243 
PHILOSOPHY. 
and a priori given conceptions, the definition of an empi¬ 
rical conception is properly an explication by which we 
indicate what marks we have received from certain objeCts 
of experience; yet we do not deprive ourfelves of the 
liberty occafionally to receive more, and to difcard others. 
The definition of an d priori given conception might cor- 
reCtly be termed its expojition; by which we mean that 
its marks are completely yet obfcurely thought by the 
Underftanding, and that we are only confcious of fome 
of them ; they are consequently only arbitrarily-thought 
conceptions which are capable of a proper definition, and 
which indeed is all that is required. The enquiry after 
the objeClive validity of fuel', a conception is foreign to 
the definition, as I do not think by it a real but only an 
imaginary objedf. The mathematical conceptions have 
objedlive validity as Soon as they can be exhibited in the 
pure intuition. But, as to the conceptions which are 
compofed from empirical data, their definitions mult be 
termed Affumptions, in fo far as we are not yet certain 
that objedts really correfpond with them. With refpedt 
to definitions in philofophy, it follows, that we mull not 
imitate the mathematician, who always makes his defini¬ 
tion precede ; for in Philofophy we have to do with given 
conceptions, whofe definitions only arife after they have 
been Sufficiently contemplated. In the Mathematics, on 
the contrary, eve mull always begin with the definition, 
becaufe from it the conception itfelf arifes. Nor can 
mathematical definitions ever err ; their faults can only 
concern precifion, as they may repeat the fame marks, or 
give too many, and fuch as may be known as confequences 
& priori. Analytical definitions, on the other hand, may 
err in various ways; when they either fail in complete¬ 
ness, or, what is worfe, when they furnifh marks w'hich do 
not lie in the conception. 
Axioms are Synthetical pofitions h priori, in fo far as 
they are immediately certain. When we go out of a con¬ 
ception, in order to unite it with another Synthetically, a 
third conception mull be found, which produces this 
fynthefis. In the mathematics, it is the intuition in 
which certain determinations can be known as univer- 
fally-valid, becaufe it is a pure intuition. The certainty 
of thefe pofitions is therefore always intuitive. Now 
there arealfo indeed Synthetical pofitions & priori in Phi¬ 
lofophy ; but thefe cannot be called Axioms, becaufe they 
never can beconfidered as immediately certain, and becaufe 
we mull not merely refer to knowledge obtained by fyn- 
thelis, fince the fynthefis in this knowledge is a limita¬ 
tion of its validity. Thepofition, Whatever happens has a 
caij'e, cannot be termed an axiom, becaufe, in order to 
comprehend this, I do not view the conception Caufe as it 
were in pure intuition, and go back from this to the con¬ 
ception as to an unaltered lphere. The producing this 
lyiuhefis is a determination of time a priori, according to 
which I perceive that the experience of an event is polfible 
only fo far as I reprefent an empirical variety in relation 
to time as necefiarily connected. It is now obvious that 
the above pofition is only fo far a Tranfcendental Princi¬ 
ple, that is, a pofiticn applicable to objedts of experience 
alone, as it renders experience itfelf polfible; and that, 
with regard to the things in themfel ves, it has no objec¬ 
tive validity. Without this tranfcendental deduction, we 
fliould be in danger improperly to extend its meaning too 
tar, and to reft upon it, for inftance, in order to prove the 
reality of Tranfcendental Liberty. This is the cafe with 
all the fynthetical pofitions H priori, in fo far as our know¬ 
ledge ol them does not reft upon Intuition. They all of 
them require a deduction, and on that account cannot be 
confidered Axioms. 
Demonstrations are apodi&ical proofs in fo far as 
they are intuitive. Proofs by means of empirical intuitions 
are not demonlirations, becaufe they do not lead to apo- 
didlical certainty ; for I may indeed include in the con-r 
tjeption, according to analogy, that which I have found 
in the empirical intuition ; but I never can be apod iff ically 
certain that it applies alfo to all objeCts to whole fphere 
the particular cafe in the intuition belongs. Nor are 
philofophical proofs demonlirations : for, though they 
mull lead to apodiClical certainty, Hill the exhibition-by 
intuition, as well as the tranfition from the univerfal tq 
the particular, is contrary to the nature of this kind of 
knowledge. It always remains with the univerfal, and 
can never be carried on otherwife than by conceptions. 
Mathematical proofs are quite different from philofophi¬ 
cal ones. They proceed from the conception to the 
intuition, and conllruCl their conceptions. As they 
ground themfel ves on other pofitions, which have already 
been eltablilhed in this manner, and on the exhibition of 
the particular in the pure intuition, they difcover deter¬ 
minations which are valid for a whole fphere ; they go 
back from this exhibition of the particular to the concep¬ 
tion, and are in this manner able to add to it new deter¬ 
minations & priori. 
All direCt fynthetic apodi&ical pofitions, are divifible 
into Dogmas or Mathemas, whether they are demon- 
ltrable or immediately certain. Dire&ly fynthetical 
pofitions are thofe which are valid in the whole meaning 
of their conceptions ; but, if they are only knowabie 
from conceptions, then they are Dogmas; on the other 
hand, if they arife folely from the conllru&ion of con¬ 
ceptions, then they are Mathemas. Analytical pofitions 
cannot be called Dogmas, becaufe they do not extend the 
conception, but only explain it. No one calls mathema¬ 
tical pofitions Dogmas: it appears therefore that the 
common ufe of language confirms our definition. Now 
pure Reafon, in fo far as it is fpeculative, contains no 
direfl fynthetical pofition from conceptions: for, as to 
Ideas, there can be no fynthetical judgments that are 
objectively valid. It has the appearance, however, as if 
thofe pofitions which have been adduced under the title 
of principles of pure Underftanding, and whofe contents 
confift entirely of pure conceptions of Underftanding, are 
dire&ly fynthetical. It has been Ihown that they have no 
objedlive validity, in fo far as the Categories are taken in 
their original fignification, but that this can only then 
belong to them when thefe conceptions are determina¬ 
tions of Time h priori; as then fuch a fynthetical pofition 
is objectively valid, in fo far as it' reprefents the va¬ 
riety or matter of the empirical intuition with regard to 
Time as necefiarily connected by the Category, and con- 
fequently contains a rule for the application of the Cate¬ 
gories to empirical objefls. But, though it is a fynthe- 
tical and apodiClicaliy-certain pofition, it is not a direCl 
fynthetical pofition, and therefore no Dogma, becaufe it 
is not valid in the original meaning of its conception. 
Thefe pofitions are called Principles, and not Theorems ; 
not becaufe they cannot be proved, but becaufe they firit 
of all render the ground of demonftration itfelf, that is 
experience, polfible. The Dogmatical method, which 
would confift in conltruCting conceptions, and which 
declares this fynthefis without deriving it from tranfcen¬ 
dental grounds, as objectively valid, cannot be admitted 
in philofophy. 
SeCl. II. Difcipline of Pure Reafon with refpect to its Pole - 
mical Uj'e. 
By the polemical ufe of Reafon we mean the protection 
of its pofitions againll dogmatical attacks. When certain 
pofitions and their oppofites are apodi&ically proved, this 
is an antithetic of Pure Reafon. But fuch pomions, in 
reality, cannot occur, and the expofure of the illufion by 
which Reafon endeavours to maintain two oppoiite afier- 
tions mull certainly be polfible. Reafon meets with a 
contradiction, when it afeends to the higheft condition of 
the conditionally-thought objective unity of empirical 
intuition. The ground of this is, that Reafon in this cafe 
miitakes the unconditioned for an objeCt of intuition, 
and mull therefore necelfarily be in contradiction with 
itfelf, whether it conliders the unconditioned as a mem¬ 
ber of the feries, namely as the lalt, or takes the feries it¬ 
felf for abfolutely unconditioned. With refpeCt to the 
3 „ plychological 
