6.

The Highest Principle and the Principle

of Origin in Hermann Cohen's
Theoretical Philosophy

Jurgen Stolzenberg

If one wants to characterize the philosophical oeuvre of Hermann Cohen

in a way that does justice to both its philosophical-historical meaning and
Cohen's self-understanding, then it should be described as working through
a double task. This task consists, on the one hand, in a defense of Kant's phi-
losophy against interpretations it had received in the course of its rediscov-
ery, after the post-Kantian idealistic systems lost their persuasiveness. On
the other hand, the task consists in a rehabilitation of Kant's philosophy in
order to contribute to the re-grounding of philosophy as a theory of sci-
ence. Indeed, from the beginning Cohen characterized and legitimated his
undertaking of Kantian interpretation as "a combination of the systematic
and the historical task" (Cohen 1987b, iv).
This assertion of Cohen's, however, is itself in need of interpreta-
tion. Already in the preface to his first draft of a systematic reconstruc-
tion of Kant's theoretical philosophy-the first edition of Kants 7heoric
der Erfahrung (Kanis Theory of Experience)-Cohen emphatically called
attention to this connection between the historical and systematic tasks as
a connection which, for its part, stands under the signature of "systematic
partisanship" (Cohen 1987b, v). This also means that the return to the origi-
nal Kantian text propagated by Cohen, and the announced rebuttal of the
"adversaries" of Kant "through simple citations" (Cohen 1987b, iv) from the
textual inventory of the Critique of Pure Reason, could only derive their
sense and force from a systematic interpretation with respect to the entirety
of the Kantian critique of reason.
Cohen placed this interpretation under the thesis, "Kant discovered a
new concept of experience" (Cohen 1987b, 3). The first edition of Kants 7heo-
rie der Erfahrung clarifies, although not until its end, what this new concept
vf experiences in 1 it consists in the totality of synthetic

132
propositions, which form the content of mathematics and the pure natural
sciences. 2 The peculiarity of this reduction of the Kantian notion of experi+
ence, for which the relation to sense-data given in intuition is constitutive (a
relation that is obviously excluded by Cohen), and which is also constitutive
for the theoretical program of the Kant ian critique of reason as a whole, is
obvious. That is to say, if the field of objects to which philosophy is to cor-
respond is limited at the outset to the mathematical-natural sciences, then
Kant's central concern for a critique of traditional metaphysics must lose all
meaning and all interest. These restrictions, however, are no less dear than
the reasons that Cohen gave for this purpose. They are precisely the reasons
with which he sought to defend Kant's philosophy against its critics and
"antagonists;' especially those who came from the field of psychology:'
Cohen's critique has its basis in the thesis that psychology is not in the
position to accept the rank of an ultimately ground-laying (letztbegrunden-
den) theory, due to its status as an empirical science; for, as an empirical sci-
ence, it may attain only hypothetically ultimate elements of consciousness.
Since this is the case, psychology cannot be recognized as the foundation
for the interpretation of an epistemology which aims at an ultimate ground-
ing (Letztbegriindung), as Cohen's interpretation construes Kant's theory of
knowledge.'
It is crucial for Cohen's reconstruction of Kant that he targeted his cri-
tique not only against the psycho logistic interpretation of Kant, but against
Kant himself. His underlying thesis is that Kant's theory of knowledge is,
in its central elements-such as the "transcendental deduction of the pure
concepts of the understanding" -merely an empirical-psychological analy-
sis of the synthesis of knowledge from subjective conditions. Hence, what
Cohen deems to be Kant's goal-the grounding of the necessity and strict
universality of mathematical- natural scientific knowledge from a priori con-
ditions-in principle could not be achieved.' From this, Cohen has drawn
the conclusion that is decisive for his overall approach in reconstructing
Kant's philosophy, namely, to consider the "synthetic doctrine" -which Kant
had posited for the Critique of Pure Reason-as a psychological-empirical
method of reconstruction of the genesis of cognition from its subjective con-
ditions. Thus, this synthetic method has to be shunned as inappropriate for
the enterprise of an objective reconstruction of Kant's theoretical philoso-
phy. An "analytic" or "transcendental method" has to replace it. This novel
method, which Cohen construes consciously and "critically" against Kant's
"synthetic doctrine;' and which attempts not to create "on its own, the sci-
ence of the objects of nature" (Cohen 1987a, 577), must instead start out from
science as a ''factum" and must lay bare the a priori conditions on which
the validity of its cognitions rest. 6 Cohen saw these conditions given in the
embodiment of those synthetic a priori laws which Kant had characterized

HERMANN COHEN'S THEORETICAL PHILOSOPHY • l.B

as "principles of pure reason;' and which form the basis of mathematics and
pure natural science. With the exclusion of a theoretical ground-
ing of the objectivity of knowledge, Cohen also had to reject as merely psy-
chological the Kantian definition of the concept of the a priori as the origin
of a concept from the sources of reason. The character of the a priori for
Cohen became equivalent with that of necessity and universality, i.e., with
the lawfulness of cognition. And since Cohen wanted to understand natural-
scientific knowledge alone as "experience;' it becomes understandable that
Cohen conceives of the principles contained in mathematics and pure natu-
ral sciences as the principles of experience thus construed.
'!he "transcendental method;' as Cohen termed it, through which these
principles ought to be sought, is not, however, without problems. These
problems will be demonstrated and clarified in the first section, through
their connection to Cohen's conception of a highest principle in the second
edition of Kanis 7heorie der Erfahrung. In the second section, we will pur-
sue the question of the way in which a continuity exists between Cohen's
conception of a highest principle and the central principle for the Logik der
reinen Erkenntnis (Logic of Pure Cognition)-namely, the principle of ori-
gin. We shall pursue this question with the intention of addressing a misun-
derstanding suggested by Cohen, and at the same time shall contribute to a
controversy that exists in Cohen-research with respect to the interpretation
of the principle of origin.

The Highest Prindple and the Transcendental Conditions of Cognition

In the following, we shall not discuss a misconstrual of Cohen's self-under-
standing-namely, that he supposedly linked his "transcendental method"
too closely with the "analytic method" which Kant had pursued in the
Prolegomena and also, as Cohen thought, in the Critique of Pure Reason.'
Rather, we shall call attention to a problem that arises from the concept of
the analytical method itself.' This method starts out from a proposition that
is accepted as certain and as given, and inquires into the premises from
which its truth follows. With respect to the demonstration of the premises,
one can initially say that more than one foundation can serve as a sufficient
condition. Concerning the truth of these conditions, it is also the case that
none of these conditions must be true simply because their consequences
are true, since true propositions can follow from false premises. Hence it
follows, concerning the truth of the propositions assumed as given, or of
a set of propositions assumed as given, that their truth can in no way he
proven through the analytic method, unless the truth of the premises is
guaranteed independently from their grounding capacity. If this is not the
case, the premises must be taken as hypotheses. With respect to the truth of

134 • J0RGEN STOLZENBERG

the premises, it follows that it is not based on their grounding capacity for
true propositions, but rather that it is to be presupposed.
Hence, a charge leveled against Cohen by Friedrich Albert Lange becomes
understandable: If the truth of the premises would be claimed under the
condition that the premises are the explanatory ground for the presupposed
experience of science9 -and Lange imputes this claim to Cohen-then a
mere tautology would be asserted, or the argument would be circular. For
then the ground for the truth of a condition of the possibility of experience
would consist in that it is a condition of the possibility of experience."
Cohen replied to this objection in two passages-once in the context
of his explication of the transcendental method in Kants Begriindung der
Ethik (Kants Grounding of Ethics);" and again, without referring to Lange by
name, in the second edition of Kants 7heorie der Erfahrung, in the context
of his explication of the highest principle of the transcendental conditions
of cognition. The argumentation in this regard will be investigated more
closely in what follows.
Cohen's explications of a highest principle of the conditions of cognition
proceed from the question about the systematic role to be assigned to pure
spatial intuition as one of the conditions of the apriority of the principles
of mathematics with respect to the whole of the "transcendental system:·
Cohen clarifies this role with reference to the fact that the explication of
valid a priori conditions for the principles of mathematics must not be seen
in isolation, but rather with regard to a "total unified science:· mathemat-
ical-natural science, whose basis is mathematics (Cohen 1987a, 137). Each
individual transcendental condition is to be construed, therefore, not as "an
encapsulated whole" but only "as part and parcel of a higher whole"; this
entirety presents the "totality of transcendental conditions which, as a unity,
hold together the unity of science" (Cohen 1987a, 138). Cohen understands
this relation of part and whole as that of species and genus. The concept of
a highest principle therefore interprets the concept of the genus of the tran-
scendental conditions of natural-scientific cognition. Cohen has given this
concept several titles, such as the "highest principle of all synthetic judg-
ments" or also the highest "principle of experience.""
It is clear that in this way merely the function of a highest principle is
designated, and not its content. So it remains to be asked: In what does the
commonality of the transcendental conditions of the cognition of experi-
ence consist? Cohen has linked his answer to this question with his answer
to the charge of circularity formulated by Lange. A variation of Lange's
accusation is formulated by Cohen as follows:

The appearance could arise that we are going in circles. We seek that which
makes possible experience or synthetic judgments as necessary, and believe

HERMANN COHEN's THEORETICAL PHILOSOPHY a 135

 that we must assert the same in a highest law as principle; this seems to
contain merely a riddle, without containing a hint for the solution. We ask:
What makes experience possible? And the answer is: the highest principle
of experience. (Cohen 1987a, 138-139)

The argumentation with which Cohen seeks to dissolve this concern

demands special attention. At first it seems as if Cohen is not preparing to
refute the accusation of a circular justification of the possibility of experi-
ence. For he starts the argument with the question: What makes possible the
highest principle itself! And his answer, "nothing other than itself" (Cohen
1987a, 139), seems to contain no discernable response to the objection of
circularity and seems to leave this objection unchanged.
Yet if one follows Cohen's argument for his answer, then something
becomes apparent that is striking and also in need of explanation: Cohen's
explication of the content of the highest principle, from which one should
be able to discern its finality and self-justification of the highest principle, is
not-as might be expected, according to the requirement of Cohen's "tran-
scendental method" and his elucidations of the notion of the genus of the
transcendental conditions-an explication of the content of the concept of
what is common to the conditions of validity of cognitions of mathematical-
natural science. It is, rather, an explication of the recognition of the factum
of mathematical-natural science itself-i.e., an explication of what one ulti-
mately accepts when one takes one's point of departure from the factum
of mathematical-natural science. What one ultimately accepts-and what
Cohen now designates as the content of the highest principle-is expressed
in the idea that there are in fact laws or, in Cohen's formulation, "that a law
ought to reign in the field of experience" (Cohen 1987a, 139). Since Cohen
conceives necessity as the modality of the validity of laws, and since the
domain of mathematical-natural science is identical with that of experi-
ence, this idea can therefore also be expressed in this manner: "we want to
recognize necessity in that field of our consciousness which is character-
ized as science, as mathematical-natural science" (Cohen 1987a, 139). What
is expressed by the content of the highest principle is nothing other than
the recognition of the moment of lawfulness as such, which is implied in
the recognition of the factum of science qua science,
and that means qua cognition from laws. This assertion is also ultimately
proven by the following assertion: "this idea, [namely,] to take as the start-
ing point the recognition of the factum of mathematical-natural science ...
posits itself as the highest law:'"
Several questions are to be addressed to this conception. Firstly, we must
explain the conspicuous alteration of the train of thought that has arist.·n
with the shift from the explication of the content of the general concept of

136 • JDRGFN STOLZENBERG

the condition of mathematical-natural science, to the explication of the con-
tent of the acceptance of the facticity of mathematical-natural science. This
explanation must, at the same time, shed light on whether this alteration
entails a change of the conception of a highest principle itself, or whether it
has only initiated the decisive, albeit non-self-evident, step to its appropri-
ate interpretation.
Further, it is unclear in what way Cohen could intend to use this concep-
tion to refute the accusation of circularity in his justification of the possibil-
ity of experiential cognition. This requires an elucidation of the exact sense
of the thesis of the self-justification and finality of the highest principle with
which Cohen opposes the accusation of circularity.
To answer these questions we must once more concentrate on what it
is that became conspicuous in the course of explicating the content of the
highest principle. We saw this conspicuousness when Cohen presented a
double description of the content of the highest principle, from which ini-
tially it is not evident how the description can be seen as the description of
the content of one and the same proposition. While Cohen, according to
the first description, understands the highest principle as a presentation of
the generic concept of the transcendental conditions of mathematical-natu-
ral science determined as experience, according to the second description,
he presupposes the idea of the recognition of the factum of mathematical-
natural science as that whose conditions are supposed to be explicated. As
he says, this idea "posits itself as the highest law" by the fact that that which
is here recognized, becomes designated as a general form of the lawful-
ness. While thereby at first the conditions and what they have in common
stand in view, from which the possibility of natural-scientific cognition is to
be explained, the essential character of natural-scientific cognition, which
should be explained through these conditions, is thematic in relation to the
general form of lawfulness. Since the conditions and that which is condi-
tioned, the ground of explanation and that which is to be explained are to
be distinguished from one another, it must appear incomprehensible that
both descriptions should count as descriptions of one and the same prin-
ciple. Thus one needs to ask: How can the epitome of the conditions of the
possibility of natural-scientific cognition be identified with what is their
essential formal character, i.e., the character of lawfulness as such?
Obviously, this can only be seen if what Cohen calls the transcendental
conditions of the possibility of natural-scientific cognition of experience are
the essential formal determinations of the logical content of the concept of
this cognition qua cognition from laws. For only then can it be said that
the concept in which the commonality of all the conditions of the possibil-
ity of this cognition is thought is identical with the concept whereby this
cognition is determined according to its essence. However, that means that

HERMANN COHEN's THEORETICAL PHILOSOPHY • 137

Cohen's demonstration of the conditions of the possibility of
entific cognition is in fact to be understood only as the explication of the
conditions of the logical possibility of the concept of this cognition, i.e., its
lawful validity.
This is precisely the sense of the "content" of the "transcendental method"
specified by Cohen himself, as is demonstrated in his following statement
on the "content of the transcendental method":

Experience is given; what is to be discovered are the conditions on which

their possibility rests. If the conditions which make the given experience
possible are found to make it possible in a way to speak of the experience
as valid a priori, and if strict necessity and unrestricted universality can be
awarded to it, then these conditions are to be designated as the con.'ititutive
characteristics of the concept of experience . ... That alone is the sole task of
transcendental philosophy.''

Several things follow from this. As the context shows, Cohen's transcen-
dental method can provide no proof for the validity of mathematical-nat-
ural scientific cognition. The only thing that it can accomplish and pro-
vide is the description of those cognitions in which the formal character of
lawfulness, which is proper to mathematical-natural scientific cognition, is
instantiated. This character of lawfulness or lawful validity, however, is not
proven in this manner; rather it will be accepted as factual together with
the acceptance of the facticity of mathematical-natural science. '!hat means
that Cohen's thesis that the form of lawfulness of a cognition warrants the
objective validity of the cognition is proven neither through the practice of
the transcendental method nor through recourse to the highest principle.
This is so because these cognitions, to which Cohen awards the status of
transcendental conditions, are for their part only species of laws; and the
highest principle, which should justify them-this has been shown by our
analysis of this highest principle-contains nothing other than the isolated
idea of the moment of lawful validity as such that is contained in all types
of laws. It therefore follows that the highest principle cannot he the kind of
principle from which the objective validity of the transcendental conditions
could be proven. Rather, it achieves only the theoretically much more mod-
est function of being the isolated expression of the unity of the moments of
lawfulness and objective validity, which are contained and always already
assumed in all lawful cognitions.
If one turns once more to the context in which Cohen has developed
the explication of his highest principle, then the accusation of a circular
grounding of the possibility of experience can be seen as a
ing. It is a misunderstanding, however, of which Cohen's "serious debate''

q8 • pJRGEN STOLZENBERG
and "resolution" is itself indeed prone to, through the reference to the
groundlessness of the highest principle and the recognition of the lawful-
ness of natural-scientific cognition that is thereby brought to expression
(Cohen 1987a, 139). For this accusation cannot be settled by referring to the
recognition of the factum of mathematical-natural science and the implied
recognition of the lawful character of its cognitions, which are elevated to
the content of such a highest principle. According to this critique, the cir-
cularity in the argument lies precisely in the fact that one attempts, in con-
junction with a highest principle of experience, to justify the condition of
the possibility of experience from experience itself Instead, this accusation
can only be "resolved" by showing that it has no target. In response to this,
we will show that one can in no way speak of a grounding of the validity
of experience from the highest principle, because there is no such ground-
ing. What exists and what is the content of Cohen's theory of experience is
solely the presentation of these laws that constitute the concept of experi-
ence concerning its assumed objective validity. The content of the epitome
of these laws is expressed in the idea that they all contain the formal-invari-
ant moment of lawfulness in which the objectivity of cognition is always
already implied, and the highest principle represents precisely this content.
And because there is no grounding for the validity of experience, no vicious
circle is set in motion with the exposition of a highest principle. To be sure,
it must be said that, since every cognition contains the moment of lawful
validity, which is expressed in the highest principle as such, it is no longer
thought to be justified through a higher proposition. Therefore, this highest
principle justifies itself and can be regarded as unconditional, insofar as it
can no longer have a ground of its validity that is distinct from it. However,
it is not this characteristic of the highest principle, but solely its content
whose consequence is that characteristic from which the accusation of cir-
cularity can be refuted.
For the same reason, the derived accusation of the impossibility of the
proof of the truth of cognition for accepted, factual mathematical-natural
science simply misses the point, and hence is to be rejected, because there
is no such proof in Cohen's theory of experience. Thus in another passage,
Cohen describes his theoretical program as follows: "The critique of cogni-
tion ... separates science into presuppositions (Voraussetzungen) and foun-
dations (Grundlagen) that are assumed in and for its laws:'" Since these
foundations are, for their part, laws in which their own objective validity
is implied, and since the highest principle only contains this moment of
lawfulness and this moment of the objective validity implied in it, one can
in no way speak of a proof for "the justification (Rechtsgrund) of certainty"
(Cohen 1984, 49) with respect to these presuppositions.

HERMANN COHF.N'S THEORETICAL PHILOSOPHY • 139

The Highest Prindple and the logk of Origin
In his investigation of the development of Hermann Cohen's theoretical
philosophy, Geert Edel has shown in an equally meticulous and convincing
manner that Cohen's main systematic work, the Logik der reinen Erkenntnis
(Logic of Pure Cognition), must be read and understood as a "sum" of the
systematic content of his reconstruction of Kant's theoretical philosophy,
rather than as an outline of an entirely new theoretical program. If one pur·
sues the question of the way in which Cohen's theory of a highest principle
has become useful for the system of the Logik der rein en Erkenntnis, then
one should refer to Cohen's remarks on the concept of a "logic of origin"
(Cohen 2005, 31-32) and, more precisely, to his conception of a "princi-
ple of origin" (Cohen 2005, 35). From these remarks, therefore, must arise
an answer to the question of how the concept of origin that is central for
Cohen's logic of cognition is different from that of the highest principle and
further, in which way the content-to which the quality of origin is now
attributed-is more specifically determined.
Before one can turn to these questions, one must call attention to a differ-
ent and more serious problem. It arises from Cohen's statements concern-
ing the systematic function of the principle of origin. In these statements
the Platonic concepts of the idea and hypothesis and Cohen's interpretation
of the same are the guiding clues." Thus he writes, "If the idea is primarily
a hypothesis, then the category of origin is the most fundamental ground-
laying (Grundlegung); it is the foundation (Grundlage) of modern science"
(Cohen 2005, 597 ).
And with reference to the grounding function of the Logik der reinen
Erkenntnis concerning the "unity of a system;' Cohen remarks: "1be unity
of the system demandsa central point in the foundation oflogic. This meth-
odological center is the idea of the hypothesis, which we have developed
into the judgment and into the logic of the origin" (Cohen 2005, 601).
If one initially follows the Platonic concept of the hypothesis, then Cohen
means by it a thesis, which presents the beginning of a line of argumenta·
tion. Here it is decisive, on the one hand, that the meaning of the thesis
consists solely in its function for the argument. This thesis consists in the
starting point and criterion for the truth of the propositions derived from
it. It does not consist in the validity that may be attributed to it indepen-
dently of this function. On the other hand, it is crucial that this thesis can
be questioned and replaced by other theses if its implications contradict it
or another higher-ranking assumption. 17
1l1is, however, results in a problem. For if one regards the concept of
hypothesis as the "methodological center" of the Logik der reinen Erkennt·
nis-as Cohen seems to do in his last -cited statement--and if one understands

140 • JI]RGEN STOLZENBERG

this function to the effect that the concept of hypothesis is utilized as an
interpretamen! of the principle of origin, then it appears that one can no lon-
ger speak of a continuity between the theorem of a highest principle and the
conception of the principle of origin. For the characteristics of being presup-
positionless, unconditioned, and necessary which Cohen had attributed to
the highest principle, no longer appear to be applicable to the principle of
origin interpreted as a hypothesis. In other words: If the principle of origin
is only to have the status of a ''ground-laying" (Grundlegung), conceived by
thinking and in principle always subject to revision, then it is not evident
how it can be understood in light of the highest principle, which Cohen
previously characterized as the unconditioned and insurmountable "sys-
tematic point of culmination" (Cohen 1987a, 143) of his system of synthetic
principles. Thus, one would have to go so far as to say that Cohen's interpre-
tation of the principle of origin as hypothesis actually presents an implicit
self-critique; it excludes and negates from the very start every attempt to
understand this principle as the successor to the former highest principle.
The alternative that results in this way between, on the one hand, proceed-
ing from a highest, unconditioned, and unsurpassable principle and, on the
other, undertaking the exposition of a "ground-laying" (Grundlegung) that is
only relatively fmal, because it is in principle subject to revision, is not only a
problem for Cohen's theoretical development. This alternative also dominates
the leading scholarship. While for Geert Edel, Cohen's concept of origin "can
only have the status of a hypothesis and can only signify a laying of grounds
( Grundlegung);'" Helmut Holzhey has interpreted Cohen's concept of origin
as "an unconditioned ground in itself" and as "absolute:'"
In order to decide between these alternatives, we have to examine whether
the respective interpretations of the relation between a highest principle, ori-
gin, hypothesis, and the unconditioned (or anhupotheton 20 ) are themselves
without alternatives. We will come to the conclusion that this is not the case.
In this light, one needs to first pursue for a moment Cohen's broader imple-
mentation of his hypothesis-theorem. It arises from a thesis that Cohen had
already formulated in the context of his first reconstruction of Kant. It claims
that the system of principles that forms the basis of natural-scientific cogni-
tion is to be conceived as a system that is incomplete, open, and capable not
only of expansion but of revision." In the Logik der reinen Erkenntnis, this
thesis appears in the assertion that "completeness" with respect to the "num-
ber of categories ... would not be a fullness, but would form an open wound
oflogic" (Cohen zoos, 396). The reasoning behind this thesis can be derived
from Cohen's orientation to modern mathematical-natural science, particu-
larly from the ''factum of becoming ( Werdefaktum) of mathematical-natural
science:· as it is now called in the Logik der reinen Erkenntnis (Cohen 2005,
76}. With this orientation, Cohen attempts to account for the circumstance

HERMANN COHEN's THEORETTCAL PHILOSOPHY a 141

that the inventory of natural-scientific cognitions cannot be seen as com-
plete but is in the process of continuous development, expansion, or revi-
sion. According to Cohen, however, this circumstance is to be acknowledged
not only for the inventory of scientific cognitions, but also for the inventory
of its basic concepts and principles-that is, for all "pure cognitions;' whose
presentation is the theme and task of the Logik der rein en Erkenntnis. "The
necessary notion of the progress of science;' as Cohen formulated his thesis,
"not only implies, but necessarily presupposes, the idea of the progress of
pure cognitions" (Cohen 2005, 396).
In the present context, it is crucial to see that from his thesis that "pure
cognitions" can only be awarded the status of "ground-layings" (Grundle-
gungen) which are capable of revision, Cohen has not drawn the conclusion
that, since this is the case, the idea of the form oflawfulness as such is to be
abandoned. Cohen's conclusion is rather that:

the ultimate foundations (Grundlagen) of logic are ground-layings

(Grundlegrmgen) whose expressions must change according to the prog-
ress of the problems and of the insights. It is a vain delusion that therefore
the law, the a priori, the eternal would become volatilized and subjectiv-
ized; rather, the eternity of reason is confirmed in the historical nexus of
ground-layings (Grundlegungen). (Cohen 2002b, 84-85)

This idea of the objectivity and invariability of the universal form of law-
fulness is, however, precisely the idea that forms the basis of the content of
the highest principle, as Cohen had posed and elucidated it in the second
edition of Kanis 7heorie der Erfahrung. Consequently one might conclude
that this idea of the form of lawfulness is now indeed able to interpret the
content of the principle of origin. The highest principle would now have to be
described as "origin;' since it is thinking alone which can generate the unity
of lawfulness and objectivity-and Cohen had already proposed this insight
in the context of the explication of his theorem of a highest principle.
This would also make another thesis understandable, which is decisive
for Cohen's conception of the principle of origin. This is the thesis that
"thinking and being are the same (in their object)" (Cohen 1928, 337). If one
adds Cohen's further theses, according to which "the basic form of being is
the basic form of judgment;' while the latter is "the basic form of thinking"
(Cohen zoo5, 47), then it becomes clear that Cohen's concept of being pri-
marily designates the moment of veridical being or being-true ( Wahrsein)
contained in a judgment, which is related to an object of cognition." Since
such a judgment, according to Cohen, has the ground of its validity solely in
thinking, one can say: "Being is the being of thinking. lherefore thinking,
as the thinking of being, is the thinking of cognition" (Cohen 2005, 18).

142 • JI]RGEN STOLZENBERG

 1his relation between thinking and being--which can be described as a
relation of the creation of being in and through thinking, and which Cohen
has misleadingly termed the "identity of thinking and being" (Cohen 2005,
t8)-would now have to be considered as the interpretation of the content
of the highest principle and at the same time as the fundamental interpreta-
tion of what Cohen, in the Logik der reinen Erkenntnis, has called the "prin-
ciple of origin:· In accordance with the initial interpretive thesis suggested
above, thinking is therefore to be understood as "thinking of the origin"
(Denken des Ursprungs) (Cohen 2005, 36), insofar as it is the sole origin of
being, and that theory which presents the manners in which thinking posits
being, is to be called the "logic of origin."
With this result, however, more questions appear to be posed than
answered. For on the one hand, it seems that with this result one can dem-
onstrate the continuity of the significance and systematic function of the
highest principle with respect to the principle of origin. On the other hand,
however, it must now appear entirely hopeless to use this interpretation to
understand Cohen's view that the concept of origin can merely be assigned
the status of a hypothesis in the sense of a "ground-laying" (Grundlegung)
which is, in principle, capable of revision. For with the conception of an
"identity of thinking and being;' the status of the unconditioned and neces-
sary is reinstated-a status which cannot be thought in the concept of the
hypothesis, but which is nevertheless to be attributed to the highest prin-
ciple. This now seems to allow for no other conclusion than that-against
Cohen's explicit self-understanding-one will nevertheless have to assign to
the principle of origin the status of a "foundation" (Grundlage) which must
thus be called absolute and therefore an anhupotheton.
These conclusions, however, turn out to be premature. For the question
raised above-namely, whether the indicated position has any alternative
entirely unanswered as yet. Consequently it has to
be shown in what follows that this alternative is in fact incomplete and that,
on its basis, Cohen's position cannot be adequately described.
This shall be demonstrated in two steps. It has to be shown, firstly, that
one has to adhere to the idea of what is unconditioned and without pre-
suppositions with reference to the status of the principle of origin, without
being forced into assenting to the idea of it as anhupotheton, which Cohen
denied earlier. Secondly, it must be shown that one has to adhere to the idea
of the hypothesis as a "ground-laying" (Grundlegung) without having to
assent to the idea of its revisability with respect to the principle of origin.
If one brings into more precise view, for this purpose, Cohen's explica-
tion of the concept of the anhupotheton, then it becomes clear that Cohen
primarily highlights with it a moment that he designates-stemming from
his interpretation of the status of the Platonic Idea of the a

HERMANN COHEN'S THEORETICAL PHILOSOPHY • 143

position "beyond being;' while he interprets this beyondness as "beyond
all thinking, all cognition" (Cohen 2002a, 32). According to Cohen, such
a beyondness characterizes the "medieval concept of the absolute" as the
"central concept" of a "false and unscientific metaphysics;' which precisely
the "logic of origin" had intended to "invalidate" (Cohen 2005, 6o6).
One can see immediately that such an interpretation of the concept of
the anhupotheton can, in fact, in no way count as the interpretation of the
status of the principle of origin. For Cohen's formula of the "identity of
thinking and being" has its point precisely in removing any claim of the
independence and self-sufficiency of being from the achievement of think-
ing, and in attributing to the concept of being significance and meaning
only as being posited or created in thinking. Nevertheless, it is not only
acceptable, but mandatory, to maintain the idea of the unconditioned in
order to characterize the status of the principle of origin. One has to char-
acterize the relation of thinking and being as unconditioned because there
is no instance (nor can there be) from which this relation could in itself be
deduced, since this relation represents nothing other than the basic form
of each and every true judgment that is related to an object of cognition.
This is a basic form that must already be presupposed in every judgment
that makes a claim to objective validity and from whirh it would have to be
deduced. Thus the principle of origin has to be construed as unconditional,
although not as an anhupotheton in Cohen's sense. 24
Cohen's idealistic main thesis, according to which every being is being
posited in thinking, also permits a use of the concept of the hypothesis as
"ground-laying" (Grundlegung) without the idea of revisability-and this
use, furthermore, is possible without a collision with Cohen's thesis of the
revisability of "pure cognitions." At this point one may recall that the con·
tent of the principle of origin only expresses the purely empty idea of the
invariant form of each of its lawful cognitions, which is objectively deter-
mined according to its content. If the principle of origin were now assumed
to be capable of revision, then Cohen's undertaking of a logic of pure cogni·
tion would be entirely questionable. For then indeed the idea of the form
of lawfulness, which forms the basis of Cohen's logic of cognition, would
itself be rendered vulnerable. Hence Cohen could appropriately, albeit dras-
tically, call it a "vain delusion" if"the law were volatilized and subjectivized,"
for the principle of origin expresses nothing other than the principle of any
given law. Consequently, one can in no way meaningfully speak of a revi·
sion with respect to the principle of origin. 25
Furthermore, one cannot say that the principle of origin could not be con·
sidered as unconditional and unrevisable (and therefore only as "hypoth·
esis") for the reason that its distinction as principle of every law (i.e., as
original positing of being attributable to thinking) would be a distinction

144 • JI]RGEN STOLZENBERG

granted to it from thinking. This would be a distinction which could in
principle not make the claim to describe the "true" essence of something,
and henceforth also could not be the principle of origin. This thesis is to
be rejected for the reason that, on the one hand, it is not at all concerned
with the description of such an essence, but only with the isolated idea of
the relation of thinking and being mentioned earlier-and this is the basis
of the analysis of any kind of essence. On the other hand, it is nonsensical
to submit this relation to the limitation of its mere being-thought, for pre·
cisely this is excluded through this relation. With this idea we think not the
merely thought relationship of thinking and being, but this relation itself,
and therein the mere being-thought as a moment of its content is not only
not contained, but precisely negated implicitly.
Nevertheless, the principle of origin can be understood as "ground-lay·
ing" ( Grundlegung), and indeed in a double sense: Firstly, insofar as it repre·
sents the common form of all particular "ground-layings" (Grundlegungen ),
determined according to their respective particular contents; and secondly,
in that it presents thinking in its original relation to being, which is in prin·
ciple not capable of revision, and this relation consists in that being is a
being created by thinking. In this sense, then, thinking is to be construed as
the "ground-laying" (Grundlegung) of being or, what amounts to the same,
the principle of origin describes the "laying of the ground" (Grundlegung)
of being through thinking. Hence, we can say with Geert Edel that Cohen's
origin is no "separated absolute, but a moment, a character or a 'quality' of
thinking, which however realizes itself in all particular cognitions"" and
which is posited by Cohen in the form of an autonomous highest prin·
ciple. Against Edel's interpretation, however, one must insist that the idea
of lawfulness that is expressed in this principle cannot as such be subject to
revision, and thus that the principle of origin can only be understood as in
that sense a hypothesis-i.e., as a "ground-laying" (Grundlegung) in which
thinking is the ground of being.
If one looks back on the previously mentioned phrase of Helmut Holz-
hey, then the origin is indeed entirely "determined in itself as uncondi·
tioned ground;' since a further ground, from which its validity could yet be
grounded, cannot be determined. One can ask, however, whether one can
also follow Holzhey's interpretation, suggested initially by Hans Wagner,"
according to which the origin so determined can be understood as "self·
determination through self-relation:' 1" This is supposed to be possible in
that the origin is thought as "self· relation to an Other that is grounded by
this origin and differentiated from it in order to determine oneself in this
self-relation to the Other in this Other:'" Such a conception is incompat·
ible with Cohen's notion of origin: Since with this notion Cohen meant
only the idea of the form of lawfulness which is as such entirely empty of

HERMANN COHEN'S THEORETICAL PHfLOSOPHY • 145

content, no ground of creation of a difference in contents can be contained
in it, and it is precisely this difference that is indicated with the moment of
the Other. Cohen's proposition obviously also means that indeed validity
is conditioned through the origin, but all pure cognitions are only "varia-
tions" (Cohen 2005, 36) of the origin, and that means instances, in which
pure thinking comes to presentation or to appearance: It does not mean
that these instances are analytical derivatives of the origin as a self-differen-
tiating principle. And this is also indicated by Cohen's thesis that the unity
of the highest principle is not a "creative unity, but merely an ideal unity,
namely the unity of the law" (Cohen 1987a, 591).
In conclusion, Cohen's principle of origin is no metaphysical absolute,
but is also not a hypothesis capable of revision; rather it is the principle
of the form of lawfulness, which qualifies thinking itself. This principle is
indeed an ultimate principle, yet is not a principle that generates difference.
Thus here as well, as so often, the truth lies in the middle.
TRANSLATED BY COLIN f. HAHN

NOTES

1. On the following, see Edc1Jg88, 55-51 and 88-89. On this topic, see further
the investigation of Winter
2. The corresponding assertion of Cohen's amounts to: "The goal is: the expla-
nation of the possibility of synthetic propositions a priori. These form the true
and entire content of experience. And this content of experience, which is given
in mathematics and pure natural science ... needs to be explained according to its
possibility" (Cohen 1987b, 206).
3. On this, see specifically Geert Edel's introduction to the second edition (also
in the reprint of the third edition) of Kants 7heorie der Erfahrung (Cohen 1987a,
ll*-12*).
4- See Cohen 1984, 5-6.
s. See Cohen 2001, 46; and Cohen 1984, 6. For this specifically, see Ede! 1988,
384ff.
6. See Cohen 1987b, 206: As Cohen says here, "the elements have to be sought,
in which the alleged apriority of the same [i.e., of the contents of natural-scientific
cognition! consist. The value characteristics of the a priori are necessity and univer-
sality. The ground of these, however, can only be known in us, in concepts; not in the
object. It is thus necessary to find the concepts in which the a priori must be secured,
which conditions the necessity and universality of the given content of experience."
7· Thus Cohen explains in Cohen 1987b, 206: "This course [i.e., the course of the
analytical methods J is dearly indicated in the Prolegomena, but also in the IfirstJ Cri-·
tique." Cohen could reach this position because of the fact that Kant has taken some
sections of the Prolegomena, in which he in fact carries out the analytic method, into
his introduction to the second edition of the Critique of Pure Reason. (1hus §§ s-6
of Kant's introduction to the second edition of the Critique of Pure Reason parallel

146 • J0RGEN STOLZENBF.RG

§§ 2-5 of the Prolegomena.) On the synthetic and analytic methods respectively, cf.
Kant's explanation in the Prolegomena: "In the Critique of Pure Reasofl I worked on
this question syflthetically, namely by inquiring within pure reason itself, and seek-
ing to determine within this source both the elements and the laws of its pure use,
according to principles. This work is difficult and requires a resolute reader to think
himself little by little into a system that takes no foundation as givm except reason
itself, and that therefore tries to develop cognition out of its original seeds without
relying on any fact (factum) whatever.... the Prolegomefla must therefore rely on
something already known to be dependable, from which we can go forward with con-
fidence and ascend to the sources, which are not yet known .... The methodological
procedure of the Prolegomena.,. will therefore be analytic" (Kant 1902, 274-75; Kant
1997, 25-26: emphasis added).
8. On the following, see Baum 1986. Baum has pointed out that the "so-called
'transcendental arguments"' discussed in the Anglo-Saxon Kant literature "are a
new version of the 'transcendental methods' of Neo-Kantianism" (17-18). To this
one can add P. F. Strawson's critique of "transcendental psychology": Strawson's
critical thesis-namely, that Kant's theory of cognition deals with an "imaginary
subject of transcendental psychology" (Strawson 1966, 97)-has its forerunner in
Cohen's corresponding critique of Kant. Against Strawson's critique of Kant in this
regard, see Carl1989, 125-26.
9. The concept of experience is henceforth used in the Cohen ian sense.
10. See Lange 1974. Lange's accusation reads verbatim: "If one pushes the empha-
sis of the mere transcendental standpoint thus far, then one arrives ... at the tautol-
ogy that experience is indeed to be explained from the conditions of possible expe-
rience" (577). The avoidance of this "tautology" and the attainment of a "synthetic
result" (577) is seen by Lange as given with the proof that the "categories [must] yet
be something other than conditions of experience" (578). Lange sees this "more"
as given in the psycho-physical "organization" (578) of humans, from which the
categories are to be grounded. On Lange's interpretation of Kant, see KOhnke 1986,
esp. 251-52.
n. See Cohen 2001, 33·
u. See Cohen 1987a, 138.
13. Loc. cit.: emphasis added.
14. Cohen 2001, 32: emphasis added.
15. Cohen 1984, 6: emphasis added.
16. Cohen's interpretation of the Platonic concepts of idea and hypothesis, which
is significant for his theoretical development, is found in Platons Ideenlehre und die
Mathematik. In: Rektoratsprogramrn der Vfliversitiit Marburg vom Jahr 1878, Son-
derausgabe Marburg 1879. Reprinted in Cohen 1928, 336-366. On this, see the com-
prehensive presentation by Geert Edel in Edel1988, 202-03. On Cohen's reception
of Plato, see further Lembeck 1991, as well as Poma 1987
17. On the Platonic concept of hypothesis, see Wieland 1982, 150-51.
18. Edell991, 81.
19. Holzhey 1986, 183 and 218.
20. [Translator's note: The Greek anhupotheton indicates a principle that is not

HERMANN COHEN's THEORETICAL PHILOSOPHY a 147

to be taken as a hypothesis, but is absolutely justified. The term is used by Plato to
describe the principle that is reached through dialectical reasoning, and by Aristo-
tle to describe the principle of non -contradiction.)
21. See Cohen 1987b, 101; and Cohen 1987a, 519.
22. On the concept of veridical being, see Tugendhat 1976, 56-57-
23. See Cohen 1928, 363.
24. Contrary to this, Geert Edel has understood the quality oft he unconditioned,
which pertains to the relation of thinking and being, always and solely as the kind
of quality that pertains to an entity located beyond thinking; such an entity would
therefore have to be thought as "anhupotheton." See Edel 1991, 81 et pass.
25. The quoted citation is also found in the context of the presentation by Geert
Edel (Edel1991, 76). However, since for Edel apparently the oppositions "ground.
laying" ( Grundlegung) and "foundation (Grund!age) given in itself;' or "hypothesis"
and "anhupothcton," respectively, form complete disjunctions, he has overlooked
Cohen's explanation that is directed against his interpretation and furthermore has
held on to the thesis that the concept of the origin itself could only be thought as
hypothesis. However, it is inconceivable how this interpretation could escape the
(absurd) consequence of the possible self-nullification of the foundation of Cohen's
logic of cognition. Against Edel's interpretation of the origin as a hypothesis capable
of revision, we can also refer to a self-commentary of Cohen's from his posthumous
work, the Religion of Reason Out of the Sources of Judaism, in which Cohen has
identified the principle of origin as the "archetype of all lawfulness" and, precisely
as such, as immune to such objections: "Reason ... means positively the lawfulness,
the archetype of! awfulness ... and there can be no firmer, no more profound foun-
dation than that oflawfulness .... {W]herever lawfulness reigns, there the domain
of reason is secured .... Lawfulness is so weighty that against it every objection has
to retreat. For what more reliable and certain basis than lawfulness could another
reason offer?" (Cohen 1966, n-12/Cohen 1972,10-11).
26. Edel1988, 433 (ed. Edel).
27. See Wagner 1959, 128-29. On Holzey's connection to Hans Wagner, see Holz-
hey 1986, xi, 183 n. 18, and 218.
28. Holzhey 1986, 183.
29. Holzhey 1986, 218. lhe cited formulation is a citation from Hans Wagner's Phi-
losophic und Reflexion, and is identified as such by Holzhey. Wagner's thesis concern·
ing the absolute, which forms the basis of Holzey's interpretation of the Cohenian
principle of origin, reads in its entirety: "1he absolute is thus essentially self-relation
to an Other that is grounded by the ahsolute itself and differentiated by it, in order to
determine in this self-relation to the Other its own self" (Wagner 1959, 128).

BIBLIOGRAPHY

Baum, Manfred. 1986. Deduktion und Beweis in Kants Transzendentalphilosophic Untcr-

suclumgen zur Kritik der reinen Vernunft. KOnigstein: Hain Verlag bei Athenaum.
Carl, Wolfgang. 1989. Der scltweigende Kant.· die Entwiirfe zu einer Deduklio11 tier
Kategorien vor 178l. G6ttingen: Vandenhoeck & Ruprecht.

148 • J0RGEN STOLZENBERG

Cohen, Hermann. 1928. Herman Cohens Schriften zur Philosophic und Zeitgeschichte.
Ed. Albert GOrland and Ernst Cassirer. Vol. L Berlin: Akadernie-Verlag.
- - . 1966. Religion der Vernunft aus den Quellen des ]udentums. 2d ed. Darm-
stadt: Melzer. (Religion of Reason out of the Sources of Judaism. Trans. Simon
Kaplan. New York: Frederick Ungar, 1972.)
- - . 1984. Das Prinzip der Infinitesimal-Methode und seine Geschichte. (Werke,
vol. 5/1). Hildesheim: O!ms.
- - . 1987a. Kants Theorie der Erfahrung. 3d ed. ( Werke, vol. tit). Hildesheim: O!ms.
- - . 1987b. Kants 1heorie der Erfahrung. tst ed. ( Werke, vo!. 1/3). Hildesheim: Olms.
- - . 2001. Kants Begrii.ndung der Ethik nebst ihren Anwendungen auf Recht, Reli-
gion und Geschichte. 2d rev. and exp. ed. (Werke, voL 2). Hildesheim: Olms.
- - . 2002a. Der Begriff der Religion im System der Pltilosophie. (Werke, vol. 10).
Hildesheim: O!ms.
- - . 2002b. Ethik des reinen Willens. zd ed. (Werke, vo!. 7). Hildesheim: Olms.
- - . 2005. Logik der reinen Erkenntnis. 2d ed. ( Werke, vol. 6/I). Hildesheim: Olms.
Edel, Geert 1988. Von der Vernunftkritik zur Erkenntnislogik: die Entwicklung der
theorelischen Philosophic Hermann Cohens. Freiburg: K. Alber.
- - . 1991. "Kantianismus oder Platonismus? Hypothesis als Grundbegriff der
Philosophic Cohens:' II Cannocchia/e 1-2: 59-87.
Ho!zhey, Helmut. 1986. Cohen und Natorp. VoL 1. Basel: Schwabe.
Kant, Immanuel. 1902. Kants gesammelte Schriften. Vol. 4. Berlin: Reimer.
gomena to Any Future Metaphysics. Ed. Gary Hatfield. Cambridge: Cambridge
University Press, 1997.)
KOhnke, Klaus Christian. 1986. Entstehung und Aufstieg des Neukantianismus: die
deutsche Universitiitsphilosophie zwischen Idealismus und Positivismus.
furt: Suhrkamp. (The Rise ofNeo-Kantianism: German Academic Philosophy Bet-
ween Idealism and Positivism. Trans. R. J. Hollingdale. Cambridge: Cambridge
University Press, 1991.)
Lange, Friedrich Albert. 1974. Geschichte des Materialismus und Kritik seiner Bedeu-
tung in der Gegenwart. With an introduction by Alfred Schmidt. Frankfurt:
Suhrkamp.
Lembeck, K. H. 1991. "Cohens fruhe P!aton-Deutung und seine Que!!en:' II Can-
nocchiale: 89-116.
Poma, Andrea. 1987. "II Platone di Hermann Cohen:' Annuario Filoso.fico 3:
2ll-242.
Strawson, P. F. 1966. 7he Bounds of Sense: An Essay on KantS 'Critique of Pure Rea-
son'. London: Methuen.
Tugendhat, Ernst. 1976. Vorlesungen zur EinfUhrung in die sprachanalytische Philo-
sophie. Frankfurt: Suhrkamp.
Wagner, Hans. 1959. Philosophic und Rejlexion. Munich: E. Reinhardt.
Wieland, Wolfgang. 1982. Platon und die Formen des Wissens. GOttingen: Vanden·
hoeck & Ruprecht.
Winter, Eggert. 1980. Ethik und Rechtswissenschaft. Eine historisch-systematische
Untersuchung zur Ethik·Konzeption des Marburger Neukantianismus im Werk
Hermann Cohens. Berlin: Duncker & Hum blot.

HERMANN COHEN'S THEORETICAL PHILOSOPHY a 149

 7.

Transcendental logic and Minimal Empiricism:

lask and McDowell on the Unboundedness of
the Conceptual

Steven G. (rowel!

Introduction: Empiricism in the Shadow of Hegel

In the preface to Mind and World, john McDowell describes his work as "a
prolegomenon to a reading" of Hegel's Phenomenology.' He also remarks
that Robert Brandom has decisively shaped his thinking by "forcing me to
get clear about the differences, small in themselves, that transform for me
the look of our wide measure of agreement" (MW ix). These remarks are
linked, for both the agreement and the differences can be seen to reflect
different aspects of the Phenomenology of Spirit, which Hegel characterizes
both as the "science of the experience of consciousness" and also as the "lad-
der" to the science of logic.' The agreement between Brandom and McDow-
ell consists in their mutual embrace of what the latter calls the "unbound-
edness of the conceptual" -the recognizably Hegelian "refusal to locate
perceptible reality outside the conceptual sphere" (MW 26). Thinking does
not stand on one side of a divide whose opposite side would be a merely
given, perceptible world, alien to the concept. What is perceived already
belongs within what Wilfrid Sellars called "the space of reasons." ·nlC differ-
ences between Brandom and McDowell, in turn, stem from something like
different assessments of the relation between the Phenomenology and this
unbounded conceptual realm.
For Brandom, the conceptual order is unbounded because the con·
tent of experience is not a mental representation, located in an individual
mind, but a function of socially mediated commitments and entitlements
incurred by participants in the game of giving and asking for reasons. 1he
indexicality of perceptual consciousness is explained causally as reliahle
differential responsiveness to the environment. Expressed as a report,
ception serves no justificatory role but is merely an "entry move" into the

150