
of all interested; and thus finally to complete our discussion of
the dialectic of pure reason.
If, in connection with a transcendental theology, we ask,
first, whether there is anything distinct from the world, which
contains the ground of the order of the world and of its 
connection in accordance with universal laws, the answer is that
there undoubtedly is. For the world is a sum of appearances;
and there must therefore be some transcendental ground of
the appearances, that is, a ground which is thinkable only by
the pure understanding. If, secondly, the question be, whether
this being is substance, of the greatest reality, necessary, etc. ,
++ After what I have already said regarding the psychological
idea and its proper vocation, as a principle for the merely regulative
employment of reason, I need not dwell at any length upon the
transcendental illusion by which the systematic unity of all the 
manifoldness of inner sense is hypostatised. The procedure is very similar
to that which is under discussion in our criticism of the theological
ideal.
P 566
we reply that this question is entirely without meaning. For
all categories through which we can attempt to form a concept
of such an object allow only of empirical employment, and
have no meaning whatsoever when not applied to objects of
possible experience, that is, to the world of sense. Outside this
field they are merely titles of concepts, which we may admit,
but through which [in and by themselves] we can understand
nothing. If, thirdly, the question be, whether we may not at
least think this being, which is distinct from the world, in
analogy with the objects of experience, the answer is: certainly,
but only as object in idea and not in reality, namely,
only as being a substratum, to us unknown, of the systematic
unity, order, and purposiveness of the arrangement of the
world -- an idea which reason is constrained to form as the
regulative principle of its investigation of nature. Nay, more,
we may freely, without laying ourselves open to censure, admit
into this idea certain anthropomorphisms which are helpful
to the principle in its regulative capacity. For it is always an
idea only, which does not relate directly to a being distinct
from the world, but to the regulative principle of the systematic
unity of the world, and only by means of a schema of this
unity, namely, through the schema of a supreme intelligence
which, in originating the world, acts in accordance with wise
purposes. What this primordial ground of the unity of the
world may be in itself, we should not profess to have thereby
decided, but only how we should use it, or rather its idea, in
relation to the systematic employment of reason in respect of
the things of the world.
But the question may still be pressed: Can we, on such
grounds, assume a wise and omnipotent Author of the world?
Undoubtedly we may; and we not only may, but must, do so.
But do we then extend our knowledge beyond the field of possible
experience? By no means. All that we have done is merely
to presuppose a something, a merely transcendental object, of
which, as it is in itself, we have no concept whatsoever. It is
only in relation to the systematic and purposive ordering of
the world, which, if we are to study nature, we are constrained
to presuppose, that we have thought this unknown being
by analogy with an intelligence (an empirical concept); that
is, have endowed it, in respect of the ends and perfection
P 567
which are to be grounded upon it, with just those properties
which, in conformity with the conditions of our reason, can
be regarded as containing the ground of such systematic unity.
This idea is thus valid only in respect of the employment of our
reason in reference to the world. If we ascribed to it a validity
that is absolute and objective, we should be forgetting that
what we are thinking is a being in idea only; and in thus taking
our start from a ground which is not determinable through
observation of the world, we should no longer be in a position
to apply the principle in a manner suited to the empirical
employment of reason.
But, it will still be asked, can I make any such use of the
concept and of the presupposition of a supreme being in the
rational consideration of the world? Yes, it is precisely for
this purpose that reason has resorted to this idea. But may I
then proceed to regard seemingly purposive arrangements as
purposes, and so derive them from the divine will, though,
of course, mediately through certain special natural means,
themselves established in furtherance of that divine will? Yes,
we can indeed do so; but only on condition that we regard
it as a matter of indifference whether it be asserted that
divine wisdom has disposed all things in accordance with its
supreme ends, or that the idea of supreme wisdom is a regulative
principle in the investigation of nature and a principle
of its systematic and purposive unity, in accordance with
universal laws, even in those cases in which we are unable
to detect that unity. In other words, it must be a matter of
complete indifference to us, when we perceive such unity,
whether we say that God in his wisdom has willed it to be
so, or that nature has wisely arranged it thus. For what has
justified us in adopting the idea of a supreme intelligence as
a schema of the regulative principle is precisely this greatest
possible systematic and purposive unity -- a unity which our
reason has required as a regulative principle that must underlie
all investigation of nature. The more, therefore, we discover
purposiveness in the world, the more fully is the legitimacy
of our idea confirmed. But since the sole aim of that
principle was to guide us in seeking a necessary unity of nature,
and that in the greatest possible degree, while we do indeed,
P 568
in so far as we attain that unity, owe it to the idea of a supreme
being, we cannot, without contradicting ourselves, ignore the
universal laws of nature -- with a view to discovering which the
idea was alone adopted -- and look upon this purposiveness
of nature as contingent and hyperphysical in its origin. For we
were not justified in assuming above nature a being with those
qualities, but only in adopting the idea of such a being in order
to view the appearances as systematically connected with one
another in accordance with the principle of a causal 
determination. 
For the same reasons, in thinking the cause of the world,
we are justified in representing it in our idea not only in
terms of a certain subtle anthropomorphism (without which
we could not think anything whatsoever in regard to it),
namely, as a being that has understanding, feelings of pleasure
and displeasure, and desires and volitions corresponding to
these, but also in ascribing to it a perfection which, as infinite,
far transcends any perfection that our empirical knowledge of
the order of the world can justify us in attributing to it. For
the regulative law of systematic unity prescribes that we should
study nature as if systematic and purposive unity, combined
with the greatest possible manifoldness, were everywhere to be
met with, in infinitum. For although we may succeed in discovering
but little of this perfection of the world, it is nevertheless
required by the legislation of our reason that we must
always search for and surmise it; and it must always be 
beneficial, and can never be harmful, to direct our investigations
into nature in accordance with this principle. But it is evident
that in this way of representing the principle as involving the
idea of a supreme Author, I do not base the principle upon the
existence and upon the knowledge of such a being, but upon
its idea only, and that I do not really derive anything from this
being, but only from the idea of it -- that is, from the nature of
the things of the world, in accordance with such an idea. A
certain, unformulated consciousness of the true use of this
concept of reason seems indeed to have inspired the modest
and reasonable language of the philosophers of all times,
since they speak of the wisdom and providence of nature and
of divine wisdom, just as if nature and divine wisdom were
P 569
equivalent expressions -- indeed, so long as they are dealing
solely with speculative reason, giving preference to the former
mode of expression, on the ground that it enables us to avoid
making profession of more than we are justified in asserting,
and that it likewise directs reason to its own proper field,
namely, nature.
Thus pure reason, which at first seemed to promise nothing
less than the extension of knowledge beyond all limits of 
experience, contains, if properly understood, nothing but regulative
principles, which, while indeed prescribing greater unity
than the empirical employment of understanding can achieve,
yet still, by the very fact that they place the goal of its
endeavours at so great a distance, carry its agreement with
itself, by means of systematic unity, to the highest possible
degree. But if, on the other hand, they be misunderstood,
and be treated as constitutive principles of transcendent
knowledge, they give rise, by a dazzling and deceptive
illusion, to persuasion and a merely fictitious knowledge,
and therewith to contradictions and eternal disputes.
***
Thus all human knowledge begins with intuitions, proceeds
from thence to concepts, and ends with ideas. Although
in respect of all three elements it possesses a priori sources of
knowledge, which on first consideration seem to scorn the
limits of all experience, a thoroughgoing critique convinces us
that reason, in its speculative employment, can never with
these elements transcend the field of possible experience, and
that the proper vocation of this supreme faculty of knowledge
is to use all methods, and the principles of these methods,
solely for the purpose of penetrating to the innermost secrets
of nature, in accordance with every possible principle of unity
-- that of ends being the most important -- but never to soar
beyond its limits, outside which there is for us nothing but
empty space. The critical examination, as carried out in the
Transcendental Analytic, of all propositions which may seem
to extend our knowledge beyond actual experience, has doubtless
sufficed to convince us that they can never lead to anything
more than a possible experience. Were it not that we are
suspicious of abstract and general doctrines, however clear,
P 570
and were it not that specious and alluring prospects tempt us
to escape from the compulsion which these doctrines impose,
we might have been able to spare ourselves the laborious 
interrogation of all those dialectical witnesses that a transcendent
reason brings forward in support of its pretensions. For
we should from the start have known with complete certainty
that all such pretensions, while perhaps honestly meant, must
be absolutely groundless, inasmuch as they relate to a kind
of knowledge to which man can never attain. But there is no
end to such discussions, unless we can penetrate to the true
cause of the illusion by which even the wisest are deceived.
Moreover, the resolution of all our transcendent knowledge
into its elements (as a study of our inner nature) is in itself
of no slight value, and to the philosopher is indeed a matter
of duty. Accordingly, fruitless as are all these endeavours of
speculative reason, we have none the less found it necessary
to follow them up to their primary sources. And since the
dialectical illusion does not merely deceive us in our judgments,
but also, because of the interest which we take in these
judgments, has a certain natural attraction which it will always
continue to possess, we have thought it advisable, with a view
to the prevention of such errors in the future, to draw up in
full detail what we may describe as being the records of this
lawsuit, and to deposit them in the archives of human reason.

Table of Contents Previous Section Next Section Search Engine