Heidemann

Hans Kelsen’s Pure Theory of Law is the most prominent

example of legal normativism. This Element traces its genesis.
In philosophy, normativism started with Hume’s distinction
between Is-and Ought-propositions. Kant distinguished
practical from theoretical judgments, while resting even the
latter on normativity. Following him, Lotze and the Baden neo- Philosophy of Law
Kantians instrumentalized normativism to secure a sphere of
knowledge which is not subject to the natural sciences. Even
in his first major text, Kelsen claims that law is solely a matter
of Ought or normativity. In the second phase of his writings,
he places himself in the neo-Kantian tradition, holding legal
norms to be Ought-judgments of legal science. In the third

Hans Kelsen’s

Hans Kelsen’s Normativism

phase, he advocates a barely coherent naive normative realism.
In the fourth phase, he supplements the realist view with a strict
will theory of norms, coupled with set pieces from linguistic
philosophy; classical normativism is more or less dismantled.
Normativism
About the Series Series Editors
This series provides an accessible George Pavlakos
overview of the philosophy of law, University of

Carsten Heidemann
drawing on its varied intellectual traditions Glasgow
in order to showcase the interdisciplinary Gerald J. Postema
dimensions of jurisprudential enquiry, University of

https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

review the state of the art in the field, North Carolina
and suggest fresh research agendas for at Chapel Hill
the future. Focussing on issues rather than
Kenneth M.
traditions or authors, each contribution
Ehrenberg
seeks to deepen our understanding of
University of Surrey
the foundations of the law, ultimately with
a view to offering practical insights into
some of the major challenges of our age.

Cover image: Jane Couroussopoulos, www.janepaint.com ISSN 2631-5815 (online)

ISSN 2631-5807 (print)
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press
 Elements in the Philosophy of Law
edited by
George Pavlakos
University of Glasgow
Gerald J. Postema
University of North Carolina at Chapel Hill
Kenneth M. Ehrenberg
University of Surrey

HANS KELSEN’S
NORMATIVISM

Carsten Heidemann
Schleswig-Holstein Bar Association
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press
 University Printing House, Cambridge CB2 8BS, United Kingdom
One Liberty Plaza, 20th Floor, New York, NY 10006, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre,
New Delhi – 110025, India
103 Penang Road, #05–06/07, Visioncrest Commercial, Singapore 238467

Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning, and research at the highest international levels of excellence.

www.cambridge.org
Information on this title: www.cambridge.org/9781108995221
DOI: 10.1017/9781108993661
© Carsten Heidemann 2022
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2022
A catalogue record for this publication is available from the British Library.
ISBN 978-1-108-99522-1 Paperback
ISSN 2631-5815 (online)
ISSN 2631-5807 (print)
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

Cambridge University Press has no responsibility for the persistence or accuracy of

URLs for external or third-party internet websites referred to in this publication
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
 Hans Kelsen’s Normativism

Elements in the Philosophy of Law

DOI: 10.1017/9781108993661
First published online: January 2022

Carsten Heidemann
Schleswig-Holstein Bar Association
Author for correspondence: Carsten Heidemann, [email protected]

Abstract: Hans Kelsen’s Pure Theory of Law is the most prominent example
of legal normativism. This Element traces its genesis. In philosophy,
normativism started with Hume’s distinction between Is-and Ought-
propositions. Kant distinguished practical from theoretical judgments,
while resting even the latter on normativity. Following him, Lotze and the
Baden neo-Kantians instrumentalized normativism to secure a sphere of
knowledge which is not subject to the natural sciences. Even in his first
major text, Kelsen claims that law is solely a matter of Ought or normativity.
In the second phase of his writings, he places himself in the neo-Kantian
tradition, holding legal norms to be Ought-judgments of legal science. In
the third phase, he advocates a barely coherent naive normative realism. In
the fourth phase, he supplements the realist view with a strict will theory of
norms, coupled with set pieces from linguistic philosophy; classical
normativism is more or less dismantled.

Keywords: Kelsen, normativism, Kant, neo-Kantianism, validity

© Carsten Heidemann 2022

https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

ISBNs: 9781108995221 (PB), 9781108993661 (OC)

ISSNs: 2631-5815 (online), 2631-5807 (print)
 Contents

1 Introduction 1

2 The Genesis of Normativism in Philosophy 2

3 Hans Kelsen’s Normativism 23

4 Summary 72

References 75
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press
 Hans Kelsen’s Normativism 1

1 Introduction*
Normativism is an important strand in philosophy. It can roughly be defined as
having three basic features. First, there is a fundamental mode of Ought, or
normativity, which cannot be reduced to the mode of Is, or factuality. Second,
normativity is in some way objective. Third, normativity is constitutive of
a certain domain, be it ontology, epistemology, language, morality, or law.
In accordance with the third feature, philosophical normativism is not
restricted to securing the objectivity and autonomy of the ‘practical’ sphere; it
also plays an important role in metaphysics. As will be shown (Section 2.2.2),
Kant’s rejection of transcendental realism without succumbing to skepticism
was only possible by his adopting a variety of normativism for his theoretical
philosophy. In present-day philosophy, normativism dominates not only parts of
the theory of meaning in a narrow sense; it is also central to Hilary Putnam’s
internal realism, Jürgen Habermas’s discourse theory, and Robert Brandom’s
inferentialism – to name some prominent examples.
Normativism in legal theory is mainly associated with the writings of Hans
Kelsen (1881–1973). While there is scarcely any legal theorist who would
subscribe to its tenets without restrictions, the Pure Theory of Law gained
worldwide recognition and had an impact on authors like H. L. A. Hart,
Joseph Raz, and Robert Alexy. The normativistic tendencies of current legal
theory are generally due to Kelsen’s immediate or mediated influence. But
Kelsen’s texts represent not only what might be called the ‘historical core’ of
legal normativism; having undergone several changes of paradigm during the
long span of time in which they were published, they also give an overview of
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

a broad range of divergent aspects of normativism. All of this justifies a closer

look at his theory.
The term ‘normativism’ seems to owe at least part of its currency to Kelsen’s
theory. It is comparatively young. The German equivalent, Normativismus, had
been in use since the nineteenth century; it gained prominence after it was
introduced into legal theory, by the right-wing legal philosopher Carl Schmitt,
as a pejorative ‘battle term,’ when polemicizing against Kelsen’s theory of law
in the 1930s.1 Schmitt saw normativism in law as the view that law is exclu-
sively a system of (normative) rules which are objective, impersonal, and
independent of factuality (Schmitt 1977: 370–1). This definition captures fairly
well the three aspects mentioned above and what Kelsen probably had in mind
as the lowest common denominator of his different conceptions of normativism.

* I am grateful to two anonymous reviewers for helpful remarks. Special thanks go to Monika
Zalewska for our enriching discussion of Kelsenian and philosophical topics.
1
On the Kelsen–Schmitt controversy, see Paulson 2017.
 2 Philosophy of Law

In the second section of this Element, the development of the Is–Ought dualism
and normativism in philosophy is traced, by sketching some historical ‘land-
marks’ which informed Kelsen’s theory. This is necessary because the meaning
and role of ‘Ought’ can hardly be understood without drafting the history of this
concept – unless one is content with the meagre and apodictic explanation of the
dualism Kelsen gave in the second edition of Pure Theory of Law:

The difference between is and ought cannot be explained further. We are

immediately aware of the difference. Nobody can deny that the statement
‘something is’ – that is, the statement by which an existent fact is described –
is fundamentally different from the statement: ‘something ought to be’ –
which is the statement by which a norm is described. (Kelsen 2005: 5–6
[Kelsen 1960: 5])

In the third section, the genesis of Kelsen’s version of normativism is sketched.

His first major text, Hauptprobleme der Staatsrechtslehre (1911), is treated com-
paratively extensively. It contains the germs of Kelsen’s later Pure Theory of Law
while connecting it to the jurisprudential tradition of the nineteenth century;
besides, though little known, it is a true specimen of normativism in its own right.
Both the second and third sections concentrate on neo-Kantianism for three
reasons. First, neo-Kantianism is the predecessor of modern analytical philosophy;
the Baden school developed normativism as a metaphysical theory in a highly
sophisticated manner so that it still informs present-day discourse.2 Second,
Kelsen’s theory emerged at a time when neo-Kantianism impregnated German
academic culture. It incorporated neo-Kantian theorems ‘subcutaneously’ from the
beginning and is not intelligible without this theoretical surrounding. Third,
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

Kelsen’s explicitly neo-Kantian writings from the 1920s, especially Das Problem
der Souveränität, Der soziologische und der juristische Staatsbegriff and
Rechtswissenschaft und Recht, though philosophically his best, are not well
known; they are written in a difficult style and have not been translated into English.
Finally, the fourth section gives a short summary of the findings of sections
two and three.

2 The Genesis of Normativism in Philosophy

2.1 Hume’s Law
Hume’s Law is the starting point of the modern discussion of the dualism of Is
and Ought.3 Although the notion that there is some form of normativity is

2
On this topic, see the instructive paper Beiser 2009: 9–37.
3
However, in secondary literature both whether Hume’s text is the origin of the philosophical
tradition of a dualism between Is and Ought and what Hume really meant is contested; see the
seminal paper McIntyre 1959: 451–68.
 Hans Kelsen’s Normativism 3

millennia-old, so that saying something is valuable, obligatory, or ought to be

the case is not just a statement of fact, it is only since David Hume (1711–76)
that it has been an explicit topos of philosophy that norms cannot be deduced
from nature, or Is-propositions.
Hume, together with Kant and Wittgenstein, is one of the great innovators of
modern philosophy (if it is deemed to be an analytical enterprise) insofar as he
tries to completely reform metaphysics, doing away with any elements which
have a supernatural flavor. Hume’s classic formulation of the dualism of Is and
Ought runs as follows:

In every system of morality, which I have hitherto met with, I have always
remarked, that the author proceeds for some time in the ordinary way of
reasoning, and establishes the being of a God, or makes observations con-
cerning human affairs; when of a sudden I am surprised to find, that instead of
the usual copulations of propositions, is, and is not, I meet with no proposition
that is not connected with an ought, or an ought not. This change is imper-
ceptible; but is, however, of the last consequence. For as this ought, or ought
not, expresses some new relation or affirmation, ‘tis necessary that it should
be observed and explained; and at the same time that a reason should be
given, for what seems altogether inconceivable, how this new relation can be
a deduction from others, which are entirely different from it. But as authors do
not commonly use this precaution, I shall presume to recommend it to the
readers; and am persuaded, that this small attention would subvert all the
vulgar systems of morality, and let us see, that the distinction of vice and
virtue is not founded merely on the relations of objects, nor is perceived by
reason. (Hume 2007a: 302)

This is ‘Hume’s Law’, usually summarized as saying that it is impossible to

https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

deduce Ought-propositions from Is-propositions. But actually the text is richer

than that. It comprises different claims. First, Is and Ought are “copulations of
propositions” (a is b vs. a ought to be b). Second, Ought-propositions concern
morality, God, or human affairs. Third, it seems to be inconceivable (though
perhaps a reason might be given) that Ought-propositions could be deduced
from Is-propositions, because the Ought-denotes some “new relation or affirm-
ation.” Fourth, Ought-propositions cannot be validated by “relations of objects”
or by reason.
Hume does not argue for these claims; he seems to take them as self-evident.
The short passage quoted above more or less exhausts what he has to say
concerning the subject; the dualism does not play a vital role in his theory,
nor is Hume consequent in observing it. In his philosophical system, the Is–
Ought dualism results in skepticism concerning the objectivity of moral pro-
positions, because Hume only accepts two categories of objective knowledge,
namely, “relations of ideas” (expressed by analytical/mathematical sentences)
 4 Philosophy of Law

and “matters of fact” (expressed by empirical sentences) (Hume 2007b: 18).

Accordingly, his description of the Is–Ought distinction differs in two (possibly
contradictory) points from the provisional definition of normativism given
above. Ought-propositions, for him, are not objective; and he does not
altogether exclude the possibility that there might be a derivation of Ought
from Is, after all.

2.2 Normativism in Kant’s Critique of Reason

Immanuel Kant (1724–1804) shares the critical concern of revising metaphys-
ics with the aim of purifying it from any supernatural elements. His approach,
however, is different from Hume’s. To put it simply, Hume’s perspective is that
of an observer of ‘human nature’; he analyzes the elements of cognition from
the outside, so that cognition seems, to him, to be a matter of associations of
perceptions. Kant’s philosophy, on the other hand, starts – as it were – inside
cognition and explores the necessary conditions for the possibility of objective
judgments – it is the perspective of a participant. And while Hume radicalizes
the theories of his empiricist predecessors by jettisoning rationalist elements,
Kant seeks a synthesis between rationalism and empiricism by assigning both
empirical and rational elements their proper place in cognition.
Kant advocates the thesis that Ought-judgments cannot be reduced to or
deduced from Is-judgments without referring to Hume in this context (of
course, he explicitly mentions him when confessing that it was Hume who
aroused him from his “dogmatic slumber” (Kant 2004: 10 [A 13]). This is
understandable considering the background that normativity plays a different
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

and more important role in Kant’s philosophy than in Hume’s. His transcen-
dental idealism tries to steer a middle path between empiricism and rational-
ism in order to avoid Humean skepticism, and normativity plays an important
part in this project. Kant points to normativity to strictly separate practical
questions from empirical or theoretical ones. But Kant is also the first major
philosopher to turn the tables and hold not only that Ought-propositions
cannot be reduced to Is-propositions; rather, any kind of theoretical or Is-
cognition presupposes normativity. For him, practical reason has, in several
ways, priority over theoretical reason.

2.2.1 The Distinction between Is and Ought

Kant maintains explicitly that there is no way of deriving what ought to be
(done) from natural facts. Conversely, the Ought is meaningless when dealing
with proceedings in nature. To describe this state of affairs, he uses terminology
very similar to that which Kelsen employed some 150 years later:
 Hans Kelsen’s Normativism 5

The ought expresses a species of necessity and a connection with grounds

which does not occur anywhere else in the whole of nature. In nature the
understanding [Verstand] can cognise only what exists, or has been, or will
be. It is impossible that something in it ought to be other than what, in all these
time-relations, it in fact is; indeed, the ought, if one has merely the course of
nature before one’s eyes, has no significance whatever. We cannot ask at all
what ought to happen in nature, any more than we can ask what properties
a circle ought to have; but we must rather ask what happens in nature, or what
properties the circle has. (Kant 1998: 540 [A 547])

This passage contains some further qualifications of the Ought which are not
important in this context, but which are noteworthy: The Ought goes along with
a kind of necessity, and it expresses a “connection with grounds,” not with
causes.
In another passage in the Critique of Pure Reason, Kant maintains:

Whereas, so far as nature is concerned, experience supplies the rules and is

the source of truth, in respect of the moral laws it is, alas, the mother of
illusion! Nothing is more reprehensible than to derive the laws prescribing
what ought to be done from what is done, or to impose upon them the limits
by which the latter is circumscribed. (Kant 1998: 398 [A 318–19/B 375])

Like Hume, in both passages Kant does not further argue for this point, but
seems to take it to be self-evident. Finally, in the Critique of Judgment, regard-
ing the sphere of practical reason, Kant holds:

Now since . . . the objective necessity of the action, as duty, is opposed to that
which it, as an occurrence, would have if its ground lay in nature and not in
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

freedom (i.e., in the causality of reason), and the action which is morally
absolutely necessary can be regarded physically as entirely contingent (i.e.,
what necessarily should happen often does not), it is clear that it depends only
on the subjective constitution of our practical faculty that the moral laws must
be represented as commands (and the actions which are in accord with them
as duties), and that reason expresses this necessity not through a be (happen-
ing) but through a should-be, which would not be the case if reason without
sensibility (as the subjective condition of its application to objects of nature)
were considered, as far as its causality is concerned, as a cause in an intelli-
gible world, corresponding completely with the moral law, where there would
be no distinction between what should be done and what is done, between
a practical law concerning that which is possible through us and the theoret-
ical law concerning that which is actual through us. (Kant 2002: 273 [A 338–
9/B 342-3])

Compared to the previously cited passages, which are pleasantly clear, this
one seems obscure and needs some explanation. What Kant seems to mean, as
far as our topic is concerned, is that the Ought (or the moral law) expresses what
 6 Philosophy of Law

is morally necessary, in contradistinction to the natural law which expresses

what is physically necessary. That which is morally necessary may be the case
or may not be the case; only in an “intelligible world,” that is, an ideal world in
which all norms are complied with (a deontically perfect world), is everything
that ought to be the case really the case – norms and natural laws would be one.
And Kant insinuates some other important points that he elaborates elsewhere
and which have an impact on Kelsen’s conception: The Ought emanates from
reason; that is why it is on the one hand subjective (having its origin in a faculty
of mind) and on the other hand objective in the sense of being ‘reasonable.’ So
(pure) reason has two different functions – a theoretical (speculative) one,
directed at “the cognition of the object up to the highest a priori principles,”
and a practical one, directed at “the determination of the will with respect to the
final and complete end” (Kant 2015: 97 [A 216]). Further, the Ought is expres-
sive of “freedom,” defined as independence from causal laws. Accordingly, the
Ought-laws are contrasted with natural laws; and by complying with norms,
a “causality by freedom” originates.

2.2.2 The Primacy of Ought over Is

Although Is and Ought have to be kept apart, there are, for Kant, at least two
relations between them where the Ought is logically prior to the Is. First, there is
a “primacy” of practical reason over theoretical reason. Second, understand-
ing’s activity of judging presupposes a minimum of autonomy and, thus, of
normativity.
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

2.2.2.1 The Primacy of Practical Reason

Kant explains the possibility of a primacy of the practical function of pure
reason over its theoretical function as follows:

If practical reason may not assume and think as given anything further than
what speculative reason of itself could offer it from its insight, the latter has
primacy. Supposing, however, that practical reason has of itself original
a priori principles with which certain theoretical positions are inseparably
connected, while these are withdrawn from any possible insight of specula-
tive reason (although they must not contradict it): then the question is, which
interest is supreme (not, which must give way, for one does not necessarily
conflict with the other)?4 (Kant 2015: 97 [A 216])

And the “practical interest” of reason overrides, because it involves original

principles a priori which are practically necessary and insofar proven, but

4
“Interest,” in this context, is defined by Kant as “a principle that contains the condition under
which alone [the] exercise [of a faculty of mind] is promoted” (Kant 2015: 97 [A 216]).
 Hans Kelsen’s Normativism 7

theoretically undecidable; they consist of the moral law, which is a “fact of

reason,” and its presuppositions (the “practical postulates,” freedom, immedi-
ately, and immortality of the soul and God in a mediated way) – or so I read
Kant.5 This is a metaphysically demanding relation between Is and Ought which
need not be discussed any further in the present context, because it presupposes
the complex conceptual edifice of Kant’s explanation of reason and has no
impact on Kelsen’s theory.

2.2.2.2 The Spontaneity of Understanding

There is, however, a less obvious role for normativity in Kant’s Critique of
Pure Reason which also indicates a primacy of Ought over Is. Kant’s explan-
ation of ‘understanding’ (Verstand) is probably the first major attempt at
a normativistic theory of cognition. In fact, it is easy to see why some
conception of normativity, or rules, should be fundamental to Kant’s theoret-
ical philosophy as soon as one considers the argument that triggered his
transcendental idealism. To be sure, the Critique of Pure Reason sets out to
vindicate the possibility of synthetic judgments a priori, i. e. of judgments that
extend our knowledge but which are not dependent on experience; yet Kant’s
transcendental idealism is foremost an answer to the aporias of philosophical
realism, that is, of a position according to which cognition deals with objects
which are independent of it:

Truth, it is said, consists in the agreement of cognition with its object. In

consequence of this mere nominal definition, my cognition, to count as true,
is supposed to agree with its object. Now I can compare the object with my
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

cognition, however, only by cognizing it. Hence my cognition is supposed to

confirm itself, which is far short of being sufficient for truth. For since the
object is outside me, the cognition in me, all I can ever pass judgement on is
whether my cognition of the object agrees with my cognition of the object.
The ancients called such a circle in explanation a diallelon. And actually the
logicians were always reproached with this mistake by the sceptics, who
observed that with this explanation of truth it is just as when someone makes
a statement before a court and in doing so appeals to a witness with whom no
one is acquainted, but who wants to establish his credibility by maintaining
that the one who called him as witness is an honest man. (Kant 1992: 557–8)

Kant tackles this problem not by succumbing to skepticism, but by dissolving

metaphysics into epistemology: “The proud name of an ontology . . . must give
way to the modest one of a mere analytic of the pure understanding” (Kant
1998: 57–8 [A 247/B 303–4]). He dismisses the idea that the truth and object-
ivity of judgments are guaranteed by their referring to a transcendent reality and
5
This topic is discussed elaborately in Willaschek 2009.
 8 Philosophy of Law

holds that they are instead the result of an interaction between receptive or
passive sensibility and the spontaneous activity of understanding (Verstand).
Other candidates not being available, this spontaneity is made up by under-
standing’s activity – that is, thinking or judging – being guided by rules, and by
the judging subject’s competence to distinguish correct from incorrect applica-
tions of these rules.
There are quite a number of passages where this indispensable normativist
level of Kant’s theory emerges – for example, when he explains concepts as
rules for uniting representations, or when he defines understanding generally as
the “faculty of rules” (Vermögen der Regeln) (Kant 1998: 242 [A 126]). In fact,
Kant’s theoretical philosophy might presuppose a solution of the third antinomy
(Willaschek 2010: 165–84). That is, in short, the cognitive constitution of the
world of objects might presuppose some form of freedom of will, or
normativity.6
Against this background, it is not astonishing that, for Kant, logic is
concerned with normative rules: “In logic, however, the question is not
about contingent but about necessary rules; not how we do think, but how
we ought to think” (Kant 1992: 529). In a similar vein, borrowing from the
language of jurists, he distinguishes between questions of what is lawful (quid
juris) and questions of fact (quid facti) (Kant 1998: 219–20 [A 84/B 116]), that
is, between empirical questions and questions of justification, or genesis and
validity.
What is noteworthy about this kind of ‘theoretical normativity’ is that, unlike
moral normativity, it does its work implicitly. Kant describes it in a way which is
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

close to Wittgenstein’s explanation of the phenomenon of rule following. In the

Jäsche-Logik, Kant holds:

The exercise of our powers also takes place according to rules that we follow,
unconscious of them at first, until we gradually arrive at cognition of them
through experiments and lengthy use of our powers, indeed, until we finally
become so familiar with them that it costs us much effort to think them in
abstracto. Thus universal grammar is the form of a language in general, for
example. One speaks without being acquainted with grammar, however; and

6
Cf., in the same direction, Walker 2017: 205: “So, after all, does not theoretical reasoning provide
as sound a basis for presupposing freedom as practical reasoning does? [Kant] seems to be
suggesting this in the Groundwork, when he says ‘one cannot possibly think of a reason that
would consciously receive direction from any other quarter with respect to its judgments, since the
subject would then attribute the determination of his judgment not to his reason but to an impulse.
Reason must regard itself as the author of its principles independently of alien influences,
consequently, as practical reason or as the will of a rational being, it must be regarded of itself
as free.’ This makes it seem as though the same case can be made for autonomy in judging as in
acting, though he is here appealing only to how one must think of oneself. One might wonder why
he never develops the idea.”
 Hans Kelsen’s Normativism 9

he who speaks without being acquainted with it does actually have a grammar
and speaks according to rules, but ones of which he is not himself conscious.
Like all our powers, the understanding [Verstand] in particular is bound in its
actions to rules, which we can investigate. (Kant 1992: 527 [A 2])

The most remarkable feature of these rules is that, although they can in
principle be made explicit or “thought of in abstracto,” in their basic form
they are not ‘objects’ to be applied in certain situations; rather, they are just
implicit in certain acts: This is the only interpretation that takes care of Kant’s
thesis that you can have a grammar without knowing it. It anticipates
Wittgenstein’s famous dictum: “When I follow the rule, I do not choose.
I follow the rule blindly” (Wittgenstein 2009: 92 [219]).
This interpretation is underpinned by Kant’s text on schematism, where he
discusses the problem of how an intuition might be subsumed under a concept.
That poses a problem because intuitions and concepts are heterogeneous
(ungleichartig), stemming from different cognitive faculties. Concepts – or
the judgments which employ them – are the result of exerting the spontaneous
faculty of understanding, while intuitions stem from the passive faculty of
sensibility. Kant’s solution is that an intuition and a concept can be related to
each other by applying a scheme, which is a

rule for the determination of our intuition, in accordance with some

specific universal concept. The concept “dog” signifies a rule according
to which my imagination can delineate the figure of a four-footed animal
in a general manner, without limitation to any single determinate figure
such as experience, or any possible image that I can represent in concreto,
actually represents. This schematism of our understanding, in its applica-
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

tion to appearances and their mere form, is an art concealed in the depths
of the human soul, whose real modes of activity nature is hardly ever
likely to allow us to discover, and to have open to our gaze. (Kant 1998:
273–4 [A 141/B 180–1])

Concepts, therefore, signify rules, and schematism guides their application in

a way which is not a possible subject of reflection. Accordingly, in the introduc-
tion to the chapter on schematism Kant says that if logic claims to show,
universally,

how we are to subsume under these rules, that is, to distinguish whether
something does or does not come under them, that could only be by means of
another rule. This, in turn, for the very reason that it is a rule, again demands
guidance from judgment. And thus it appears that, though understanding is
capable of being instructed, and of being equipped with rules, judgment is
a peculiar talent which can be practised only, and cannot be taught. (Kant
1998: 268 [A 133/B 172])
 10 Philosophy of Law

Some lines later on Kant maintains that a physician, lawyer, or politician

may have at command many excellent pathological, legal, or political rules,

even to the degree that he may become a profound teacher of them, and yet,
none the less, may stumble in their application. For, although admirable in
understanding, he may be wanting in natural power of judgment. . . . Or the
error may be due to his not having, through examples and actual practice,
adequate training for this particular act of judgment. (Kant 1998: 268–9 [A
134/B 173])

This is, again, a striking anticipation of Wittgenstein’s rule-following con-

siderations – even as far as the terminology is concerned. Kant recognizes the
problem of applying a rule, if taken to be an object, to its “material”; he
discusses the problem of an infinite regress of rules/interpretations, and he
deems a competent application of a rule to be a question of “training.” These
arguments are at the core of Kant’s theoretical philosophy, because any cogni-
tion of an object presupposes the correct application of the categories; and
categories, being pure concepts of understanding, are, like all concepts, rules
which are applied in the procedure of schematizing. Since, for Kant, classical
ontology dissolves into an analytic of understanding, we might say that, for
a Kantian-minded philosophy, implicit normativity is one of the most basic
elements of any metaphysics.

2.2.3 Relations between Is and Ought

Although Is and Ought are fundamentally different functions of reason, there
are, for Kant, several relations between them. First, as was shown in
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

Section 2.2.2, the “primacy” of practical reason postulates the existence of

freedom, the immortality of the soul, and God which, for theoretical reason,
were just a possibility. Second, the cognition of any natural object (and thus the
possibility of nature itself) presupposes a set of implicit rules guiding cognition,
the ‘theoretical Ought.’ Third, there is a structural analogy between laws of
nature and practical laws. Fourth, the categorical imperative is a rule guiding the
choice of maxims, thus influencing the empirical will. Fifth, any act that is
exerted by following such a maxim, which is in accordance with the categorical
imperative, starts a new chain of “causality from freedom.” Sixth, any Ought
aims at producing a certain state of affairs, and this state of affairs must be
possible:

Now although there is an incalculable gulf fixed between the domain of the
concept of nature, as the sensible, and the domain of the concept of freedom,
as the supersensible, so that from the former to the latter (thus by means of the
theoretical use of reason) no transition is possible, just as if there were so
 Hans Kelsen’s Normativism 11

many different worlds, the first of which can have no influence on the second:
yet the latter should have an influence on the former, namely the concept of
freedom should make the end that is imposed by its laws real in the sensible
world; and nature must consequently also be able to be conceived in such
a way that the lawfulness of its form is at least in agreement with the
possibility of the ends that are to be realized in it in accordance with the
laws of freedom. (Kant 2002: 63 [A/B XIX])

2.3 Normativity and Validity: Lotze and the Neo-Kantians

Kant’s thesis of a primacy of practical reason was taken up and radicalized in the
highly idiosyncratic theory of Johann Gottlieb Fichte (1762–1814), who influ-
enced especially the neo-Kantians of the Baden school, but made no important
contribution to elucidating the concepts of normativity and validity. However,
the theory of validity of Hermann Lotze (1817–81) had an enormous impact on
neo-Kantian and later analytical philosophy, influencing among others Wilhelm
Windelband, Heinrich Rickert, Edmund Husserl, and Gottlob Frege – and
informing Kelsen’s conception of normative validity. Lotze’s influence lingers
on in current normativistic philosophy; accordingly, he has aptly been called
“the grandfather of the concept of normativity, which has become such a mantra
in contemporary philosophy” (Beiser 2013: 127).

2.3.1 Hermann Lotze

Both a physician and a philosopher, Lotze published his main works in the
middle of the nineteenth century, at a time when philosophical idealism had lost
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

ground in Germany (as had philosophy in general), natural sciences were more
and more successful, and psychology was getting established as a science.
Lotze’s ambition was to reconcile natural sciences and idealism in a system
which he called “teleological idealism.” And like Kant and Fichte, although in
a different way, Lotze assumes a primacy of practical over theoretical
philosophy.7 In an early text, he defines his position as follows:

[The appearances] come and emerge, held and connected not by some exter-
nal object, but by their interrelation, which a higher power without needing
material forces for its acts has prescribed. Our study has dissolved into this
teleological idealism. . . . [T]he beginning of metaphysics is not in itself, but
in ethics. (Lotze 1841: 329, my translation)

The first part of the quotation sounds almost Kantian; the difference is
that objectivity is not derived from the correct application of the

7
On this topic, see Thieme 1887.
 12 Philosophy of Law

concepts of understanding, but from a higher power which Lotze labels

the value of “the good” (das Gute) (Lotze 1841: 324–6). Cognition aims
at this value; that is why Lotze’s theory may be termed a ‘teleological’
variant of idealism.
An important means in Lotze’s endeavor to reconcile natural sciences and
idealism is the concept of “validity,” in contradistinction to “existence.”
Existence is the way natural objects are given; validity is the normative way
values, truths, or true judgments are given. It is a basic, undefinable meta-
physical concept: “[W]e must not ask what in its turn is meant by Validity,
with any idea that the meaning which the word conveys clearly to us can be
deduced from some different conception” (Lotze 1884: 440).
Validity is not subjected to space and time. Its normative character is not
stressed by Lotze; however, it is a consequence of his combination of
idealism with the view that “truths” or true judgments must not be hyposta-
tized. And he maintains, when dealing with Descartes, that there is a “matter
or content of our cogitation, which supplies not only the original datum from
which we start, but the sole source from which that which we ought to think
or that which we cannot but think can be derived” (Lotze 1884: 454). This
normative “matter of our cogitation” is a “valid truth.” From such meta-
physical priority of normativity, Lotze shapes one of the first powerful
normativistic post-Kantian arguments against psychologism (i.e. the view
that logic and epistemology might be reduced to psychology):

What particular tone of mind is required for successful thinking, how

the attention is to be kept up, distraction avoided, torpidity stimulated,
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

precipitation checked, all these are questions which no more belong to

the field of logic than do enquiries about the origin of our sense-
impressions and the conditions under which consciousness in general
and conscious activity is possible. We may presuppose the existence of
all these things, of perceptions, ideas, and their connexions according
to the laws of a psychical mechanism, but logic only begins with the
conviction that the matter cannot end here; the conviction that between
the combination of ideas, however they may have originated, there is
a difference between truth and untruth, and that there are forms to
which these combinations ought to answer and laws which they ought
to obey. It is true that we may attempt by a psychological investigation
to explain the origin of this authoritative consciousness itself; but the
only standard by which the correctness of our results could be meas-
ured would be one set up by the very consciousness to be investigated.
The first thing, then, that has to be ascertained is, what the contents of
this authoritative conviction are; the history of its growth can only
have the second place, and even then must conform to requirements of
its own imposing. (Lotze 1884: 7–8)
 Hans Kelsen’s Normativism 13

The importance of this passage cannot be overvalued; it signifies a different

thrust of the argument from the dualism of Is and Ought. While Hume and, in his
practical philosophy, Kant, made use of the argument to delimit the practical
from the theoretical sphere, Lotze uses it to show that there cannot be a purely
naturalist or psychologist explanation of cognition, because any combination of
ideas in a judgment presupposes that this combination is either correct or
incorrect, valid or invalid, so that it is subjected to normative standards. Kant
advocated a similar thesis (as shown in Section 2.2.2), but it does not explicitly
play an important role in his philosophy. This is not so in Lotze’s conception:
When he developed his theory in the middle of the nineteenth century, philoso-
phy faced its own fragmentation and the onslaught of the successful natural
sciences. Normativism seemed to be the only way to preserve the autonomy of
the humanities and philosophy itself; normative ‘validity’ designated a realm
exempt from the grasp of the natural sciences because it was, in fact, a necessary
presupposition for any scientific judgment.
If one takes into account this central historical role of normativism, it is not
surprising that Lotze does not focus as much on the ‘moral’ or practical Ought.
Although he planned to write a philosophy of morality, aesthetics, and religion
as a third part of his System der Philosophie, it was never accomplished.
There is, however, one point where Lotze differs from the Kantian tradition
as regards normativity. While Kant’s notion of normativity is, foremost, one of
Ought-rules which are not objects of any kind, Lotze explains the notion of
validity by referring to Plato’s doctrine of ideas. From the notion that the reality
of a truth consists in its validity, according to Lotze,
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

some light is thrown on a surprising statement which is handed down to us in

the history of Philosophy. Plato, we are told, ascribed to the Ideas of which he
had achieved the conception an existence apart from, things, and yet, as these
same critics tell us, of like kind with the existence of things. . . . The truth
which Plato intended to teach is no other than that which we have just been
expounding, that is to say, the validity of truths as such. . . . But the Greek
language then as afterwards was wanting in an expression for this conception
of Validity as a form of Reality not including Being or Existence. (Lotze
1884: 440–1)

It was this assimilation of ‘valid truths’ to Platonic ideas that led some neo-
Kantians, notably Heinrich Rickert, to ascribe validity – as will be shown in
Section 2.3.2 – to “values,” which are not norms directed at human subjectivity,
but which have an objective reality independent from it. Although this does not
seem to be quite in accordance with Lotze’s intentions, he certainly encouraged
it, not only by referring to Plato’s philosophy but also by maintaining that some
“good” or “truth” is the datum from which an Ought can be derived. This seems
 14 Philosophy of Law

to imply that validity is not identical with an Ought, but is, after all, some kind of
being.

2.3.2 The Baden School: Wilhelm Windelband and Heinrich Rickert

Normativism figures prominently in parts of neo-Kantian philosophy. The term
‘neo-Kantianism’ is not used here to denote a philosophical theory which
somehow attempts to reconstruct or incorporate theorems from Kant’s writings;
rather, it specifically denotes a philosophical movement which dominated
German philosophy between approximately 1870 and 1920, peaking at around
1900. It is characterized by two notorious slogans: ‘Zurück zu Kant’ (back to
Kant),8 and ‘Kant verstehen heißt über ihn hinausgehen’ (to understand Kant is
to go beyond him) (Windelband 1907: IV). They express the lowest common
denominator of neo-Kantianism: On the one hand, philosophy must, after
decades of confusion and decay, be reformed by returning to Kant’s philosophy;
on the other hand, Kant’s philosophy itself is not thought through in all its parts,
so it needs to be complemented and improved.
Most neo-Kantians are antirealists – more so than Kant himself. For them, the
objective world is not given independently of cognition; rather, it is ‘constituted’
by it, and they do not assign any function to the notion of a thing-in-itself.9 It is not
ontology but epistemology that is at the centre of most neo-Kantian philosophies.
Accordingly, the notion of ‘judgment’ is one of their most fundamental elements:
It is the basic unit of cognition. By comparison, the sphere of sensibility or
intuition is notoriously neglected by most neo-Kantians.
There are two main schools of neo-Kantianism: the Marburg school, and the
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

Baden school. The Marburg school, represented principally by Hermann Cohen

and Paul Natorp, holds transcendental philosophy in a Kantian vein to be
a theory of the exact sciences. It accepts the judgments of the culturally
established and successfully operating natural sciences as a ‘given’ fact and
critically explores the conditions of their possibility – this is the Marburg
version of the ‘transcendental method.’ It has been aptly termed “a method-
critical theory of science with extended policing powers” (Hoeschen 1999:
113). This brand of neo-Kantianism is of great importance for the further
development of philosophy. It influenced logical positivism and analytical
philosophy; tying philosophy to a comprehensive sociocultural practice pos-
sibly had an impact on Husserl’s concept of Lebenswelt, and even on

8
Otto Liebmann enthusiastically ended each chapter of his influential book Kant und die Epigonen
(Schober: Stuttgart, 1865) with this slogan.
9
See Windelband’s harsh dictum: “From the standpoint of epistemology, the thing-in-itself is
a completely meaningless and useless, confusing and annoying fiction” (quoted in Beiser 2014:
524).
 Hans Kelsen’s Normativism 15

Wittgenstein’s later concept of Lebensform. And it is highly important for the

self-conception of the Pure Theory, being responsible for Kelsen’s presenting it
as a theory of legal dogmatics, performing a presuppositional analysis.10 The
theory of the basic norm makes sense only when seen as an outcome of applying
Cohen’s ‘transcendental method.’ Yet the Marburg school is of no significance
for the development of normativism, so it can be disregarded in the present
context. The Baden school, on the other hand, represented mainly by Wilhelm
Windelband and Heinrich Rickert, explicitly defines philosophy as a theory of
‘values’ in the wake of Lotze, thus giving it a normativist twist.
In a way, both schools of thought defended philosophy, in different ways,
against the naturalist menace of being ‘swallowed up’ by the expanding empir-
ical sciences: the Baden school, by demarcating an irreducible sphere of
normative validity, and the Marburg school, by lifting philosophy onto a meta-
level vis-à-vis the sciences.

2.3.2.1 Wilhelm Windelband

Wilhelm Windelband (1848–1915), the founder of the Baden school, was
tutored by Lotze when writing his thesis on ‘certainty,’ which was published
in 1870. He started as a historian of philosophy and never wrote a monograph
containing his philosophical thoughts in a systematic way. His views are mainly
contained somewhat rhapsodically in a collection of essays titled Präludien
(Preludes).
Windelband follows Kant in maintaining that metaphysics must dissolve into
epistemology, and that normativism must rule epistemology. In fact, he is a lot
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

more explicit regarding these points. He takes Kantian philosophy to focus on

the question how reference to objects is possible, and he takes Kant’s answer to
be that knowledge is not brought about by the correspondence of
a representation with an object, but by the cognizing act being in accordance
with the rules guiding the way representations have to be related to each other;
for Windelband, Kant replaced the concept of object with that of a rule
(Windelband 1907: 157, 160).11 (The terms ‘Regel’ [rule] and ‘Norm’ [norm]
are used by Windelband interchangeably.12) Truth is normativity of thinking; it
lies in thinking’s compliance with valid norms – the Lotzean origin of
Windelband’s philosophy is obvious. Likewise, his anti-psychologism: He
maintains that “representations come and go, and psychology may explain
how this comes about: philosophy analyses their value from the critical stand-
point of truth” (Windelband 1907: 48, my translation).

10 11
On this status, see Heidemann 2020: 81–98. On this topic, see further Beiser 2014: 497.
12
Cf. Beiser 2014: 497.
 16 Philosophy of Law

‘Critical’ (i.e. Kantian) philosophy, accordingly, boils down to a science of

necessary and universally valid valuations (Windelband 1907: 49). However,
Windelband is not quite clear on the problem of how to justify the rules guiding
cognition and the ‘basic norms’ or axioms of the three disciplines of philoso-
phy – science, ethics, and aesthetics (whose basic norms are, respectively, truth,
goodness, and beauty; they apparently are values in a Lotzean sense and take the
place of the Kantian ‘thing-in-itself’). He seems to have different strategies.13
On the one hand, he adopts the approach of the Marburg school of neo-
Kantianism in considering philosophy as a theory of science (Windelband
1907: 42); ‘science’ in this sense being regarded as a successful and sociocul-
turally established practice with a claim to objectivity. Taking science as
‘given,’ it follows that both the basic norms of cognition and the rules guiding
cognition can be extracted (though not justified) by analyzing scientific practice.
On the other hand, Windelband adopts a teleological and hypothetical approach:
The rules for guiding cognition are (only) valid for anyone who aims at the basic
values of truth, goodness, or beauty. In an early text, he further maintains that
these values are a necessary consequence of the unity of self-consciousness
(Windelband 1874: 90)14 (which sounds both Kantian and convincing;
Windelband possibly later eliminated this thesis in order to purge his theory
from any seemingly psychologist elements). In yet another context, he main-
tains that the basic values can be derived as presuppositions of the ‘normal
consciousness’ (Normalbewusstsein) (Windelband 1907: 67–9), thus exploit-
ing, in a bit doubtful manner, the ambivalence of the term ‘normal,’ meaning
both ‘average’ and ‘according to norms’.
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

It is noteworthy that normativity seems to be important for Windelband only

in connection with the analysis of cognition; the moral Ought does not play any
important part in his philosophy. While it seems an accident that Lotze did not
write a systematic text on ethics, it seems to be part of Windelband’s program
that he neglects this domain. Just one paper from Präludien deals with the moral
Ought, and, without elaborating the difference between moral and cognitive
normativity, Windelband simply expounds his position that the fundamental
moral Ought is a principle even more formal than Kant’s categorical imperative;
it is the rule that one ought to do one’s duty (Windelband 1907: 380–414). This
can be explained, in part, by the aim of neo-Kantian philosophy to instrumen-
talize normativity to thwart any attempt of the natural sciences (including
psychology) to usurp the domain of philosophy and the humanities, partly
through the insight that philosophy was not able to yield specific ‘moral content’
anyway.

13 14
Cf. Beiser 2014: 506–7. See further Köhnke 1993: 361.
 Hans Kelsen’s Normativism 17

One further trait of Windelband’s theory is important for normativism. He

distinguishes a “critical” from a “genetical” method. This distinction is based on
Kant’s differentiation between quaestio facti and quaestio juris, mentioned in
Section 2.2.2. It concerns the method of philosophy, seen as a theory of validity of
axioms. The genetic method looks into the origins of these axioms by employing
the methods of empirical sciences; it is based on psychology and cultural history.
But employing the genetic method can never answer the question why an axiom is
valid, it can only yield an answer to the question why an axiom is taken to be
valid; the axioms are, for this method, “factual ways of conceiving something
[Auffassungsweisen], which took shape during the development of human ideas,
feelings, and volitions” (Windelband 1907: 328, my translation). By contrast, the
critical method takes these axioms, “no matter how far they are factually recog-
nized, to be norms which are valid under the presuppositions that thinking aims at
being true, willing aims at being good, and feeling aims at appreciating beauty”
(Windelband 1907: 328, my translation). It is the only true philosophical method.
Accordingly, philosophy’s task is it to make explicit, using the critical method, the
norms necessarily followed when pursuing the aims of finding out what is true,
good, and beautiful – like his teacher Lotze, Windelband combines a normative
approach with a teleological conception.

2.3.2.2 Heinrich Rickert

Heinrich Rickert (1863–1936), in turn, wrote his thesis under the supervision of
Windelband, so that he is in several respects the philosophical ‘grandson’ of
Lotze. His theory can be regarded as a systematic development of Windelband’s
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

philosophy. From the beginning, he is not concerned with ethics or practical

philosophy, but strives to dissolve metaphysics into epistemology, and epistem-
ology into a philosophy of values (Wertphilosophie).
Rickert is probably best known for elaborating Windelband’s distinction
between nomothetic disciplines, which concentrate on general laws, and idio-
graphic disciplines, which examine particular events in the light of certain
values, by distinguishing natural sciences from cultural sciences. This theorem
strongly influenced Gustav Radbruch (the other great legal philosopher of the
German language in the first half of the twentieth century besides Kelsen) and
was successful enough to provoke Kelsen to write a lengthy paper against it, Die
Rechtswissenschaft als Norm- oder als Kulturwissenschaft (‘Legal science as
a science of norms, or as a science of culture’), which was published in 1916.
As far as his theoretical philosophy is concerned, Rickert advocates a strictly
nonrealist position. For him, there is no objective reality that is logically prior to the
sphere of judgments; reality is constituted in objective cognitive judgments. To
 18 Philosophy of Law

grasp the notion of an objective judgment, according to Rickert, it is necessary to

distinguish between the psychical act of judging and its meaning-content. There are
two kinds of meaning-contents which might be correlated to the act of judging. The
subjective or “immanent” meaning-content is “that which is inherent to the real acts
of judging by virtue of their capacity” (Rickert 1921: 143). By contrast, the
objective or transcendent meaning-content is something that the concrete individ-
ual performance of an act of judgment ‘aims at’ and which is external to it. Rickert
characterizes it as follows:

That structure which we have in mind when we are talking about the objective
meaning-content is what is sometimes called the “truth” of the judgment, its
content per se which is not immanent to the judging act of the subject but
which must be thought of as something independent of it and attained by it,
and which all of us mean and understand wherever we judge or assert
something as true at all. (Rickert 1921: 143, my translation)

Such an anti-psychologist interpretation of the objective meaning-content of

a judgment seems to require a Platonic ‘hyper-realism’; that is, an ontology of
abstract meaning-contents. But Rickert wants to avoid such a metaphysics,
which is scarcely compatible with critical philosophy, and simply shifts the
problem of realism to a different level. His theory strives for a third alternative
beyond psychologism and Platonism. To achieve this aim, Rickert – following
Lotze (Rickert 1921: 229) – employs the concepts of validity (Geltung), Ought
(Sollen), and theoretical value (theoretischer Wert). He argues as follows: The
transcendent meaning-content of a judgment is objective by being necessarily
correlated to the theoretical value of truth; truth is the ‘form’ of this meaning-
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

content, as it were. The theoretical value, and with it the objective meaning-
content, does not exist like an object, instead, it is valid. It gives rise to an Ought,
namely, a demand to think in a certain way.
Validity must not be understood as an ideal mode of being of some kind of
nonsensual reality (Rickert 1921: 247); rather, it is the indefinable way in which
normativity is given, and which is prior to any existence:

[The concept of existence is not the only one] under which something can be
subsumed; besides, there is the concept of value. We use this word which
denotes a concept that can be defined just as little as the concept of existence,
for structures that do not exist but which, nevertheless, are “something”, and
this can best be expressed by saying that they are valid. (Rickert 1921: 229–
30, my translation)

Rickert argues in different ways for the case that validity has not only to be
distinguished from existence, but is also prior to it; the most convincing argu-
ment runs as follows:
 Hans Kelsen’s Normativism 19

[The transcendent meaning] lies . . . “above” and “before” everything that

is existing and cannot be captured by any ontology. This simply follows
from the fact that any recognition “that something exists”, presupposes the
meaning inherent to the sentence, that something exists, no matter whether
it is a physical or a psychical, a real or an ideal, a sensual or supersensual,
a given or an inferred existing. If the meaning of the existential sentence is
not true, then nothing exists at all. Therefore, the meaning cannot be part
of the things existing, it must be prior to it. (Rickert 1921: 229, my
translation)

An object x exists only insofar as there is an objective meaning-content of

a judgment (so that a sentence expressing the judgment is true), which affirms
the existence of x. Therefore, the validity of the objective meaning-content is
a necessary logical presupposition of any existence.
Once you accept its premises, the argument seems to be sound. It estab-
lishes, at least for transcendental idealism, the primacy of Ought, or value,
over Is. The problem with it is that Rickert is not able to balance the relation
between the immanent and the transcendent meaning-content in
a satisfactory manner. This matter deserves a digression. On the one hand,
this problem is not only the crux of neo-Kantian Wertphilosophie, its des-
cendants also haunt parts of contemporary philosophy. On the other hand,
Kelsen refers to it in some crucial passages when explaining the notions of
Ought and, especially, validity (see Section 3.2.4.3 below). It can best be
assessed by examining Husserl’s (because he is the main protagonist of anti-
psychologism), Windelband’s, and Rickert’s attempts to answer the question
whether and how logical laws and cognitive content might be regarded as
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

being normative. Philosophically, it is the most thrilling aspect of ‘old-

school’ normativism.

2.3.3 Digression: Immanence and Transcendence, Norm and Value

Edmund Husserl (1859–1938), one of the main protagonists of anti-

psychologism, seems to worry that taking logical principles to be ‘prescriptive’
for the acts of a factually existing subject may impair their absolute character
and necessarily drag them back into the mire of psychologism. Therefore, he
distinguishes a ‘pure’ logic, which is not a normative but a theoretical discip-
line, from an ‘applied’ logic, which determines how we ought to think. Logics,
understood as a set of norms for thinking, seems to belong to a strange inter-
mediate region between psychological acts and the transcendent sphere of pure
logics, while the latter does not contain any normative Ought. Husserl admits
that the laws of pure logic might, like every general truth, serve as “a general
norm for judging correctly.” But
 20 Philosophy of Law

by their own nature, they are not normative, but theoretical truths, and as such
they can serve for regulating the activity of judging just as much as truths
from any other disciplines. . . . The anti-psychologists erred in taking the
regulation of cognition as the essence of the laws of logic. . . . It was
overlooked that the so-called logical principles as such are not norms, but
may be used as norms. With respect to this function we got used to speaking
of laws of thinking, and so it seemed as if these laws had a psychological
content, and as if the difference to the so-called laws of psychology was just
that they regulate, while the other laws of psychology do not. On the other
hand, the psychologists erred. . . . It is a matter of course that every general
truth, be it of psychology or not, founds a rule of correct judging; and thus not
only the possibility, but the existence of rules for judging is confirmed, which
are not founded in psychology. (Husserl 1900: 157, my translation)

Windelband holds in a similar way that

the pure logic or logic in the narrow sense is usually defined as the theory of
the forms of thinking. But this has to be specified; it is concerned with the
forms of correct thinking, which are a selection from the psychologically
possible forms of moves of mental representations. Logic does not teach how
we really think, but how we ought to think, if we want to think correctly. . . .
One has to keep in mind, however that the validity of these forms has, in the
end, to be totally independent from the quest for knowledge of an empirical
and especially the human consciousness. . . . From this, a fundamental duality
of all logical laws follows: On the one hand, they are rules for empirical
consciousness which all thinking that is directed toward truth should comply
with; on the other hand, their inner and autonomous meaning and being is
quite independent from the question whether there are factual mental pro-
cesses of representing which conform to them. The former might be called
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

their validity in itself [Geltung an sich], the latter their validity for us [Geltung
für uns]; where “us” does not just mean us human beings, but all individual
subjects who – like us – distinguish, in their representations, between true and
false, correct and incorrect. Seen from “us,” logic is an “Ought” – but this
Ought must be grounded in something whose validity subsists in itself, and
which turns into a norm or Ought only if it is related to a consciousness
capable of erring. (Windelband 1912: 17–18, my translation)

Finally, Rickert devotes an almost book-length text (Rickert 1909) to show-

ing – in a manner comparable to that of Windelband – that although normativity
is a fundamental element of cognition, it is not simply by applying norms for
cognition that the correctness of judgments comes about. According to Rickert,
there are two different ways for cognitive theory to proceed. On the one hand, it
is possible to start at the psychical act of judging and explore the normative rules
necessarily involved in it. Rickert calls this procedure the “transcendental-
psychological way.” The problem with it is that its results are by necessity
relative to the act of judging; so that the norms recognized in the act of judging
 Hans Kelsen’s Normativism 21

are still somehow part of the psychological sphere – or so Rickert holds.

Therefore, the transcendental-psychological way has to be complemented by
a “transcendental-logical” way or method. The transcendental-logical method
does not start with the psychological act of judging, but with the transcendent
sphere of values from which the norms for judging are derived. The reason, for
Rickert, is that thinking must have an object which is not part of it, but external
to it; otherwise thinking would not have an object at all, because an object must
offer “resistance” to cognition.
Characteristic of the conceptions of Windelband and Rickert is that, in order
to ward off psychologism (and its associates, relativism and naturalism), they
think it necessary not to see Ought-normativity as basic, but to ground it in
something more abstract. In order for the norms of thinking not to be part of the
psychological sphere, they have to be derived not as a presupposition of the act
of thinking, but from values which rest in themselves, detached from human
affairs. And thus, though they fly the flag of transcendental idealism (and deny
a knowledge-independent existence of everyday objects), in the end
Windelband and Rickert run danger of falling prey to hyper-realism, that is,
of endorsing the thesis that there are transcendent, knowledge-independent,
values.
The way Rickert tries to deal with this consequence and to find a point
where the transcendent value and the immanent norm or Ought might have
contact, is remarkable for its meandering. The first step is to remove the
transcendent value from the realm of existing objects. This is made possible,
as shown in Section 2.3.2.2, by introducing the notion of ‘validity.’
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

The second step is to determine some graspable thing, which is not reducible
to a psychical entity and which, in a way, embodies the transcendent value.
This graspable object, for Rickert, is the true “sentence,” understood as
a series of signs (Rickert 1909: 194). Its meaning is “transcendent,” independ-
ent of cognitive acts, and has to be distinguished from the “immanent”
subjective meaning of these acts. Like Frege, Rickert calls this objective
meaning a “thought” (Gedanke) (Rickert 1909: 196). It is logically prior to
any existence, because – as was shown in Section 2.3.2.2 – for Rickert,
existence presupposes the validity of an objective-meaning content. Being
logically prior to existence, the true thought can only be made sense of as
a “theoretical value,” and, accordingly, the task of cognitive theory is to
determine its presuppositions (Rickert 1909: 207).
There is an obvious objection to this argumentation: Why should it be
possible to get at the transcendent meaning by simply connecting it to the
empirical object of a ‘sentence,’ to a series of signs? After all, the meaning of
a sentence is accessible only via the psychical act of ‘understanding’ it. So,
 22 Philosophy of Law

either the transcendent meaning is identical with the immanent meaning of an

act of understanding a true sentence, or it is, again, inaccessible.
Rickert tries to answer this objection, first, by maintaining vaguely that
transcendental psychology might bridge the gap; “we,” as practically free
beings, might autonomously recognize “the Ought for its own sake, and the
value for its own sake.” Only as practically free beings might we take posses-
sion of the realm of transcendent values (Rickert 1909: 222). But just a few
sentences later he seems to capitulate and to admit the validity of the objection:

Nevertheless, there might still be concerns about the worth of the transcen-
dental-psychological analysis, especially as regards its ability to bridge the
gap between the world of theoretical values and the psychical acts of think-
ing. To support these doubts, a very radical argument may be brought forward
that at this point the theory of cognition hits upon a problem which cannot be
solved at all. Meaning and being, value and reality, transcendence and
immanence conceptually exclude each other. Philosophy cannot overcome
this dualism, not even by introducing the conception of autonomously recog-
nising the values. (Rickert 1909: 222, my translation)

However, in the passage following this concession, Rickert maintains that

one should not give up over this problem. The seeming dualism of immanence
and transcendence is just

the necessary product of any reflection about cognition. We have cognition,

and grasp the truth at the same time. There is an immediacy of the unity of
meaning and being which is “experienced”. But, alas, by asserting that “we”
have and experience this “unity” we must destroy this immediate unity and
divide it into a real cognitive act and its non-real object. This division is
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

necessarily connected with the concept of cognition. We cannot even say that
there is this unity, because that would be cognition of the unity as unity, which
is impossible. The unity cannot be thought, it can only be experienced.
(Rickert 1909: 223–4, my translation)

This is no true solution. If the unity of immanence and transcendence is to be

proved by an unreflected “experience” of this unity in an act of cognition, then
transcendence is dependent on immanence. However the problem of the relation
between the immanent and the transcendent aspects of normativity might be
solved, it is certainly not by pointing to the dichotomy of immanent norm and
transcendent value.15 As meritorious as Lotze’s idea was, of demarcating the
domain of cognition, which cannot be ‘naturalized,’ by introducing the notion

15
The problem of the relation between immanence and transcendence troubles even so late
a descendant of neo-Kantianism as Jürgen Habermas. In Wahrheit und Rechtfertigung he holds
that “it is the aim of all justification to discover a truth which is independent of any justification”
(Habermas 1999: 53, my translation). This sentence is paradoxical not only at first sight.
 Hans Kelsen’s Normativism 23

of validity, it was not really helpful to his successors that he conceptualized

those elements to which validity might be attributed as transcendent and object-
like. A more plausible solution, faintly hinted at in the last Rickert quote above,
would be to go back to the notion of ‘implicit normativity,’ which can be found –
as was shown in Section 2.2.2 – in Kant’s philosophy and which is prominent in
Wittgenstein’s conception of rule-following in the Philosophical Investigations.

2.4 Summary
The genesis of normativism as outlined in this section might be summarized as
follows: Hume was the first major philosopher to make the dualism of Is and
Ought explicit, taking the Ought to be a matter of moral, religious, or political
propositions and holding Ought-sentences not to be objective. Kant took the
moral Ought to be expressive of practical reason (and, therefore, to be object-
ive). Although he explicitly dealt only with the moral Ought in an elaborate way,
there is a ‘theoretical’ Ought (recognized in any act of judging) in Kant’s theory
as well. Lotze gave an ontological touch to this theoretical Ought by taking
a truth to be the “content of cogitation” which is valid and from which an Ought,
or rules demanding that a subject should think in a certain way, can be derived.
This interpretation was taken up by Windelband in order to make the normativ-
ity of values the basic philosophical element and to forestall any psychologistic
explanation of cognition (while neglecting the moral or practical Ought).
Rickert systematized and radicalized Windelband’s approach by distinguishing
between the Ought directed at the judging person’s subjectivity and the even
more fundamental value of truth which is valid independent of any subjectivity.
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

His theory, however, hit upon new problems by distinguishing between

a “transcendent” Ought, embodied by values independent of any human cogni-
tion, and an “immanent” Ought, embodied by norms directing cognition which
somehow emanate from the transcendent Ought.

3 Hans Kelsen’s Normativism

Section 2 outlined the philosophical setting concerning normativism when
Kelsen entered the stage. It was the outcome of a deep, highly sophisticated,
multifaceted and not always commonsensical discussion over more than
a century.
Now, especially due to the writings of Stanley L. Paulson,16 it is largely
accepted that Kelsen’s legal-theoretical writings must be assigned different
phases to avoid grave inconsistencies and to do justice to their philosophical

16
See especially Paulson 1990; see further Kubeš 1980.
 24 Philosophy of Law

fundamentals. According to my count, there are four different phases, distin-

guishable by the philosophical basis of the writings assigned to them, and their
consequences (cf. Heidemann 2007). The first phase covers Kelsen’s habilita-
tion thesis, Hauptprobleme der Staatsrechtslehre (Main Problems of the Theory
of State Law) from 1911, and some subsequent shorter texts. In the writings of
this period, there are a lot of anticipations of theorems of what was later called
“Pure Theory of Law,” but there is no coherent philosophical or epistemological
underpinning as yet. Between about 1916 and 1935, Kelsen develops most of
the special tenets of the Pure Theory in full and grounds them in a specific neo-
Kantian epistemology. From around 1940 until 1960, he does without any
explicit philosophical foundation for his theory but relies on a rather naive
cognitive realism according to which the legal norm is an entity ‘given’ to
cognition; after 1960 he introduces a rather heterogeneous jumble of philosoph-
ical conceptions into his theory – especially from linguistic philosophy –and
couples them with a strict will-theory of norms.17

3.1 The First Phase: Juristic Constructivism

With his first major publication,18 Hauptprobleme der Staatsrechtslehre,
Kelsen did not emerge as a legal philosopher: He was trained as a jurist, and
he understood his text in the first instance as a methodology of legal dogmat-
ics. Its aim is to answer “preliminary questions” for any legal cognition
(Kelsen 1911: III). Kelsen presupposes as fact that there is an autonomous
legal cognition – that is, a cognition which is neither sociological nor ethical –
in which positive19 law is viewed as something that is objective and normative
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

(Kelsen 1911: VI, 44)20, namely, the cognition of practical jurists and legal
dogmatists. It is reconstructed from the internal perspective; Kelsen aims to
uncover its tacit presuppositions and to criticize its constructions (Kelsen
1911: V).
The methodological assumption underlying this enterprise is the distinction
between an explicative view, employed by causal sciences, and a normative
view, employed by legal science (that is, legal dogmatics). Both views exclude

17
For a detailed depiction of the genesis of Kelsen’s normativism, see Heidemann 1997.
18
He published a book on Dante’s theory of state six years earlier (Hans Kelsen, Die Staatslehre
des Dante Alighieri, Wiener Staatswissenschaftliche Studien, series 6, vol. 3 [1905]), but it does
not contain any traces of the later Pure Theory. Kelsen himself regarded it as “a student’s
unoriginal effort” (see Kelsen 2006: 37).
19
Kelsen takes for granted that the term ‘law’ covers exclusively the positive law, cf. Kelsen 1911:
367.
20
The ‘normative perspective’ of legal dogmatics is, for Kelsen, as much a postulate as a reality.
Calling legal dogmatics a ‘normative’ discipline only means that norms are its subject matter; it
does not mean that it posits norms (Kelsen 1911: VI–VII).
 Hans Kelsen’s Normativism 25

each other and must not be blended in one and the same science (Kelsen
1911: 6). Therefore, Kelsen aims at restricting juristic construction by eliminat-
ing from it any elements alien to the normative view – especially arguments
from psychology or sociology – in order to get at a logically closed system of
purely juridical fundamental concepts (Kelsen 1911: XI).
This alone, however, would not have been a radical innovation of contem-
porary conceptions of law and state. In fact, Kelsen himself – in a later text,
Allgemeine Staatslehre (1925) – saw his theory in the positivistic tradition of
Gerber and Laband, who applied Begriffsjurisprudenz (conceptual jurispru-
dence) to state law (Kelsen 1925: VII). And stressing the normativity of law
was not new either; Kelsen was preceded, among others, by Georg Jellinek who
was closely related to the Baden school of neo-Kantianism. Kelsen had attended
his seminars at Heidelberg.
So what was actually new about Kelsen’s approach? On the one hand, it
was the radicality with which he defined the dualism of Is and Ought and
attempted to exclude any ‘Is-elements’ from law. On the other hand, it was
his ‘semi-reductionism’ concerning legal concepts, that is, his attempt to
construct these concepts from the one central normative element of law, the
Rechtssatz.

3.1.1 The Dualism of Is and Ought and the Teleological Perspective

The basis of the distinction between the explicative and the normative point of
view is the dualism between Is and Ought. Accordingly, Kelsen emphasizes the
importance of separating these elements. He does not develop a fully fledged
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

argument; instead, Kelsen points to diverse lesser known or nonphilosophical

writers, like Georg Simmel and Arnold Kitz (Kelsen 1911: 7–8), or Julius
Kirchmann (Kelsen 1911: 19 fn.). The dualism is described in several different
ways:

(1) According to an epistemological explanation, Is and Ought are exclusive

and comprehensive Denkbestimmungen (determinants of thought), original
categories which cannot be derived from anything else (Kelsen 1911: 7). The
Ought is pure form; it is not necessarily connected with any content (Kelsen
1911: 10). It is a mode of thinking, like future, simple past, subjunctive, or
optative (Kelsen 1911: 7–8).

(2) According to an ontological explanation, Is and Ought are, rather, expres-

sive of “two separate worlds” (Kelsen 1911: 8). By constituting natural law and
norm, respectively, they constitute “reality” (understood as bodily or psychical
events), on the one hand, and “ideality,” on the other.
 26 Philosophy of Law

(3) According to a “formal-logical” explanation, the mutual independence of

Is and Ought manifests itself in the fact that, logically, an Ought can be justified
only by pointing to another Ought; and an Is can only be justified by pointing to
another Is. The chain of justification inside the respective spheres is infinite; and
it is just a “psychological need of limitation” which wants to ground the Ought-
chain on an Is, and the Is-chain on an Ought (Kelsen 1911: 9).
It is a special feature of Kelsen’s version of the Is–Ought dualism that he not
only maintains that an Ought cannot be derived from an Is; he also holds that an
Is cannot be derived from an Ought. And there is a relativization of the
dichotomy. In a prominent passage in the foreword to the Hauptprobleme,
Kelsen writes that the dualism of Is and Ought emanates from
a comprehensive dualistic world view according to which “I and world, soul
and body, subject and object, form and content” are unbridgeably separated. At
the same time, he relativizes the dualism by holding that the choice between
such a dualistic view and a monistic view is a matter of Weltanschauung (Kelsen
1911: 5–6).
A further candidate (besides Is and Ought) for a fundamental perspective is
dismissed by Kelsen rather easily: the teleological view. As was shown in
Sections 2.3.1 and 2.3.2 teleology was a dominant factor both in Lotze’s
philosophy and in the philosophy of the Baden school, where norms directed
at human acting and thinking are derived from abstract values which are the
telos of acting and thinking. Besides, teleological thinking dominated parts of
the juristic discourse. According to Kelsen, however, the teleological view is
a special case of the explicative view. It is directed at understanding occur-
rences, especially human acts, by the “means–end-scheme.” And the means–
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

end-scheme is just an application of the causal scheme: An end can only be

achieved by some means if this means can work as a cause to bring about the end
as its effect. Although it does not follow, according to Kelsen, that the teleo-
logical view can be reduced to the causal view; they just offer different
perspectives of the same principle which governs the explicative way of cogni-
tion: the principle of sufficient reason (Kelsen 1911: 57–63).

3.1.2 Relations between Is and Ought

Even though it is not possible to deduce an Ought from an Is, or an Is from an
Ought, there are several relations between them.

(1) Both the ways an Ought comes into existence and is ‘destroyed’ are part of
the sphere of Is (Kelsen 1911: 10). An example of this phenomenon is Jellinek’s
formula of the “normative force of the factual,” according to which custom can
produce a consciousness of Ought (Kelsen 1911: 9). However, it is not possible
 Hans Kelsen’s Normativism 27

to answer the question why something ought to be done, by pointing to custom

or psychical processes or something else from the factual sphere (Kelsen 1911:
9, 71).

(2) Is and Ought are related insofar as the Ought “aims”at an Is. It “wants to
bring about” changes in the sphere of Is. Accordingly, the “demands of the
Ought” (Sollanforderungen) are motives for the will, and these demands are
complied with by the ones they are directed at, or they can be “applied” by other
individuals (Kelsen 1911: 14).

(3) There is a close relation between the Ought and the will. Willing is
a factual psychical process belonging to the sphere of Is. It is relevant to the
Ought in two different ways: as the will of somebody who posits an Ought-
demand, and as the will of somebody who the Ought-demand is directed at. In
both cases, the Ought must be strictly separated from the act of will; for it is
possible that the process of willing has terminated while the Ought is still valid
(Kelsen 1911: 10–11). Kelsen, however, is not clear on the point whether there
can be an Ought without any volitional procedure founding it.21

(4) Kelsen holds that Is and Ought can have a content which is “comparable”
(Kelsen 1911: 69); this content or point of reference of both Is and Ought is
a “notion” (Vorstellung) (Kelsen 1911: 7). But he also maintains that the relation
between Is and Ought is a relation between content and form (Kelsen 1911: 10);
this would mean that the content of an Ought is an Is.

(5) Norms, the units of Ought, exist in reality by evoking a certain behavior.
A norm is “efficacious” if it is complied with, that is, if it really determines the
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

will of the individual subjected to the norm. A norm is “applied” if it is used as

a scheme of evaluation for some factual act as being, or not being, in accordance
with the norm. A secondary function of any norm is that it demands that an act
which is in accordance with it is approved, and that an act which is not in
accordance with it is disapproved, so that approval/disapproval are secondary
ways of applying a norm (Kelsen 1911: 14–17).

3.1.3 Law of Nature and Norm; Value and Duty

Like Kant, Lotze, and the neo-Kantians, Kelsen does not give a detailed defin-
ition of ‘norm’ in this early phase and, like them, he juxtaposes norm and law of
nature. A law of nature describes and explains reality, that is, factual processes
(Kelsen 1911: 6). The explanatory function is due to the causal principle which
is the basis of any natural science. Every natural law is “a judgment which
21
See Kelsen 1911: 67, on the one hand and Kelsen 1911: VI, on the other.
 28 Philosophy of Law

comprehends a process as a necessary consequence of another process in

a group of sequences of the same kind, thus explaining it” (Kelsen 1911: 5,
my translation). By contrast, norms are rules which constitute an Ought. Kelsen
lists diverse types of norms, namely, moral and legal norms, logical norms, and
grammatical and aesthetic rules (Kelsen 1911: 5).
Natural laws aim at explaining something which is given, while norms aim at
bringing about something which is not yet given (Kelsen 1911: 14). The causal
principle is not valid in the sphere of norms. Facts are relevant for norms not by
being real; rather, as content of norms, they are part of the ideal sphere (Kelsen
1911: 6). And any human act is subject to natural laws, but not every human act
is regulated by norms (Kelsen 1911: 13).
The most important normative element for Lotze and the neo-Kantians of the
Baden school was the ‘value,’ whose validity removed it far away from any
human willing or thinking. Not so for Kelsen. For him, the norm is the
fundamental Ought-element. Values in an objective sense are a function of
norms; a judgment stating a value concerns the relation between a fact or an
individual and a norm. The validity of a norm is not derived from some
transcendent value; a norm is valid because it contains an objective Ought
(Kelsen 1911: 14), that is, because it ought to be complied with or applied.
Whether the norm has any effects in the sphere of factuality or not (its efficacy)
is of no avail to its validity (Kelsen 1911: 352).
Finally, a duty is not the same as a norm, either. The objective norm contains
an abstract general ought, while the duty is a concrete individual Ought directed
at someone subjected to the norm.22
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

Any norm is a structure of norm-subject and norm-object. In the formula

‘x ought to (do) a,’ x is norm-subject and a is norm-object (which might be an
act or some state of affairs)23. Norm-subject and norm-object do not exist as
parts of the factual world but as ideal reference points of the Ought, which
relates them to each other in the norm. Accordingly, there need not be any causal
relation between them, and the norm-subject is not a physical human being, but
a ‘person’; so a person is nothing but an ideally constructed reference point of
the Ought (Kelsen 1911: 72–4).
Kelsen labels the relation between norm-subject and norm-object somewhat
contra-intuitively as ‘Zurechnung’ (imputation); a term which will become
important especially in his neo-Kantian phase. It stays opaque. Kelsen uses

22
See the extensive discussion in Kelsen 1911: 313–46.
23
Although he was one of the first legal theorists to introduce formulas for describing legal
relations (cf. Kelsen 1925: 49), Kelsen does not yet use any of them in the Hauptprobleme. He
defines the norm-subject as “dasjenige, was soll” (the one who ought to do something) and the
norm-object as “dasjenige, was gesollt wird” (that which ought to be done); Kelsen 1911: 71–2.
 Hans Kelsen’s Normativism 29

this term not only to denominate the relation between the abstract elements
inside the norm, but also to denote the relation between certain elements of
factual reality, a human being and some factual act or process, which are related
as the result of an application of the norm (Kelsen 1911: 72). The difference can
be illustrated as follows: Originally, imputation connects ‘help the needy’ with
‘everybody’ in the norm ‘Everybody ought to help the needy’; but it also
connects the factual process of helping a needy human being with the one
who is helping him. This ambiguity will accompany the notion of Zurechnung
in Kelsen’s later phases.

3.1.4 The Normative Domain of Law

Normativism in law was provisionally defined in Section 1 as the view that law
is exclusively a system of (normative) rules which are objective, impersonal,
and independent of any factuality. Accordingly, for Kelsen, the law consists of
norms; and what makes his view normativistic is that it contains nothing but
norms. It follows that every concept that is relevant for law must be derived
from a legal norm, its parts, or the set of all legal norms. It is a construction; that
is why Stanley L. Paulson’s description of this phase of Kelsen’s writings as
“critical constructivism” (Paulson 1990: 24–8) is appropriate.

3.1.4.1 The Legal Norm

The legal norm differs from the moral norm insofar as the latter is posited
“autonomously,” that is, by the norm-subject him-/herself (Kelsen 1911: 21–2,
41), while the former is posited “heteronomously,” that is, by somebody else
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

(Kelsen 1911: 36–7). Kelsen identifies the legal norm with the Rechtssatz (Kelsen
1911: 3) – a hypothetical judgment (in a logical sense) about the will of the state.
It is not “given” to legal dogmatics; rather, it has to be put together by lawyers
from the fragmented empirical material supplied by the legislative procedure; so
it is, on a different level, a “construction,” as well. The hypothetical judgment is
the “logical” or “ideal” form of the legal norm. This further distinguishes the legal
norm from the moral norm, which is an imperative (Kelsen 1911: 236–7).
Every Rechtssatz can be formulated as “Under conditions c, the state wants to
act in the way a” (Kelsen 1911: 207, 211–12). Like the norm in general, it is
a rule of imputation. But while the norm in general relates to a certain act or state
of affairs for a norm-subject as something which she ought to do or bring about,
the Rechtssatz is a law imputing an act to the state as being willed by it. The
Rechtssatz in a wider sense may impute any act to the state; in a narrow sense it
imputes coercion to the state – the state not being a given entity, but a unifying
construct.
 30 Philosophy of Law

This definition of the legal norm raises questions: On the one hand, the
formulation of the Rechtssatz seems to be a pure stipulation; on the other
hand, the Ought seems to be lost somehow – at least it does not appear in the
formula. With regard to the first point, Kelsen argues that law is necessarily the
will of the state; but a will cannot be directed at somebody else’s behavior,
which can only be wished for; so the will of the state can only be directed at ‘its
own’ behavior. Regarding the second point, Kelsen holds that the Ought is
somehow connected with the hypothetical judgment about the will of the state,
but he does not settle in a satisfactory way how this connection might be
conceived. The most relevant passage runs as follows:

If the norm drapes itself in the form of a judgment, then it raises a general
claim to truth, and it repeats itself in anybody who hears it, like any judgment.
As a judgment, the proposition: If somebody steals, then the state wills to
commit him to prison, is as true and universally – i.e. for any reasonable
being – valid as the proposition: The state is powerful, or: The horse is
a mammal. But this logical universal validity is of no consequence for the
validity of the norm posited together with the judgment. (Kelsen 1911: 259,
my translation)

This explanation is, at best, ambiguous: On the one hand, the judgment is the
“form” of the norm; on the other, there is a difference between the logical validity
of the hypothetical judgment of the Rechtssatz and the validity of the Ought,
which is somehow connected to the judgment. And it almost seems as if the norm
is “posited” together with the judgment. This does not go together well.
The source of the validity of a single legal norm is, in this early phase, tabooed
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

by Kelsen. He maintains that, for legal dogmatics, it is not permissible to ask why
a single norm is valid because this question concerns its own presuppositions. To
ask it would mean, for the jurist, “to saw off the bough he is sitting on” (Kelsen
1911: 353). To illustrate the analogous problem of ‘Is-science’ to ascertain a final
ground of the world, Kelsen cites Franz Grillparzer: “Geläng’ es mir, des Weltalls
Grund, / somit auch meinen auszusagen, / so könnt’ ich auch zur selben Stund’ /
mich selbst auf meinen Armen tragen” (Should I succeed the reason to name / of
the universe, and, thus, mine own, / I could – it would be just the same – / carry
myself on my arms alone) (Kelsen 1911: 466).
This quaintly expressed ‘taboo’ is a congenial placeholder for the basic norm,
which will be introduced only in the next phase.

3.1.4.2 The Construction of Legal Concepts

The second ingredient of Kelsen’s normativism is that every element of the law,
every legal concept, must be shown to be a construction based on legal norms.
 Hans Kelsen’s Normativism 31

The will of the state – or the state as a person, which is nothing else but this
will – is not an empirical entity; it exists just as a “point of imputation” which is
constituted by the Rechtssätze. It is something quite different from the empirical
will which might be part of the legislative procedures but which is, inside the
law, irrelevant. The same applies to the concepts of will and person in general:
Both elements only exist as reference points, that is, as constructed points of the
imputation of acts by legal norms (Kelsen 1911: 145–6). The concepts of “duty”
and “(subjective) right” can be shown to be constructions from the Rechtssatz as
well: A duty of the state is the result of a “subjectification” of the Rechtssatz
(this procedure is not further explained by Kelsen) (Kelsen 1911: 435). Besides,
there is a duty for a citizen if the Rechtssatz says that, under certain conditions,
the state wants the citizen to be sanctioned (Kelsen 1911: 447). And somebody
has a right to a certain act of somebody else if he is able to ‘trigger’ the state’s
will to sanction this somebody else in case he omits this act (Kelsen 1911:
618–29).

3.2 Neo-Kantian Transcendental Idealism

The second phase of the Pure Theory, lasting from approximately 1920 to
1935, is its heyday. Not only does it offer the most sophisticated theory of
legal cognition Kelsen ever advocated, it is also his most productive period.
Between 1920 and 1934 Kelsen published four monographs: Das Problem
der Souveränität und die Theorie des Völkerrechts (The Problem of
Sovereignty and the Theory of International Law, 1920); Der soziologische
und der juristische Staatsbegriff (The Sociological and the Legal Concept
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

of State, 1922); Allgemeine Staatslehre (General Theory of State, 1925);

and Reine Rechtslehre (Pure Theory of Law, 1934). He also wrote numerous
longer treatises which were published in journals or as brochures; the most
important ones are Die Grundlagen der Naturrechtslehre und des
Rechtspositivismus (The Foundations of the Natural Law Doctrine and
Legal Positivism, 1928), and a book-length text dealing with the theory of
Fritz Sander, Rechtswissenschaft und Recht (Legal Science and Law, 1922).
The books on sovereignty and on the concept of state, in particular, are brilliantly
written and contain more elaborate arguments than almost anything Kelsen wrote
after 1940; the polemical Rechtswissenschaft und Recht is rather chaotic and
sloppily formulated, but it gives a unique glimpse into Kelsen’s nascent neo-
Kantianism whose final version is summarized in the Foreword to the second
edition of Hauptprobleme der Staatsrechtslehre (1923). The short Reine
Rechtslehre is in many respects the closing text of this phase; in concise and
slightly Kafkaesque language, Kelsen gives a summary of his legal theory, stating
 32 Philosophy of Law

his theses in a rather apodictic way without elaborating the foundations of his
theory.
In secondary literature, it is sometimes suspected that Kelsen’s neo-
Kantianism is no more than a pretentious label. This is, however, only justified
as far as his writings from 1940 onward are concerned. For the 1920s, there can
be no doubt that Kelsen really is a true-blue neo-Kantian.
But what is neo-Kantian about Kelsen’s theory in this phase? According to
Kelsen himself, he was not aware of the affinities between his theory and neo-
Kantianism until 1912, when he read a review of his Hauptprobleme in the
Kant-Studien by Oscar Ewald (Kelsen 1998a: 15 [Kelsen 1923: XVII]). Ewald
pointed specifically to his treatment of the concept of will, which he thought was
“remarkably in accord” with what Cohen had advocated in his Ethik des Reinen
Willens. And he appreciated the book as an attempt to introduce the “transcen-
dental method” into the theory of state law. The review spurred Kelsen to focus
on neo-Kantian philosophy. This had two main impacts on his theory: First,
Kelsen conceived his legal theory as a neo-Kantian transcendental philosophy
of law, that is, as a theory of legal science as embodied by legal
dogmatics; second, Kelsen identified the legal norm with the hypothetical
judgment of legal science.

3.2.1 Transcendental Idealism

Transcendental idealism is at the core of Kant’s theoretical philosophy. Kelsen

explains it, approvingly, as follows:
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

It is impossible for cognition to play just a passive part in relation to its

objects; it cannot be restricted to reflecting things somehow existing “in
themselves”, i.e. in a transcendent sphere. As soon as we can no longer
assume objects to have a transcendent, i.e. knowledge-independent, exist-
ence, cognition has to play an active, creative part in relation to these objects.
It is cognition itself which creates its objects from the material given to it by
the senses according to its immanent laws. Cognition’s being guided by rules
guarantees the objective validity of its results. . . . What takes the place of
metaphysical speculation is the task of determining the rules guiding the
process of cognition, i.e. the objective conditions of this process. (Kelsen
1928a: 62, my translation)

This is the classical (neo-)Kantian argument: It is impossible to compare

a cognition of an object with the object itself; it is only possible to compare one
cognition of an object with another cognition of the same object. Therefore,
philosophical realism must fail as a comprehensible theory of cognition, at least
as far as it involves an adequation theory of truth or cognition. Accordingly, any
acceptable metaphysics is not concerned with the most fundamental elements as
 Hans Kelsen’s Normativism 33

they might be ‘in themselves’; rather, it deals with the necessary rules guiding
cognition: Ontology turns into epistemology (understood as a theory of cogni-
tive validity).

3.2.2 The Status of the Pure Theory as a Theory of Legal Science

But Kelsen’s neo-Kantianism is not restricted to adopting this transcendental-
idealistic conception of cognition in its classical form. As shown in
Section 2.3.2, according to Cohen the object of philosophical analysis is
‘given’ science, that is, science successfully operating as a sociocultural prac-
tice: Traditional metaphysics, which, for Kant, dissolved into an analytic of
understanding, further dissolved, for neo-Kantians who were following Cohen,
into a theory of the necessary elements and presuppositions of established
science.
Conceiving the Pure Theory in this way as a theory of legal science is at
first glance not much of an innovation. There seem to be affinities between
a ‘methodology of legal dogmatics’ and a ‘theory of legal science.’
However, the former is just a heuristic means, while a theory of science in
a neo-Kantian sense takes the place of classical metaphysics; it is prima
philosophia. What Kelsen had expounded as an unpretentious “method-
ology” in Hauptprobleme, all of a sudden becomes an ambitious equivalent
to an ‘ontology of law.’ No wonder that he was enchanted by this outlook:
“[R]eferring to the fact of science is the essential of transcendental philoso-
phy, the only basis from which it – as a theory of scientifical experience –
performs its analyses of its only possible object, the synthetic judgments of
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

experience as science” (Kelsen 1922a: 128, my translation). It follows that

the scientific judgment is “the cornerstone of transcendental philosophy,
which, therefore, can only be critique of science, critique of cognition,
because it is restricted to analysing the synthetic judgment” (Kelsen
1922a: 128, my translation).
The importance of adopting this conception, taking the fact of science and
scientifical judgments as ‘given,’ can hardly be overestimated. It is responsible
for the hypothetical approach of the Pure Theory (if the results of legal dogmat-
ics as a practice are taken to be valid, then the following elements of law and
their presuppositions have to be accepted), and at the same time offers an
additional argument for Kelsen’s relativism. Furthermore, it is expressive of
Kelsen’s view of the Ought which deviates from the Kantian tradition: For Kant,
the Ought was not a category of understanding employed in theoretical judg-
ments; rather, it emanated from (practical) reason. For Kelsen, there is no
practical reason; the Ought is a category of understanding, employed in the
 34 Philosophy of Law

synthetical judgments of legal science. Its status is analogous to that of the

category of causality. Accordingly, legal science is not a practical, but
a theoretical science.
But the Pure Theory as a theory of legal science does not only perform
a presuppositional analysis of legal dogmatics; it has a ‘critical’ function, as
well: All those elements of established legal dogmatics which are not in
accordance with its presuppositions, or with the requirements any scientific
practice has to meet, are eliminated (Kelsen 1992: 1 [Kelsen 1934: 1]). This
function of the Pure Theory results in the ‘purity thesis’ which makes up
Kelsen’s version of the ‘third feature’ of normativism. The argument would
run as follows: Established legal dogmatics expresses itself in normative sen-
tences. If we take this seriously, its judgments are Ought-judgments. As Ought-
judgments cannot be derived from Is-judgments, and to satisfy the scientific
requirement that ‘methodological syncretism’ is to be avoided, legal science
must consist exclusively of Ought-judgments.24

3.2.3 The Hypothetical Judgment as the Fundamental Cognitive Unit

While the legal norm, in Hauptprobleme, was necessarily connected with the
Rechtssatz, these elements were nevertheless not identical: The validity or
existence of the legal norm was different from the validity of the hypothet-
ical judgment constructed from the legislative material given to legal dog-
matics. Or so Kelsen maintained. Now, in his neo-Kantian phase, Kelsen
explicitly identifies the legal norm with the hypothetical judgment of legal
science; the validity of a legal norm is its validity as a cognitive judgment.
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

A pithy formulation of this thesis can be found in Rechtswissenschaft und

Recht:

[Sander fails to see] that legal science may “create” its object – law as
embodied by legal rules [Rechtssätze], by judgments – just as natural
science may “create” its object, the nature of natural science, nature as
a system of synthetic judgments without in the least running any danger of
becoming a “source” of law. . . . Though “created” by science itself, the
synthetic judgments of natural science are determined by the material
which they unify (therefore, they are judgments “about” [über] nature)
just as the synthetic judgments of legal science, the legal rules, in which
the material given to legal science . . . is formed, formed into legal rules in
the same way as the perceptual material is formed in the synthetic judg-
ments of natural science. . . . If we are inclined to qualify the acts which are

24
The fight against methodological syncretism, or against the ‘metabasis eis allo genos,’ is
a central theme for neo-Kantian anti-psychologists and for Kelsen; see Kelsen 1920: V;
Kelsen 1922b: 34, 45.
 Hans Kelsen’s Normativism 35

the material of the legal rules as acts of will [Willensakte], then the legal
rules as judgments “about” them have to be expressed by an Ought. In
cognition, i.e. in the judgments of legal science appearing as legal rules, the
willing [Wollen], which is objectified in the sphere of judgment, expresses
itself as an Ought. Only in this sphere can we talk of law as an object of
legal science. As the object of legal science, law is a system of judgments
about law, just as nature as the object of natural science is a system of
judgments about nature. (Kelsen 1922a: 181–2, my translation)

Accordingly, “[both the state and the law are] normative order[s], i.e. system[s]
of norms which are linguistically expressed by Ought-sentences and logically
in hypothetical judgments in which the condition is connected to the conse-
quence by the ‘Ought’ (if a, then b ought to be)” (Kelsen 1922b: 75, my
translation).
The above may be summarized as follows: Cognition creates its objects by
synthesizing judgments. These judgments are objectively valid without corres-
ponding to something ‘given’ to them by being in accordance with the rules that
guide cognition. ‘Nature’ as the object of knowledge is not a conglomeration of
knowledge-independent ‘things,’ but a system of synthetic judgments. In these
judgments the category of causality is applied to perceptual material. In an
analogous way, the law as the object of legal science is identical with the system
of judgments created by legal science. It is not the task of legal science to
‘passively’ describe norms given by the legislator; rather it applies the Ought to
a nonnormative substratum consisting of the legislator’s acts of will (or their
content), thus constituting hypothetical normative judgments which are identi-
cal with the legal norms.25
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

3.2.4 Is and Ought

It was shown in Section 2 that Hume thought the dualism of Is and Ought to be
one between objective analytical or empirical sentences, and subjective prac-
tical sentences. Kant, by contrast, took practical sentences to be objective, and
introduced normativity into epistemology, so that there was a bifurcation of the
dualism of Is and Ought. The neo-Kantians concentrated on the epistemological
part of the dualism, equating the theoretical Ought with a cognitive judgment’s
validity.
It is not easy to locate Kelsen’s version of the dualism in this context. He still
maintains that the reason or ground (Grund) for the validity of a norm can only
be another norm (Kelsen 1920: 95). Besides, there is another ‘formal–logical’
explanation of the dualism which is more precise than what Kelsen brought

25
This summary is mainly taken from Heidemann 1999: 348.
 36 Philosophy of Law

forward in Hauptprobleme: It is impossible that an Ought-judgment contradicts

an Is-judgment (Kelsen 1928a: 62).26
The most prominent version of the Is–Ought dichotomy is Kelsen’s juxta-
position of causality and imputation. However, he also equates the dualism with
the seemingly ontological divide between nature and mind, nature and society,
reality and ideology. And in the manner of the neo-Kantians, he instrumenta-
lizes the Ought to thwart any psychologistic tendencies in legal science; fur-
thermore, he mentions a theoretical Ought which is a purely formal concept,
standing for the autonomy of law or, what amounts to the same, the autonomy of
legal science.

3.2.4.1 Causality and Imputation

The main case of the dichotomy between Is and Ought, for legal science, is the
dualism between normative imputation and causality. This is expounded by
Kelsen in the foreword to the second edition of the Hauptprobleme:

[T]he norm as a judgment of Ought is contrasted with the law of nature, and
the Rechtssatz, which is qualified as a norm, is contrasted with the specific
causal law of sociology. Therefore, my main concern is the Rechtssatz seen as
an expression of the specific legal autonomy [Rechtsgesetzlichkeit]; and from
this perspective the law of legal science is a system of Rechtssätze, a system
of judgments, just as nature as the object of natural science is – for transcen-
dental philosophy – a system of judgments. The difference lies simply in that
the latter judgments express the causal relation of the system of nature
whereas in the Rechtssatz the specific normativity of law manifests itself.
Both in the judgment, which is the causal law, and in the Rechtssatz, a certain
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

condition is connected with a certain consequence. But instead of the causal

nexus, it is a different principle of connection which binds the consequence to
the condition [in the Rechtssatz]. (Kelsen 1923: VI–VII, my translation)

This connection to causality, the analogon in the sphere of law, is the specific
legal Ought. Kelsen dubs it “periphere Zurechnung” (peripheral imputation)
(Kelsen 1992: 50 [Kelsen 1934: 57];27 it must be distinguished from “zentrale
Zurechnung” (central imputation) which consists in linking an act to a person.
The background to this version of the dualism is Kelsen’s specific brand of

26
It is more precise because the terms ‘ableiten’/derive or ‘(be-)gründen’/ground (which Kelsen
uses) are not well defined. A sophisticated formulation of the dualism might run as follows: “It is
impossible to deduce a nontrivial pure normative sentence from a consistent set of sentences
which does not contain any pure or mixed normative sentences.” But that does not seem to say
more, for this impossibility seems to be given exactly if Is-judgments and Ought-judgments
cannot contradict each other.
27
As far as I can see, Kelsen never explained why he called this central normative function
“peripheral imputation” and denoted the rather unimportant ascription of an act to a person (a
person being just a reification of a set of norms) as “central imputation.”
 Hans Kelsen’s Normativism 37

neo-Kantianism according to which the object of cognition must be an

Urteilszusammenhang, a coherent set of valid cognitive judgments. These
judgments must be hypothetical judgments, that is, judgments of the form “if
a, then b,” because the objectivity of cognition demands that it is expressed by,
or can be derived from, universal propositions (what Kelsen calls
“Gesetzlichkeit”; Kelsen 1923: VII). The category of peripheral imputation
constitutes the Rechtssatz, the general legal norm qua cognitive hypothetical
judgment, by joining together two states of affairs and determining that if one is
given, the other (an exercise of coercion) ought to follow. As such, it does not
have much in common with the ‘classical’ practical Ought; it is just a purely
formal linkage. There is no norm-subject, and there is no a priori content of any
legal norm, that is, some legal content which is valid without some explicit
legislative or judicial act whose ‘meaning’ it is. In Allgemeine Staatslehre
Kelsen gives the following ‘average’ formula for the general norm: “If Ha +
C (or Hna + C), then Co>H,” that is, if a human being acts in way a (and certain
conditions are fulfilled), or if a human being fails to act in way a (and certain
conditions are fulfilled), then coercion ought to be exerted (normally against
this human being) (Kelsen 1925: 49).
Perhaps this relation, which is closer than material implication and not quite
the same as entailment, can best be explained by the notion of a deontically
perfect world: In a deontically perfect world, that is, in a world where no norm is
violated, coercion follows upon fulfillment of the conditions with the necessity
of a law of nature.
A notable by-product of this conception is that legal norms, being logical
judgments, are as a matter of course capable of standing in logical relations –
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

an issue that will become a major problem in Kelsen’s third and fourth
phases.28
Kelsen, like Kant, takes Ought-judgments to be objectively valid, that is,
genuine cognition. But, as noted in Section 2.2, Kant has it that practical
cognition, cognition of Ought, is a special function of reason, while empirical
cognition is the result of an interplay of understanding with intuition, of
concepts with nonobjective perceptual material delivered by the senses. By
contrast, Kelsen’s conception of legal cognition is an exact analogy to empirical
cognition. The Ought, for him, is not a concept of reason, but a category of
relation, constituting a judgment which processes empirical material, guided by
the rules determined in a “higher” norm:

I have maintained that questioning the specific validity of an individual legal

rule leads – step by step – to a next higher legal rule and finally to the basic

28
See Section 3.3.5 for the third phase and Section 3.4.6 for the fourth phase.
 38 Philosophy of Law

norm, while questioning the content of the legal rule, i.e. asking why a single
legal rule of a specific legal system has just this and no other content, leads to
the legislative acts or judicial judgments which constitute the “material” of
the Rechtssätze. This distinction is analogous to the distinction between
concept and perception, between the logical form and its perceptual material,
made in transcendental philosophy. I have distinguished between the “con-
tent”, meaning the material still to be formed, and “validity”, meaning the
form of the material when construed into a valid logical judgment. . . . The
logical creation of the law . . . from the basic norm proceeds step by step and
under constant reference to a parallel fact. (Kelsen 1922a: 214–5, my
translation)

This is one of Kelsen’s first depictions of the legal hierarchy, the Stufenbau of
law. It is noteworthy that the higher legal norms do not employ the category of
imputation; rather, they are meta-rules stating empirical criteria for lower norms
to be valid. I will come back to this point later (Section 3.2.5). For now, it
suffices to say that peripheral imputation, for Kelsen, is identical with the legal
Ought; it is a purely formal category of relation employed in the hypothetical
judgments of legal science whose content and validity depend on factual acts
which are named as criteria for the norm’s validity in “higher” rules.

3.2.4.2 Nature and Mind, Reality and Ideology, Nature and Society
Kelsen’s most prominent and coherent interpretation of the dualism of Is and
Ought in his neo-Kantian phase has it that these elements are identical with
causality, in the sphere of nature, and (peripheral) imputation, in the sphere of
law. However, Kelsen brings forward several other explanations of the dualism.
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

One strand of explanation is seemingly an ontological one. Kelsen declares the

dualism of Is and Ought to be identical with, constitutive of, or derivative from, the
dualisms of nature and mind, reality and ideology, nature and society. It is not
necessary to go into detail regarding these ontological versions of the dualism,
because they do not concern the legal Ought as part of the transcendental-idealistic
approach. Kelsen himself summarizes them in one grand sweep as follows:

The conditions, stemming from the sphere of causally ordered events, for the
genesis of certain representations (understood as psychic acts) of norms
might be called the “real substructure” on which the norms and systems of
norms, as specific ideal contents, rise like a superstructure, as “ideologies”
which are autonomous vis-à-vis the natural laws governing the substructure.
But this is just an image, borrowed from the terminology of the materialist
conception of history, for the relation between the systems of nature and
mind, and, hence, between the systems of “reality” and of value, of which the
relation between nature and society is a special case. (Kelsen 1925: 21, my
translation)
 Hans Kelsen’s Normativism 39

There is one important manifestation of this dualism which will later be

discussed in a bit more depth (Section 3.3.5): the dualism between a psychic
intentional act and its ideal meaning-content, as introduced by Husserl when
fighting against psychologism. The psychic act is part of nature. The Ought is an
ideal or abstract content, so it belongs to the level of the meaning of the act. On
the one hand, Kelsen emphasizes that both levels are autonomous; on the other,
he maintains that it is peculiar for meaning-contents always to be sustained by
psychic acts (Kelsen 1920: 97 fn.; Kelsen 1922b: 81 fn.).

3.2.4.3 Ought and Validity

In the second phase, the notion of validity is, superficially, defined as it was in
the first phase; it is the “specific existence” (Kelsen 1992: 12 [Kelsen 1934: 7])
of the norm which consists in the fact that one should ‘really’ behave in the way
determined by the norm. This definition of validity stays constant throughout all
phases of the Pure Theory.29 Kelsen rarely explains it in detail; but in the neo-
Kantian context, validity is not special to normative judgments; rather, it is the
‘logical’ existence of any cognitive judgment which is given if the sentence
expressing the judgment is true. It is normatively charged itself. Among the
most challenging parts of all Kelsen’s writings are those passages in
Staatsbegriff and Rechtswissenschaft und Recht where he deals with Husserl,
Windelband, the concept of validity, normative logic, and psychologism. They
mark two of the few occasions where he delved somewhat deeper into neo-
Kantian philosophy, and they are remarkable for the philosophical flights he was
sometimes ready to take. In Staatsbegriff Kelsen writes:
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

Even in my Hauptprobleme, in order to explain the non-psychological mean-

ing of the Ought, which, if attributed to any content, is completely different
from the assertion that anybody wants or wishes this content, I referred to the
depiction which Husserl gave of this problem in his Logische
Untersuchungen, 2nd ed., 40 ff. In this connection, I want to call attention

29
In secondary literature, Kelsen’s concept of validity is widely held to be obscure, covering
different things, like the existence of the norm, its binding force, its being a member of the legal
system, etc. (see, e.g. Nino 1998 and the papers by E. Bulygin and E. G. Valdés in the same
volume). This is not justified; the core of Kelsen’s conception of validity is clear throughout all
the phases: The norm, that one ought to do a, is valid (i.e. in the neo-Kantian context, is a valid
ought-judgment) exactly if one ‘really’ ought to do a, or, which means the same, if the sentence
“one ought to do a” is true. In a similar way, the question whether the legal Ought is
a “legitimate” or “justified” Ought is idle (on this question, see, e.g. Paulson 2012). The legal
norm is “binding,” “legitimate,” or “justified,” exactly if it is valid, that is – again – if one ‘really’
ought to do what the norm says. But this validity is relative to the frame of reference of the
established practice of legal dogmatics; as a matter of course, a legal norm can only decree what
ought to be done legally. Inside the autonomous ‘language game’ of law, one ‘really’ ought to do
what the law says. But there might still be reasons for not letting this language game determine
how one acts.
 40 Philosophy of Law

to the far-reaching parallel between the dualism of the ideal normative-juridic

view and the real causal-psychological view and the dualism, depicted
brilliantly by Husserl, between the ideal-normative-logical and the real-
psychological view (cf. especially Husserl, ibid., 50 ff.). It cannot be investi-
gated in this place how far the parallel reaches. But it is remarkable that
normative logic, which stands in contrast to psychological logic, tends to
become pure logic, just as the normative theory of law and state, which stands
in contrast to sociological theory, tends to become a pure theory of law and
state. I cannot determine whether the “pure” logic can no longer be
a “normative” logic, as Husserl maintains. But it strikes me that Husserl’s
terminology is not quite clear in this regard. Although he replaces the dualism
of Is and Ought, natural law and norm, by the dualism between natural law
and ideal law, real science and ideal science, although he abominates the
identification of ideality and normality and wants to conceive the “pure” logic
as a system of theoretical sentences, not as a system of norms, he himself talks
about the fundamental difference between the purely logical norms and the
“technical rules of a specific human craft” of thinking. . . . And it seems not to
be a coincidence that especially those theorists who oppose the psychologist
logicians by setting up logic as a normative discipline (Kant, Herbart,
Drobisch, Sigwart) postulate a “pure” logic at the same time. For them, the
“Ought” achieves “purity” vis-à-vis psychology, just as it does for the theory
of law and state vis-à-vis psychological sociology as a natural science.
(Kelsen 1922b: 81 fn., my translation)

And in Rechtswissenschaft und Recht he maintains:

The pivotal point for the non-psychological meaning of Ought which is

independent of any human subjectivity is what Kant claimed emphatically
and in all clearness: The validity of the judgment that something ought is
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

independent from the validity of the judgment that this something is, was, will
be, or even can be. This independence of Ought-validity from Is-validity, i.e.,
the validity of the being of nature, must by necessity exclude any reference to
the actual wanting or wishing, recognising or acting of the content of the
Ought on the part of the human being at whom the Ought is “directed”. The
Ought . . . in the judgments of legal science, i.e., the legal rule, is no more
directed at a human being than the law of nature is. . . . Striving for a non-
psychological meaning of the legal norm as an Ought-judgment, I am in the
same situation as the logicians vis-à-vis the logical norm. In his Prinzipien
der Logik, Windelband says about the pure logic: “It does not teach how one
really thinks, but how one ought to think if one wants to think correctly. This
common definition suffices to express the principal demarcation of logic
against psychology: But it must not be overlooked that the formulation
makes allowance for the fundamental fact of empirical thinking, mentioned
above, that it is prone to the possibility of error, . . . so that one has to take
care that the validity of these forms is in the last instance completely
independent from the striving of empirical and especially human conscious-
ness for knowledge. . . . From this it follows that all laws of logic have
 Hans Kelsen’s Normativism 41

a double status: On the one hand, they are rules for empirical consciousness,
which all thinking ought to comply with, if striving for truth. On the other
hand, their inner meaning and being is completely independent from the
question whether there are any mental processes which are in accordance
with them or not. The latter might be called their validity per se, the former,
their validity for us. . . . From ‘our’ perspective, logic is an ‘Ought’. But this
Ought must ground in something whose validity exists per se and which
becomes a norm, an Ought, only by being related to a consciousness capable
of error.” The double meaning of the logical norm, its validity per se, that is
its “objective” validity, . . . and its validity for us corresponds to the objective
validity of the legal norm claimed by me, and its subjective validity which
might better be called “efficacy”. . . . The Ought is the suitable expression
precisely for this objective validity of the logical law, the validity per se;
accordingly, Windelband himself emphasizes that this Ought expresses the
demarcation of logic against psychology. (Kelsen 1922a: 206–8, my transla-
tion, italics added)

These passages build a bridge to the neo-Kantian discussion concerning

validity and normativity described above, and they are a key to understanding
Kelsen’s concepts of Ought and validity: Kelsen sides with Windelband in
saying that validity is what distinguishes objective meaning-contents from
natural world objects; he also holds that validity is a normative concept. He
chooses to differ, however, as far as the question whether ‘Ought’ is an adequate
expression for this normative feature of validity is concerned. In their attempt to
fend off any psychologistic connotations, Husserl and Windelband answer in
the negative, because the Ought refers to human subjectivity. Husserl grounds
normativity in a sphere of ideality, while Windelband anchors it in the sphere of
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

values. Both run the danger of falling prey to Platonism; this problem was
discussed in the first part.30 Not so Kelsen: For him, the Ought is an adequate
expression of this kind of objective normativity; he thinks that the psychological
connotations of the Ought are taken care of by distinguishing between norma-
tive validity and factual efficacy (in the special sense of the meaning-content
being in sufficient degree embodied in psychic acts). He adds that there is no
individual who the legal norm is directed at, anyway; after all, the hypothetical
normative judgment of legal science has no addressee.
I tend to think that Kelsen is right. There is, however, a problem. Equating the
Ought with normative validity results in a peculiar duplication of the Ought. On
the one hand, the legal Ought is identical with the cognitive category of
peripheral imputation, which is at work inside the hypothetical judgment of
the Rechtssatz. On the other hand, the Ought is identical with the validity of the
Rechtssatz as a judgment, which is a matter of its logical existence. This kind of

30
See Section 2.3.3
 42 Philosophy of Law

Ought is not specific to normative judgments; it also adheres to objective Is-

judgments. It might best be rendered as

One ought to judge/think that way.

If this implicit normative element is made explicit, a valid Rechtssatz might

be formulated, using two different kinds of Ought, as

One ought1 to judge/think that, under conditions c, coercion ought2 to be

exerted.

Kelsen does not clearly distinguish between these two types of Ought; and he
sometimes, as in the passages cited above, seems to hold that the validity-Ought
is quite enough to guarantee the autonomy of legal science.31 This is distantly
echoed in some less characteristic passages where Kelsen maintains that the
Ought is just a formal marker, the ladder to get into the legal system; once you
are inside, it makes no sense to speak of Ought and of the dualism of Is and
Ought.32 That, in turn, is vaguely reminiscent of the neo-Kantian notion that law
as a social and cognitive practice is ‘given,’ that is, ‘always there,’ and that the
legal Ought is problematic only ‘from the outside,’ or from the level of meta-
theory.

3.2.4.4 The Objectivity of Ought

Kelsen holds that the Ought is a category constituting theoretical judgments of
legal science; it is in a strong sense objective: If there is a norm according to
which you ought to act in a certain way, then it is not just the case that someone
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

thinks that you ought to act in a certain way; rather, you really ought to act in
a certain way. Thus, validity and objectivity of the norm coincide.
This seems, at first sight, to be at odds with Kelsen’s assumption that there
can be no objective morality, that is, no objective ‘absolute’ Ought. For he is
quite adamant on this point. In Reine Rechtslehre, he holds that, from the
perspective of rational cognition, justice as part of social morality is just
a compromise between conflicting interests. This compromise is a matter of
will, not of reason (Kelsen 1992: 16–18 [Kelsen 1934: 15–16]). There is no
practical reason, and there can be no absolute morality.

31
When dealing with Max Weber in Staatsbegriff, however, Kelsen seems to argue (not quite
convincingly) the other way round: Validity is not Ought1, rather, Ought2 is validity, so that the
normative meaning of an ideology seems to guarantee its validity; see Kelsen 1922b: 162.
32
Cf. Kelsen 1920: 10 fn. (It is a peculiarity of Problem der Souveränität that the footnotes are
often more significant than the main text. Kelsen had the text more or less finished in 1916, but
publication was delayed until 1920, and some of the notes seem to have been added in the
meantime. So the book mirrors Kelsen’s ‘thought in progress.’)
 Hans Kelsen’s Normativism 43

However, these views might go together, even if Kelsen’s relativistic view of

morality is a somewhat superfluous add-on to his theory. After all, although he
claims to offer a transcendental philosophy of law, this transcendental philoso-
phy is just a critical reconstruction of legal science, which he equates with legal
dogmatics, from the inner perspective. If transcendental philosophy in a neo-
Kantian vein adheres to some factum of science, working out its essentials and
presuppositions, then its perspective is at the same time internal (taking the
scientific judgments to be valid) and detached (not determining whether they
really are valid). It does not say that the sentences of sciences are true because
their presuppositions are true (or vice versa); rather, it maintains that the
presuppositions are a necessary condition for the judgments of legal dogmatics
to be valid, while the validity of these judgments is a sufficient condition for the
presuppositions to hold. There is a kind of relativity ingrained in taking the
factum of legal science as a starting point. The neo-Kantian philosopher is
a meta-theorist. He does not maintain:

As the judgments of legal science are valid, their presuppositions hold,

or vice versa; rather he tells us:

The judgments of legal science are the best we can get, if we want to find out
what the law is, and if the judgments of legal science are valid, certain
presuppositions hold.

3.2.5 Relations between Norms: the Legal Hierarchy and the Basic Norm
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

Besides founding his theory explicitly on a neo-Kantian basis, Kelsen intro-

duces further innovations during this second phase which are related to each
other: He presents the Stufenbautheorie, i. e. the theory of a hierarchy of legal
norms, and the concept of a ‘basic’, or rather ‘apex,’33 norm.
Both conceptions are part of a solution to the ‘taboo’ concerning the
question of the validity of legal norms. To recap: In the first phase, Kelsen’s
opinion was that the law consists of just one level of general norms, the
Rechtssätze, expressing a conditioned will of the state to act in a certain
way. It was regarded as a presupposition of legal knowledge that these general
norms are valid. Now, Kelsen distinguishes several layers of norms. A judicial
sentence or an executive order is a legal norm, as well – namely, an individual
norm which is valid if it was issued in accordance with a general norm. The

33
Kelsen calls it “Grundnorm” but actually, in the hierarchy of meta-levels, it is the highest norm of
the legal system – if it belongs to the legal system at all.
 44 Philosophy of Law

general norms, in turn, are valid if they are issued in accordance with the
constitution. The validity of a given constitution may be derived from its being
issued in accordance with an earlier constitution; but once the justifying
process arrives at the historically first constitution, it will be impossible to
find a higher ‘positive’ norm which might invest it with validity. In this case,
Kelsen says, any jurist presupposes a norm which is the ground for the validity
of the highest positive legal norm. This nonpositive presupposed norm is
called “Grundnorm,” that is, basic norm.
Neither conception originated with Kelsen alone. Even though there were
indications of it in Kelsen’s earlier work, the conception of a basic norm was
mainly developed by Alfred Verdross and Leonid Pitamic, while the
Stufenbautheorie was developed by Adolf J. Merkl – as Kelsen himself
acknowledges (Kelsen 1998a: 13 [Kelsen 1923: XV]). Ironically, both theorems
are, on the one hand, often regarded as major achievements of Kelsen’s theory,
and on the other, they do not work flawlessly – they are not part of the Pure
Theory ‘at its best.’ They are not, however, necessary ingredients of Kelsen’s
normativism; so I will confine myself to a few remarks.
Both conceptions are expressive of Begründungsdenken, or the principle of
sufficient reason. Even in the first phase, Kelsen maintained that there are in
principle endless chains of deriving an Is, via causal laws, from another Is, and
of deriving an Ought from another Ought. But he did not elaborate this, and he
did not explain how an Ought might be derived from another Ought. This
explanation seems, for the legal sphere, to be achieved by the theory of the
Stufenbau. But is it?
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

One problem in adequately understanding the Stufenbautheorie is that Merkl,

its creator, tended to explain it metaphorically, by using images like that of the
“cataracts” of a river, where the water is more and more refined and distributed,
just as the legal Ought is when it reaches the lower levels of norms (Merkl 1917:
col. 175). Kelsen is more matter-of-fact, but nevertheless the relation between
a norm of a higher level and a norm of a lower level is not quite clear from his
writings. There are two different versions which are not necessarily exclusive,
but difficult to bring together.
According to the first, a higher norm (necessarily) regulates the “creation”
and (possibly) the content of the lower norm (Kelsen 1992: 63 [Kelsen 1934:
74–5]). Now, what does this mean? In this phase Kelsen takes the legal norm to
be a logical Ought-judgment; and you obviously cannot create a logical judg-
ment in the same way you might, for example, create a painting or a cake. Still,
“creating” a valid Ought-judgment can be achieved by fulfilling the conditions
for the judgment to be valid, or, what amounts to the same, for the correspond-
ing normative sentence to be true. After all, in a Fregean or neo-Kantian spirit,
 Hans Kelsen’s Normativism 45

a true judgment in a logical sense is the same as a fact,34 and there is nothing
against saying that one can bring about (normative) facts. In this case, the higher
norm is a meta-level rule. Simplified, it has a form somewhat like:

If there is a performative/utterance under certain conditions and with a certain

content, the meaning of the performative/utterance gives rise to a valid norm.

In other words, it contributes to the conditions for imputing an act of coercion

to some state of affairs. The circumstances of the performative and the limita-
tion to its content would be criteria for the (lower) norm to be valid. There
would be a legal hierarchy as a logical Stufenbau of object- and meta-levels.
There are, however, two problems. First, the higher norm as a validity-rule
would not be comparable to a norm in the traditional sense (to be sure, neither is
the Rechtssatz). Second, as a validity-rule, it does not fit into the form of the
hypothetical judgment connecting a coercive act with certain conditions in
a normative way. The definition of the legal norm as a hypothetical judgment
is tailored to the one-level conception of law presented in the Hauptprobleme; it
seems oddly misplaced when confronted with the Stufenbautheorie.
According to the second version, the higher norm serves as a scheme for
transferring the Ought onto some empirical or ‘alogical’ material, thus bringing
a new norm into existence. This version is prominent in the specific neo-Kantian
writings. As a system of cognitive normative judgments, law is created by legal
science, using ‘higher’ norms to bestow the mode of Ought or validity on some
content ‘given’ by the empirical acts of legislature or judiciary:

I have distinguished between content, meaning the material still to be formed,

https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

and validity, meaning the form of the material when construed into a valid
logical judgment. The logical creation of the law (meaning, of course, the law
of legal science, the Rechtssätze) from the basic norm proceeds step by step
and under constant reference to a parallel fact. (Kelsen 1922a: 214–15, my
translation)

And Kelsen even employs the specific Kantian terminology of the norm as
a “Schema” for interpreting empirical material given to legal science (although
he seems to have borrowed the term from Max Weber) (Kelsen 1992: 10
[Kelsen 1934: 5]).
A problem with this version is that Kelsen sometimes comes close to saying
that the empirical law-giving act, as a performative, is somehow transformed
into a norm by applying the higher norm as an interpretive scheme (Kelsen
1925: 249). This is not, of course, the case; the performative is just ‘triggering’

34
Cf. Frege 1986: 33: “Was ist eine Tatsache? Eine Tatsache ist ein Gedanke, der wahr ist” (What is
a fact? A fact is a thought which is true.)
 46 Philosophy of Law

the norm. The fact which is interpreted as giving rise to law is part of evaluated
reality; it is a ‘legal act,’ but not a norm.
But no matter which version we prefer, is the Stufenbau of law really a matter
of deriving one norm from another, a question of a ‘chain’ of validity or Ought?
In the case of Is, or causality, it is indeed possible to construct a chain by stating
that a causes b, b causes c, c causes d, and so on. But there seems to be no
equivalent for such a chain in the legal sphere.35 The lower norm is not
determined or ‘grounded’ by the higher norm alone. Without the performative
of the law-giving act, there would be no ‘derived’ norm. Accordingly, the Is-
sphere and the Ought-sphere seem to mingle. Kelsen tries to get out of this
predicament by saying that all empirical factors can just be the conditio sine qua
non for the norm’s validity, while the higher norm is the conditio per quam
(Kelsen 1920: VI); but this is not really convincing. If the higher norm were
a sufficient condition for the lower norm’s validity (that is, existence), then it
would be possible to deduce the lower norm from the higher norm. It is more
plausible when Kelsen maintains, as in the passage cited above, that the lower
norm gets its content from empirical acts, while it gets its validity from a higher
norm. But this is metaphorical and vague, and in the end it comes down to the
meta-level conception of the legal hierarchy: The higher norm is a meta-rule
according to which, under certain conditions, the content of certain utterances or
performatives is, or contributes to, a valid norm. It is a ‘bridging principle,’
sewing together Is and Ought.
Another major innovation in this phase is the introduction of the basic norm.
There is abundant literature concerning this specific concept,36 but as it is not
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

really important in the present context it is not necessary to dwell on it

extensively. The basic norm appears in two different characters. On the one
hand, in the legal hierarchy, it is the highest norm. It answers the question why
the constitution (or, rather, the historically first constitution from which its
validity can be derived) is valid. One answer to this question would be to say
that its validity is simply presupposed by legal science. For legal science as an
existing sociocultural practice, the law is ‘always there’; and as no jurist
questions the validity of the constitution, its validity is presupposed. It simply
is the fundamental constitutive rule for law. This, however, is not the way
Kelsen proceeds. He does not hold that the validity of the constitution is
presupposed (though sometimes he comes close to doing so) (cf. Kelsen
1928b: 24); rather he says that legal cognition, once justifying the validity of

35
In fact, Kelsen holds the endless causal chain to be the analogue of the chain of central
imputation (imputing an act to a person, see Section 3.2.7); central imputation, however, always
ends somewhere (Kelsen 1968a: 134–5).
36
For two prominent examples, see Paulson 2000 and Raz 1974.
 Hans Kelsen’s Normativism 47

a norm comes to an end for the lack of another higher positive criterion for valid
norms, presupposes a norm according to which the content of a performative –
be it custom or the intentional act of someone or a group exerting factual
power – counts as the highest legal norm, that is, as constitution (Kelsen
1992: 56–7 [Kelsen 1934: 65–6]). But why should this be necessary?
Famously, H. L. A. Hart held the basic norm to be a needless reduplication
(Hart 1961: 246). If the neo-Kantian task of the Pure Theory is to reconstruct the
intellectual content of a given sociocultural practice claiming normativity and
objectivity, why not just say that inside this practice the constitution (or, what
amounts to the same thing, its validity) is regarded as an axiom, that is, as an
assumed nonverifiable judgment naming the highest criteria for validity inside
the system?
For Kelsen, there are several possible reasons why this might not be suffi-
cient. First, law, according to Kelsen, is only positive law; that is, any legal norm
must get its content from a factual act which is in accordance with a ‘higher’
norm. So, for the constitution to be positive law, it must ‘rest’ on a performative,
which fulfills the conditions for being norm-triggering, named in a yet higher
norm; this higher norm cannot be a positive norm, it can only be presupposed as
a basic norm. Second, the practice legal cognition is embedded in only takes the
constitution to be valid if it adheres to a legal system which is effective and
regularly applied in a way that takes the constitution to contain the highest
criteria of legal validity. So the basic norm is necessary to capture the legal
system’s being conditioned by its efficacy – sometimes Kelsen even seems to
identify it with the principle of efficacy.37 Third, the ‘positive’ constitution often
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

consists of disparate provisions, posited both by those who had the power to do
so and by custom. To unify these provisions and, thus, the law, it is necessary to
presuppose a norm under whose antecedent they can be subsumed as highest
criteria for legal validity. This, in turn, is necessary so that law is a system and
possible object of science.
On the other hand, and this is mainly responsible for the vast literature concern-
ing the basic norm, Kelsen holds that it is the (or a) “transcendental-logical
condition” (transzendental-logische Bedingung) for legal cognition, and thus for
law; or a “hypothesis” in the sense of Cohen and Plato, that is, a fundamental
constitutive element (Grundlegung) posited by cognition (Kelsen 1998b: 174).
This is argued for by Kelsen in the following well-known passage:

In formulating the basic norm, the Pure Theory of Law . . . aims simply to
raise to the level of consciousness what all jurists are doing (for the most part
unwittingly) when, in conceptualizing their object of enquiry, they reject

37
See Section 3.2.6.
 48 Philosophy of Law

natural law as the basis of the validity of positive law, but nevertheless
understand the positive law as a valid system, that is, as norm, and not merely
as factual contingencies of motivation. With the doctrine of the basic norm,
the Pure Theory analyses the actual process of the long-standing method of
cognizing positive law, in an attempt simply to reveal the transcendental
logical conditions of that method. (Kelsen 1992: 58 [Kelsen 1934: 67–8])

Is this rather high-flown claim justified? In part, it is, and in part, it is not. For
Kantian philosophy, transcendental-logical conditions are those conditions of cog-
nition which make reference to objects possible. They are the categories of
understanding which might be found, according to Kant, by taking the table of
judgments as a guideline; so, for Kelsen, who levels out the difference between
cognitive judgment and cognitional object and only takes into account hypothetical
judgments, they are identical with the ‘connector’ inside these judgments.
Peripheral imputation, the specific legal Ought constituting the hypothetical judg-
ment of the Rechtssatz, is such a category of relation. It might truly be said to be
a transcendental-logical condition. The basic norm, on the other hand, is not
identical with the category of Ought; rather, it is a presupposed, most fundamental
rule offering criteria for legal validity. And if one takes a closer look, it is not really
an applicable criteria-rule, but just a necessary go-along if one takes the constitution
to be valid. There is no real equivalent for this function in Kantian philosophy, and
to call the basic norm a transcendental-logical condition is at least oblique. Yet, as
shown, neo-Kantianism tends to downgrade transcendental philosophy to a critical
meta-theory of scientifical practice, that is, a critical analysis of the presuppositions
of the established sciences – those implicit nonfactual conditions which are neces-
sary to make it possible. Nothing speaks against taking the basic norm to be
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

a transcendental-logical condition in this very weak sense of not being an object,

but a general presupposition of legal science as it is practiced. These matters,
however, need not be further pursued; they are not a special problem of the Is–
Ought dualism or of normativism.

3.2.6 Relations between Is and Ought

Even though it is not possible, according to Kelsen, to derive an Ought from an
Is, or vice versa, there are numerous relations between Is-elements and Ought-
elements.

(1) Norms, as objective cognitive judgments, are abstract meaning-contents. In

order to ‘exist’ in reality, abstract meaning-contents need a psychical act, like
willing, thinking, or feeling, as their bearer. They are not ‘given’ without being
willed, thought of, or felt, although they cannot be reduced to these psychical
processes. Their validity is independent from any Is:
 Hans Kelsen’s Normativism 49

The psychological determination is irrelevant for the logical validity of

a notion, the truth of its content: The process of thinking which has
Pythagoras’ theorem as its content is irrelevant for the correctness of this
theorem. (Kelsen 1920: 16, my translation)

Is-elements are the conditio sine qua non of Ought-elements, but not their
conditio per quam (Kelsen 1920: 97 fn.; Kelsen 1925: 14). The empirical acts
which ‘bear’ the abstract meaning-contents are “parallel phenomena” in a different
dimension which necessarily conditions them (Kelsen 1920: 16). Talk of necessary
and sufficient conditions, in this context, is again a bit awkward; the problem of
“parallel phenomena” in different dimensions can be more aptly captured by the
notion of ‘supervenience’: An element a supervenes upon an element b, if there
cannot be a difference in a without a difference in b.38 Applied to law, and put
somewhat sloppily: There cannot be a change in the legal sphere without a change
in empirical reality.39 More importantly, Kelsen would probably have done better to
distinguish, in this context, not only two, but three different phenomena: the
abstract content as such, the empirical (neurological or psychical) act as such,
and the immanent content of the psychical act – in the way Rickert did, and in spite
of the (solvable) problems Rickert had. No doubt, the immanent content is always
accompanied by the act itself; and we cannot imagine any difference in it while the
psychic act, or its neurological counterpart, remains the same. But does Pythagoras’
theorem really ‘disappear’ when it is not thought of, and can we not imagine that it
is accompanied by very different psychical or neurological processes?

(2) A legal norm comes into existence, that is, gains validity, only as the result of
empirical norm-positing acts. Either there must be a custom, a habitual applica-
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

tion of a norm, or an intentional act of positing a norm, authorized by a higher

norm, so that this norm can be regarded as valid.

(3) The validity of legal norms is conditioned by the fact that the legal system
as a whole is effective (or, rather, not ineffective); either the majority of norms
should be complied with in general, or the norms should be applied by the
bodies responsible to exert coercion.40 Kelsen’s argument why this should be

38
See, for example, Bennett and McLaughlin 2018.
39
Recently, Monika Zalewska applied supervenience theory specifically to Pure Theory concep-
tions. In Zalewska 2020a: 87–8, she interprets the relation between the basic norm and the
efficacy of the fundamental rules of the legal system as one of supervenience; in Zalewska
2020b, she interprets peripheral imputation (208–10) and the connection between act of will and
legal norm (211) as relations of supervenience. She couples these interpretations with an
ambitious attempt to gain material results from them, like the principle of equality.
40
It is not easy to explain how a ‘higher norm’ validating a lower norm might be efficacious;
possibly, the best solution is to say that it is efficacious if it is used by the individuals empowered
by the legal system to posit norms and to execute sanctions as an interpretive scheme for
validating norms.
 50 Philosophy of Law

the case is, however, rather diffuse. In Problem der Souveränität, he maintains
that the principle of efficacy follows from an extralegal principle, namely, the
principle of economy of thought (Kelsen 1920: 98–9). This principle demands
that using a legal system to interpret human actions must find that most actions are
in accordance with it; otherwise, it could not function as an interpretive scheme.
The problem with this argument is that it cannot explain why law, which Kelsen
takes to be an autonomous normative structure, must be used as an interpretive
scheme in this way at all. In the first edition of Reine Rechtslehre he circumvents
this problem in a peculiar way by pointing to the fact that the principle of efficacy
is part of international law; so, as a positive norm of international law, it is at the
same time the basic norm of any municipal order (Kelsen 1992: 61 [Kelsen 1934:
70–2]). And sometimes Kelsen insinuates, quite correctly, that the efficacy
condition is due to legal dogmatics choosing to deal only with those legal orders
that are effective (Kelsen 1920: 94).

(4) As mentioned, it is possible to use the normative ideology of law as

a scheme for interpreting human acts, individuals, objects, or occurrences,
that is, to give them some special meaning by relating them to norms: They
might be legal acts in a narrow sense, acts of applying or positing norms; they
might be criminal acts; an individual might be a judge; an occurrence might
be a theft or a violation of duty. Kelsen expounds this function of legal norms
in a prominent and lengthy passage at the beginning of Reine Rechtslehre
(Kelsen 1992: 8–11 [Kelsen 1934: 2–6]), giving the misleading impression
that law is interpreted reality, or that law’s primary function is it to interpret
reality. This, again, is not the case; law, for Kelsen, in this phase, is an
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

autonomous system of normative judgments which might derivatively be

used to characterize certain acts and objects as ‘legal act,’ ‘theft,’ ‘last
will,’ etc.

(5) Is-judgments and Ought-judgments share an “indifferent substrate,”

something which is neither part of Is-reality nor of the normative sphere,
but which is, in the case of Ought-judgments, that-which-ought-to-be and, in
the case of Is-judgments, that-which-is (Kelsen 1968b: 38, 65; Kelsen
1922b: 99). It is some pre-objective material which is not yet determined
by the categories. Kelsen does not further define it; it might be best under-
stood either as a ‘state of affairs’/‘possible fact’ or as something totally
exempt from conceptualization.

(6) There are, however, two cases where, contraintuitively, Ought and Is are not
related. First, the Ought does not ‘aim’ at an Is. The system of legal norms is
self-contained and neither capable nor in need of being realised (Kelsen 1922a:
 Hans Kelsen’s Normativism 51

78). Second, as the Rechtssatz is just a hypothetical judgment connecting two

states of affairs, legal norms, strictly speaking, have no addressee.

3.2.7 Normativism as Conceptual Reductionism

Kelsen no longer regards those legal concepts which are not a necessary part of
law as a system of Rechtssätze as unproblematic constructions; rather, he takes
them to be “fictions” which may grudgingly be admitted insofar as they simplify
grasping the legal material. Nevertheless, one has to keep in mind that they are
hypostatizations; there are not any objects that correspond to them. This is most
succinctly expressed in Problem der Souveränität, where Kelsen criticizes
contemporary legal thinking:

[T]he view of the legal theorist, which is just as much diverted from the
proper object of his cognizance – the legal norm – to other areas of cognition
as it is marred by hypostatizations and fictions causing pseudo-problems,
cannot get at its original object which is veiled by the desire for visualization,
habitual ways of thinking and complacency of thought. Metaphors are solidi-
fied into independent entities, mere protheses of thought are posited as
realities. They are like ghosts obstructing juridical cognition. (Kelsen 1920:
IV, my translation)

The most prominent of these ghostly protheses of thought is the concept of

state; Kelsen deems the state to be a person and the legal system and the state
to be one and the same – in fact, arguing the case for this is the focus of his
writings in the 1920s. In a similar way, both the natural and the legal person
are identical with parts of the legal system: sets of norms regulating the
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

behavior of one or several individuals (Kelsen 1992: 46–50 [Kelsen 1934:

52–7]). This regulation of behavior is a function of the secondary norms of
‘duties’ and ‘rights,’ which, in turn, can be reduced to the necessary bare-
bones of law: A legal duty is just part of the impersonal objective Rechtssatz;
not performing it is part of the conditions for triggering the imputation of
coercion. The subjective right is a higher-level part of these conditions; its
exercise triggers the existence of a duty (Kelsen 1992: 43–5 [Kelsen 1934: 47–
52]). Both duties and rights and certain human acts can be connected to
a person by “central” imputation, which must be distinguished from “periph-
eral” imputation; as a person is identical with a system of norms, this does not
mean any more than that they are part of the content of a set of norms (Kelsen
1992: 50 [Kelsen 1934: 57]).
Ironically, the Pure Theory set out as an analysis of an existing practice that
makes a claim to objectivity, taking seriously the normative talk contained in
this practice:
 52 Philosophy of Law

If one deprives the norm or the “ought” of meaning, then there will be no
meaning in the assertions that something is legally allowed, something is
legally prescribed, this belongs to me, that belongs to you, X has a right to do
this, Y is obligated to do that, and so on. In short, all the thousands of
statements in which the life of the law is manifest daily will have lost their
significance. (Kelsen 1992: 33 [Kelsen 1934: 35])

And it ends up by reducing these normative elements of everyday legal life to

the theoretical hypothetical judgment of the Rechtssatz, which, as a general law,
is the only irreducible element of the legal sphere, securing its objectivity, not
being directed at anybody.
The extensions of these secondary legal concepts cannot be determined by
relying on the structure of the primary legal norm alone. For all elements
conjointly making up the conditions for imputing coercion are logically equal,
whether it be a delict, an exertion of a right, or some action of an attorney.
Additionally, the content of the specific legal system has to be considered. This
forces Kelsen to distinguish “Rechtswesensbegriffe” (concepts concerning the
essence of law) from “Rechtsinhaltsbegriffe” (concepts concerning the content
of law). While Rechtsinhaltsbegriffe are not necessary for defining law, they
structure its content and are dependent on it. Rechtswesensbegriffe are neces-
sary for defining the law (Kelsen 1925: 18); they consist of the Rechtssatz, its
elements, and its presuppositions and consequences (Kelsen 1925: 60, 262–5).

3.3 The Third Phase: Commonsense Realism

The transition from Kelsen’s second to his third phase marks a break in the
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

philosophical development of the Pure Theory. Kelsen abandons neo-Kantian

transcendental idealism and instead turns to a naive form of commonsense
realism, according to which legal norms are not identical with the cognitive
judgments of legal science; rather, the latter just descriptively reproduce the
norms which are “given” to them (Kelsen 1957a: 268).
Kelsen scarcely makes the realist premises explicit, and he does not explain
why he thought this change of paradigm necessary. There seem to be two main
reasons. First, in the late 1930s, Kelsen dabbled in diverse nonlegal theoretical
matters, in physics and in ‘ethno-sociology’ or cultural history, among other
areas.41 What he found unsettled his assumption that causality, as an a priori
category, was the fundamental constituting principle of natural science. Modern
physics taught him, or so he thought, that a rationally refined causal principle
could only be conceived as expressing a statistic regularity. Ethno-sociology
taught him that causality was no stable or fixed principle; rather, it eventually

41
See especially Kelsen 1941.
 Hans Kelsen’s Normativism 53

emerged in diverse cultures from the normative principle of retribution. During

a chaotic transition period in the 1930s, Kelsen seems to drop Kantian and neo-
Kantian philosophy altogether and to prefer Hume’s theory of causality, as well
as Hume’s version of the Is–Ought dualism (Kelsen 1939: 129). Kelsen had
constructed the legal category of Ought, that is, peripheral imputation, in strict
analogy with the (neo-)Kantian category of causality. Therefore, deconstructing
the concept of causality shook the basis of his theory. Second, there might be
a biographical reason. The rise of National Socialism and the beginning of the
Second World War forced him to emigrate from Europe to the United States in
1940. While neo-Kantianism, although on the wane, was still a philosophical
trend in Central Europe, it was practically nonexistent in the United States, at
least as far as legal theory was concerned. Pragmatism and legal realism ruled
the day. Thus, Kelsen had lost his institutionalized zone of discourse.42
Kelsen’s writings during this phase need not be treated extensively here (even
though they include what is widely regarded as Kelsen’s opus magnum, the
heavily reworked second edition of Reine Rechtslehre from 1960), for two
reasons. First, he altogether avoids digging any deeper into the metaphysical
foundations of his theory. Second, a main tenet of his theory survived the change
of paradigm seemingly unscathed, namely, the thesis that law exclusively
consists of norms so that all legal concepts must be reducible to legal norms.
Therefore, in what follows I will confine myself to outlining Kelsen’s new
epistemology and its repercussions for his theory.

3.3.1 Normative Realism

https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

To see what the turn in Kelsen’s theory around 1940 amounts to, it is instructive
to compare two short programmatic writings, one from the neo-Kantian phase
and one from the realist phase: Kelsen’s foreword to the second edition of
Hauptprobleme der Staatsrechtslehre (1923), on the one hand, and The Pure
Theory of Law and Analytical Jurisprudence (1941), on the other (both are
easily accessible in English). According to its foreword, Hauptprobleme

takes as its point of departure the fundamental dichotomy between Sollen and
Sein [Ought and Is]. . . . Following Wilhelm Windelband’s and Georg
Simmel’s interpretation of Kant, I take the “ought” as the expression for the
autonomy of law – with the law to be determined by legal science – in
contradistinction to a social “is” that can be comprehended “sociologically”.
The norm qua ought-judgment, then, is contrasted with the law of nature, and
the reconstructed legal norm (Rechtssatz), understood as a norm qua ought-
judgment, is contrasted with the law of causality that is specific to sociology.

42
See Paulson 1988a.
 54 Philosophy of Law

For me, therefore, the core problem becomes the reconstructed legal norm,
understood as the expression of the specific lawfulness, the autonomy, of the
law. . . . And from this point of view, the law that is the subject-matter of legal
science emerges as a system of reconstructed legal norms, that is to say, as
a series of judgments. In the same way, nature – the subject-matter of natural
science – represents for transcendental philosophy a system of judgments.
(Kelsen 1998a: 4–5)43

According to The Pure Theory of Law and Analytical Jurisprudence,

if jurisprudence is to present law as a valid system of norms, the proposi-

tions by which it describes its object must be “ought” propositions,
statements in which an “ought”, not an “is”, is expressed. But the proposi-
tions of jurisprudence are not themselves norms. They establish neither
duties nor rights. Norms by which individuals are obligated and
empowered issue only from the law-creating authority. The jurist, as the
theoretical exponent of the law, presents these norms in propositions that
have a purely descriptive sense, statements which only describe the
“ought” of the legal norm. It is of the greatest importance to distinguish
clearly between legal norms which comprise the object of jurisprudence
and the statements of jurisprudence describing that object. These state-
ments may be called “rules of law” in contradistinction to the “legal
norms” issued by the legal authority. (Kelsen 1957a: 268)

According to the earlier text, legal norms are constituted by, or identical with,
the normative judgments of legal science. According to the later text, it is of the
“greatest importance” to confine jurisprudence to “describing” the legal norms
which are “given” to it.
But it is not only commonsense realism which is the more or less tacit
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

background assumption of Kelsen’s theory in this phase; he comes close to

advocating legal realism, as well. Again, he does not do so explicitly. But it is
conspicuous that he avoids any terminology which might imply that legal norms
are abstract entities, belonging to a sphere of ideality. While speculating, in the
neo-Kantian phase, that the spheres of ideality and of normativity might con-
verge, and speaking of validity as the “ideal” existence of the norm, there is no
longer any such talk. With some justification, Alf Ross wrote that it was only the
undefined concept of validity which detained Kelsen from becoming a legal
realist in this phase (Ross 1957: 564–70).

43
There is a caveat concerning Paulson’s translation: Rendering “Rechtssatz” as “reconstructed
legal norm” conceals the fact that the Rechtssatz, though identical with the legal norm, is
constituted as a logical judgment by legal science and that there is no normativity prior to the
Rechtssatz; so it is not a matter of reconstruction, but of constitution. For my own translation see
Section 3.2.4.
 Hans Kelsen’s Normativism 55

3.3.2 Overview: Consequences of the Realist Approach

The consequences of this commonsense realism for Kelsen’s theory are rather
devastating; they can be summarized as follows:

(1) The categories of Is and Ought can no longer be classified as “forms of

cognition”; rather, they must be forms of reality. Kelsen, however, recoils from
acknowledging that there might be something like a “normative reality.”
Accordingly, his explanation of the Is–Ought dualism dwindles down to the
not further developed thesis that an Ought-sentence cannot be derived from an
Is-sentence (and vice versa) (Kelsen 2005: 5–6 [Kelsen 1960: 3]).

(2) The status of the legal norm becomes hazy. Although Kelsen defines ‘norm’ as
the meaning of an intentional act directed at human behavior (Kelsen 2005: 4–5
[Kelsen 1960: 4]), he neither explains what ‘meaning’ comes up to in a realist
context, nor elucidates how the (objective) meaning of an intentional act can
become the object of a cognitive judgment. Besides, the counterpart of the legal
norm in the sphere of Is becomes doubtful. In the neo-Kantian phase, the (general)
legal norm and the causal law, as hypothetical judgments, were juxtaposed. Now it
seems as if the empirical fact from causally ordered reality is the equivalent of the
norm in the sphere of Is, while the natural law is part of the cognitive level – its
counterpart in the Ought-sphere is the Rechtssatz, which is now defined as the
judgment of legal science describing the legal norm.44

(3) The possibility of a logic of norms gets problematic. Kelsen assumes that the
rules of logic are effective between norms “indirectly,” that is, mediated by the
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

sentences describing the norms (Kelsen 2005: 205–6 [Kelsen 1960: 210]); but this
solution is rather absurd. It seems to be a remnant of the neo-Kantian epistemology,
but it is not compatible with holding that norms are “given” to cognition.

(4) The status of the basic norm becomes ambivalent: On the one hand, it is part of
the sphere of objects, because it is a norm which is described by the major premise
of a syllogism establishing the validity of the constitution;45 on the other hand, it
does not rest on an intentional act directed toward human behavior, but on an act of
thinking, and it is not the object, but a presupposition of legal cognition (Kelsen
2005: 9–10, 23 [Kelsen 1960: 9, 23]). These aspects are scarcely compatible with
each other; the basic norm does not make much sense in a realist context.

44
This is not clear, however. See Kelsen 2005: 76 (Kelsen 1960: 80), for the natural law being
placed on the cognitive level, and Kelsen 2005: 75 (Kelsen 1960: 78), for the causal principle
being at work in nature.
45
See Section 3.3.5.
 56 Philosophy of Law

(5) In addition, the role of legal science becomes questionable. On the one hand,
it is restricted to describing the legal norms given to it; on the other hand, the
legal norms are just the result of an interpretation by the legal scientist (Kelsen
2005: 45–7 [Kelsen 1960: 46–8]). It is difficult to see how these tenets might go
together.

3.3.3 Causality, Imputation, and Empowerment

Even though Kelsen’s engagement with the results of modern physics and cultural
history had shattered his belief that causality and imputation were original and
equal-ranking categories of cognition, he does not express his doubts in the main
text of this period, Reine Rechtslehre (1960). And the dualism, for him, is still one
of the manifestations of the more fundamental dualism of Is and Ought. But it
loses importance and, without its defining neo-Kantian context, is somehow an
alien element in Kelsen’s theory. He has difficulty in ascribing imputation either
to the sphere of objects or to the level of cognition. According to some passages, it
is a connection on the level of the norms themselves (Kelsen 2005: 77–8 [Kelsen
1960: 81]); in other places, it appears to be a connection on the level of norma-
tively interpreted reality (Kelsen 2005: 76 [Kelsen 1960: 79]). According to
a third explanation, imputation is a principle which is applied when describing
a normative order of human behavior (Kelsen 2005: 76 [Kelsen 1960: 79]), while
causality rather connects the elements making up nature and is described by the
natural laws (Kelsen 2005: 75 [Kelsen 1960: 78]). Yet this would not be coherent.
If cognition is just a descriptive isomorphous reproduction of a world ‘given’ to it,
then a description of a norm cannot employ the principle of imputation, unless
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

this principle is given in some way in the normative world itself.

This matter, however, need not further be discussed here. For imputation, which
was the most fundamental Ought-modality in the neo-Kantian phase, steps back
together with the assumption that the legal norm is identical with the impersonal
judgment of legal science. It is superseded by a very different normative function.
Kelsen now regards the Ought as a portmanteau concept, covering different
kinds of normative functions: commanding, prohibiting, permitting, empower-
ing, and derogating (Kelsen 2005: 4–5 [Kelsen 1960: 4–5]).46 And the norma-
tive function of the general legal norm – and, thus, the primary normative
function in law – is “empowerment” (Kelsen 2005: 25, 56–7 [Kelsen 1960:
26, 57-8]);47 all other normative functions are part of “dependent” norms which
can be reduced to the general norm.

46
Thus, other than in the second phase, the legal norm has an addressee.
47
On this topic, see especially Paulson 1988b.
 Hans Kelsen’s Normativism 57

Basically, the general norm says that, under certain conditions, somebody
is empowered to exert coercion. How can somebody be empowered to exert
coercion? Kelsen distinguishes two ways of conceiving the general norm
(Kelsen 2005: 57, 231 [Kelsen 1960: 57–8, 237]). From a “dynamic” per-
spective, the law is a hierarchy of norms. General norms constitute just one
of several levels; they decree that under certain conditions a judge is
empowered to issue an individual norm. “Empowerment” in this case
means that the judge is able to bring about the conditions for the individual
norm (the sentence she pronounces) to be valid; so this empowerment does
not directly concern the exercise of coercion as a brute act but only the
ordering of an act of coercion. From a “static” perspective, comparable to
the conception of Hauptprobleme, there is just one level of law, the level of
heavily inflated (because they contain all conditions for legally exercising
coercion) general norms that run, in simplified form, something like:

If the constitution validates a law which empowers a judge, under certain

conditions, to decree that an act of coercion ought to be performed, and if
under such conditions there is such a decree, then the act of coercion ought to
be performed.

In both cases, the normative function is that of empowerment. In the case

of the general norm according to the dynamic perspective, empowerment
confers the ability to fulfill the conditions for an individual norm to be valid;
in the case of the general norm according to the static perspective, it confers
the ability to perform an act of coercion which is qualified as a ‘legal act.’ So
empowerment is, in a way, the successor to imputation as the basic norma-
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

tive function of legal norms. There is a direct line from attributing an act – as
a legal act – to the constructed person of the state as ‘willed’ by it, in the first
phase of Kelsen’s theory, via peripheral imputation, as the relation between
condition and sanction inside the Rechtssatz in the second phase, to
empowerment, as bestowing the capacity to bring about the conditions to
make a lower norm valid or to vest an act with the property of being a legal
act.

3.3.4 Prescriptive and Descriptive Ought; Subjective and Objective Ought;

Validity
As a consequence of the separation of the cognitive level of judgments from
the level of “given” objects, Kelsen distinguishes a “prescriptive” from
a “descriptive” Ought. The prescriptive Ought is the Ought contained in the
norms, while the descriptive Ought is its descriptive reflection at the level of
dogmatic judgments (Kelsen 2005: 73–5 [Kelsen 1960: 75–7]). This
 58 Philosophy of Law

distinction is scarcely argued for, and Kelsen does not make a similar differ-
entiation for the sphere of Is.48
Another differentiation makes more sense. Kelsen distinguishes
a “subjective” Ought from an “objective” Ought. This distinction is an
advancement of a similar differentiation from the neo-Kantian phase:
According to Kelsen’s writings before 1940, the nonnormative material
given to legal cognition to be formed into a norm is the content of some
human act which claims to have normative impact according to its “self-
interpretation.” This is its “subjective” meaning. Its “objective” meaning,
that is, whether it really gives rise to an Ought, is a matter of its interpretation
in the light of a valid norm (Kelsen 1992: 9–10 [Kelsen 1934: 2–5]). In
a similar vein, the “objective Ought,” according to Kelsen in 1960, is the
“valid” ought, the “real” Ought, which is given if it is not only claimed that
something ought to be done, but if something really ought to be done.49 The
subjective Ought is given – this is the most plausible interpretation if one
considers the genesis of this conception – if it is (only) claimed that something
ought to be done. According to another interpretation, a subjective Ought is
given if someone wants somebody else to behave in a certain way.50 The
subjective Ought is bound to an intentional act, whose meaning it is, while the
objective Ought is valid even if it is no longer backed by an intentional act
(Kelsen 2005: 7–8 [Kelsen 1960: 7]).
So, the underdetermined concept of validity indeed seems to bear the brunt in
dragging Kelsen’s notion of “objective normativity” through the realist phase
without succumbing either to naturalism or to Platonism. This is reminiscent of
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

Rickert’s efforts to fend off psychologism without giving in to Platonic hyper-

realism as depicted in Section 2.3.2 – only, Kelsen, in this phase, lacks the
appropriate philosophical background theory to argue for his case.

3.3.5 The Legal Hierarchy and the Basic Norm

In this third phase, there are two major innovations concerning the legal hierarchy.
On the one hand, Kelsen leaves the opaque quasi-Kantian terminology behind in

48
The distinction makes sense on a linguistic level: A prescriptive Ought might, cum grano salis,
be said to be involved in those performatives which are uttered in order to bring a norm in to
existence, while the descriptive Ought would be contained in assertions of a valid norm which do
not aim to ‘bring it about.’
49
Kelsen 2005: 7–8 (Kelsen 1960: 7–8). The German original contains a footnote in which Kelsen
declares that he could also have chosen the expression “real Ought” to denote the objective
Ought, if it had not been for the connotation of the expression “real” (tatsächlich) which point to
the sphere of Is.
50
This is implied by Kelsen’s most famous example of a (solely) subjective Ought: the command
of a criminal to hand over money; cf. Kelsen 2005: 44 (Kelsen 1960: 46).
 Hans Kelsen’s Normativism 59

describing the relation between higher and lower norm, and he maintains that this
relation can be captured in the form of a syllogism. On the other hand, he brings
forward the odd thesis, mentioned in Section 3.3.2, that logical relations are at work
between the given norms not directly, but indirectly, mediated by the judgments or
statements about the norms. The (statement of the) higher norm plays the role of the
major premise in this syllogism, determining factual conditions for the existence or
validity of the lower norm. The minor premise is the (statement of the) fact
fulfilling these conditions; the conclusion is the (statement of the) lower norm. In
Kelsen's own words:

The foundation of the validity of a positive norm, that is, one established by
an act of will and prescribing a certain behaviour, is the result of a syllogistic
procedure. In this syllogism, the major premise is the assertion about a norm
regarded as objectively valid, according to which one ought to obey the
commands of a certain person, that is, one ought to behave according to the
subjective meaning of these commands; the minor premise is the assertion of
the fact that this person has commanded to behave in a certain way; and the
conclusion is the assertion of the validity of the norm: that one ought to
behave in this particular way. Thus the norm whose validity is stated in the
major premise legitimizes the subjective meaning of the command, whose
existence is asserted in the minor premise, as the command’s objective
meaning. (Kelsen 2005: 202 [Kelsen 1960: 205])

The main flaw of this passage has been mentioned: The idea of an “indirect”
logic of norms is absurd; there is no way that a norm understood as
a nonpropositional entity in a realist sense could be subjected to a propositional
calculus. This apart, there are two possible ways of reconstructing the syllogism.
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

According to the first, the major premise, simplified, would run as follows:

If there is a (normative) utterance ‘n’ under certain conditions and with

a certain content, then ‘n’ is true (so n is a valid norm).

This is comparable to the conception of the preceding phase: The higher norm
would not be a norm in the usual sense, but a rule of a meta-level that states
criteria for a lower norm to be valid, or, what amounts to the same, for a statement
about a norm to be true. According to the second reconstruction, the higher norm
does not just state general validity conditions; rather, it directly establishes an
Ought for the addressees of the lower norms. Simplified, it runs as follows:
If the legislator (or some other body) has, under certain conditions,
commanded you to do a, then you ought to do a.

This sounds more like a genuine norm. However, either the lower norm is
superfluous or the Ought is, again, reduplicated. If ‘Oa’ can be inferred from '(If c,
then Oa), and c,’ then there is no normative surplus in it. After all, a syllogism
 60 Philosophy of Law

does not create ‘new’ truths; it just makes explicit conceptual relations. This
seems to be different in the meta-level reconstruction, because the meta-level rule
does not contain an Ought. Besides, only the meta-level reconstruction seems to
capture the feature that the Ought-function, on all levels of the hierarchy, is
empowering (not commanding).
Anyway, neither reconstruction really corresponds to what one would nor-
mally expect a derivation of an Ought from another Ought to look like, and both
express the necessary role of ‘higher’ rules – no matter whether they are norms
or meta-rules – in naming factual criteria for the existence of lower norms.
Positivistic normativism is not possible without such rules.
Even though the basic norm, as the highest (or most fundamental) norm,
survives Kelsen’s turn to realism seemingly unscathed, it is, like imputation, an
alien element in the realist phase. It is described by the major premise of the
most fundamental syllogism in the legal hierarchy, establishing the validity of
the constitution. It is still called a “hypothesis” (Kelsen 1957b: 260–2) and
a “transcendental-logical condition,” the latter with the restriction that this term
is used per analogiam (Kelsen 2005: 201–2 [Kelsen 1960: 203–5]). But it is
difficult to see what this analogy might look like. Transcendental-logical con-
ditions are only conceivable as necessary conditions of cognition; having
a system of ‘given’ legal norms depend on a presupposition of cognition is,
however, contradictory. Either the norms are given to cognition, but then they
cannot depend on a quasi-transcendental-logical condition; or they depend on
a condition of thinking, but then they are not ‘given’ to cognition.
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

3.3.6 Relations between Is and Ought

Compared to the second phase, there are several changes concerning possible
relations between Is and Ought.
First, the theory of the “indifferent substrate” is not dropped, but neglected.
Kelsen just vaguely talks of a “something” which either might be the case, or
ought to be the case, as the possible common or comparable content of Is and
Ought (Kelsen 2005: 6 [Kelsen 1960: 6]).
Second, the norm, as an objective meaning-content, is conditioned by, or the
meaning of, an intentional act, which belongs to the Is-sphere and is directed at
somebody else’s behavior (Kelsen 2005: 4–6 [Kelsen 1960: 4–6]). It is not quite
clear what Kelsen means by an “intentional act.” Considering his philosophical
background, one might reasonably assume that the term is borrowed from
Husserl’s philosophy. For Husserl, an intentional act is a psychic act directed
at some (propositional or nonpropositional) object. So a norm is the meaning of
a psychic act directed at the propositional object “that someone acts in a certain
 Hans Kelsen’s Normativism 61

way.”51 Sometimes, Kelsen seems to hold that such an act is an act of will. But
this would contradict his former theses that one might only will one’s own
behavior, that norms are not necessarily connected with a ‘real’ act of will, and
that possibly there is no psychic function of willing at all which is not
a normative construct.52 And Kelsen does not maintain that being the meaning
of an act of will is a necessary feature of any norm. Apart from the possibility of
deducing norms from other norms, there is the counterexample of the basic
norm which is the meaning of an “act of thinking” (Kelsen 1960: 9).53 Besides,
it is not reasonable to condition the legal norm on an internal act of will; it is
more plausible to condition it on the expression of a corresponding will. And
this is exactly what Kelsen does in several passages.54
However this may be, the norm, as an Ought, has to be strictly distinguished
from the act whose meaning it is. The statement that somebody ought to do a,
cannot, for Kelsen, be reduced to the statement that somebody else wants this
somebody to do a (Kelsen 2005: 5 [Kelsen 1960: 4]). And Kelsen maintains that

it is entirely possible to describe the relation between a behaviour and the

norm stipulating that this behaviour ought to be, without taking into consid-
eration the act of command or custom by which the norm was created. This is
obvious, for example, when we think . . . of norms created by the custom of
earlier generations, so that the men whose conduct is regulated by these
norms are aware of them only as meanings. (Kelsen 2005: 22 [Kelsen
1960: 23])

This passage marks a central point where Kelsen’s writings of the third phase
differ from those of the fourth phase: He considers it possible to think of a norm
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

without thinking of the act of will by which it came into existence.

Third, the validity of every single legal norm is conditioned on its efficacy
(Kelsen 2005: 11 [Kelsen 1960: 11]). By contrast, in the neo-Kantian phase
Kelsen conditioned (only) the validity of the whole legal system on its efficacy
in general. This innovation is highly problematic. It is not argued for, and it is
not quite clear how it was motivated. It does not seem to apply to statutory law;
and it seems strange to say that, for example, an adjudication is invalid because

51
It must be noted, though, that this interpretation is at odds with Kelsen’s position, in this phase, of
avoiding tagging the norm as an abstract entity.
52
On Kelsen’s changing opinions concerning the concept of will, see Section 3.4.4.
53
Kelsen changed the relevant passage for the English translation which was published during the
previous phase; in Kelsen 2005: 9–10, he maintains that it is impossible to think of a norm
without thinking, along with it, of the act of will whose meaning it is.
54
Cf. Kelsen 2005: 5 (Kelsen 1960: 4): “If an individual by his acts expresses a will directed at
a certain behaviour of another, that is to say if he commands, permits or authorizes such
behaviour – then the meaning of his acts cannot be described by the statement that the other
individual will (future tense) behave in that way, but only that he ought to behave in that way”
(my emphasis).
 62 Philosophy of Law

it was not subsequently carried out. This problem seems to be related to the
problem of how Kelsen can account for customary law in his theory; however, it
need not be pursued in this context.
Fourth, as shown, other than in the second phase, the (primary, independent)
legal norm has an addressee, namely, the empowered individual.

3.3.7 Legal Concepts

As far as the reducibility of central legal concepts to the autonomous system of

legal norms is concerned, there are no relevant changes. Central imputation
(zentrale Zurechnung) is now called “attribution” (Zuschreibung), but it is still
the relation of an act to a “person,” that is, to a set of norms regulating (via
secondary norms like commands or prohibitions) the behavior of one or several
human beings which are identified by general criteria (Kelsen 2005: 150
[Kelsen 1960: 154–5]). The concepts of duty and (subjective) right, and the
secondary norms of prescription, prohibition, etc. are ‘reduced’ to law as a set of
(primary) norms in the same way as before (see Kelsen 2005: chapters 28–33
[Kelsen 1960: chapters 28–33]).

3.4 The Fourth Phase: The Will Theory of Norms

The fourth and last phase of Kelsen’s theory comprises his writings after 1960.
Around the time of the publication of the second edition of Reine Rechtslehre,
Kelsen was engaged in an exchange by letter with Ulrich Klug who had recently
published a book on the logic of law relying on modern formal logic (Kelsen and
Klug 1981). Apparently influenced by Klug, Kelsen abandoned the thesis that it
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

might be possible that logical rules define the relations not only between judg-
ments about norms, but also, mediated by these judgments, between the norms
themselves (Kelsen and Klug 1981: 48–9). Independently from the exchange with
Klug, he also abandoned the thesis that a norm might be the meaning-content of
an act of thinking, and tied the norm strictly to an act of will: “no norm without an
act of will whose meaning it is” (Kelsen 1991: 234 [Kelsen 1979: 187]).
But there are more changes. Kelsen seems to have realized around 1960 that
his philosophical background might need to be updated and expanded, and he
gathered material from innumerable contemporary sources – especially from
linguistic philosophy, moral philosophy, and logics – to undergird his
conceptions.55 But his theory remains something of a patchwork. This is, in
part, due to the fact that there is no comprehensive monograph dealing with it as
a whole. An exception seems to be Allgemeine Theorie der Normen (General

55
On the literature studied by Kelsen, see Opałek 1980: 11–18.
 Hans Kelsen’s Normativism 63

Theory of Norms). But it was published posthumously, compiled from Kelsen’s

Nachlass. It contains many inconsistencies, and it is an open question whether
Kelsen himself would have published it without major emendations.
In secondary literature, the writings of this last phase are often regarded as
inducing a ‘break’ in the genesis of the Pure Theory, and the innovations they
included are mostly rejected. The new conceptions have been berated as “norm-
irrationalism” (Weinberger 1981: 94), and as destructive of the main tenets of
the classical Pure Theory (Paulson 2013). But I tend to think that the ‘break’ in
the theory occurred around 1940, and that Kelsen’s writings after 1960 are an
attempt to deal with the problems caused by his previous half-hearted realist
conception of normativity. And if he wrecked his theory, at least he did so in
style, and not by philosophical parsimony.

3.4.1 Foundations

The times of philosophical austerity are past; the foundations of the Pure
Theory are fleshed out, once more, albeit in different directions. First, Kelsen
sticks to the commonsensical realism from the third phase – again, without
elaborating it: “As a science, legal science can only know and describe the
norms which are given to it” (Kelsen 1991: 153 [Kelsen 1979: 123]). But at the
same time, the sphere of ideality is rehabilitated; Kelsen no longer avoids
calling the validity of the norm its “ideal” existence (Kelsen 1968c: 150).
Second, Kelsen undergirds his theory with a theory of speech-acts on the one
hand, according to which norms are based on acts of “commanding” (Kelsen
1991: ch. 10 [Kelsen 1979: ch. 10]), and a theory of intentional acts, according
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

to which norms are based on psychic acts of will (Kelsen 1991: ch. 9 [Kelsen
1979: ch. 9]), on the other. This is supplemented by a semantic theory,
according to which norms and assertions are the meaning-contents of different
types of sentences.

3.4.2 Is and Ought

This new abundance in respect of the philosophical foundations of the Pure
Theory has consequences for Kelsen’s conception of the dualism of Is and
Ought:

(1) Sometimes Kelsen still denotes Is and Ought as “categories,” but he does not
explain these elements in terms of neo-Kantian philosophy. Instead, citing
Henry Sidgwick, he calls them “undefinable basic concepts.” And he still relies
on the “logical principle” that it is not possible to deduce a statement about
a norm from a statement about a fact (Kelsen 1991: 58 [Kelsen 1979: 44]).
 64 Philosophy of Law

(2) In addition, there is a rudimentary ontological explanation of the dualism,

according to which the Is is the aggregate of objects or occurrences subjected to
causal laws, while the Ought is the aggregate of norms, understood as ideal
entities (Kelsen 1968c: 150).

(3) The main explanation is based on an amalgam of semantics, linguistic

pragmatics, and a theory of intentional acts. It might charitably be reconstructed
as follows: Is and Ought are different modes in different classes of sentences.
The Is is the mode of assertive sentences and the Ought is the mode of normative
sentences. The mode of an assertive sentence is that something is the case; the
mode of a normative sentence is that something ought to be the case. Both
sentences have a “content” which might be the same (or at least comparable);
Kelsen labels it “modally indifferent substrate” (Kelsen 1991: 60–1 [Kelsen
1979: 46]).

Sentences of these types are embedded in different pragmatics, and there

are different types of intentional acts that coordinate with them: Assertive
sentences are used to let somebody else know something, while normative
sentences are (foremost) used to let somebody want something (Kelsen
1963: 2). Norms, as the meaning of normative sentences, are necessarily
based on acts of will; propositions, as the meanings of assertive sentences,
are based on acts of thinking. The concept of a ‘meaning’ of a sentence is not
further explained by Kelsen. Sometimes he seems to think that the meaning of
a sentence is identical with its Is or Ought mode (Kelsen 1963: 2; Kelsen 1991:
26 [Kelsen 1979: 21]); it would be more plausible to say that it is made up of
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

mode and content.

This is certainly an improvement compared to the conception offered in the
preceding phase. The status of the norm as a meaning-content of a sentence is
clarified; and its linguistic counterpart concerning the sphere of Is is determined
in an unequivocal way. However, there are new problems.

3.4.3 Subjective and Objective Ought and Their Relation to Empowerment

Again, Kelsen distinguishes a subjective from an objective Ought. The object-

ive Ought is still defined as before; it is the valid Ought which is mandatory for
its addressee (Kelsen 1991: 27 [Kelsen 1979: 22]). But while the subjective
Ought appeared, in the realist phase, to be (predominantly) a purported or
claimed objective Ought, it is now the subjective Ought which is more
fundamental.
The subjective Ought is the meaning of an act of commanding. An act of
commanding is an act of will directed at the behavior of somebody else. As it is
 Hans Kelsen’s Normativism 65

impossible ‘directly’ to want the behavior of someone else – here, Kelsen

resuscitates a thesis from his first phase – the meaning of the act of will is not
that somebody behaves in a certain way, but that somebody ought to behave in
a certain way Kelsen 1991: 26, 31, 44–5 [Kelsen 1979: 21, 24, 35]). Kelsen,
however, does not further explain how (impossibly) willing that somebody else
acts in a certain way miraculously transforms into wanting that he ought to act
in a certain way.
Two aspects are noteworthy. On the one hand, Kelsen seems to think that the
internal, psychic, intentional act of willing somebody to do something is, in this
context, necessarily connected with giving it expression by uttering a command
(Kelsen 1991: 26 [Kelsen 1979: 21]). On the other hand, it seems to be the
meaning of Ought that someone wants somebody else to act in a certain way.
Accordingly, the objective Ought seems, for Kelsen, always to be ‘grounded’ on
a subjective Ought;56 it is given if there has been an act of commanding (and
a subjective Ought with it), which has been empowered (Kelsen 1991: 233–4
[Kelsen 1979: 27–8]).
Even if we presume that this conception is coherent, it raises a number of
questions. First, from the beginning, it was Kelsen’s concern to secure the status
of legal science as a true, autonomous science. This affords in Kelsen’s opinion,
that its subject matter is both normative and objective. In the neo-Kantian phase,
this was achieved by taking norms to be cognitive normative judgments; a valid
normative judgment contains an objective Ought, and a subjective Ought is
contained in claims or beliefs about this objective Ought. But if, in this late
phase, the meaning of Ought is fully established by there being an act of
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

commanding, how could there possibly be some surplus quality which trans-
forms it into an ‘objective’ Ought, that is, morphs it into ‘one really ought to act
like this’? And what, actually, does it mean, in the will theory of norms, to say
that one ‘really’ ought to act in a certain way? Second, and connected with the
first point, what does ‘empowering’ mean if it does not somehow ‘legitimize’
the Ought claimed by someone; and can a ‘legitimate’ Ought mean anything but
that one ‘really’ ought to act in the way prescribed, so that the objective Ought
would, after all, be more fundamental?
Kelsen’s explanation of the concept of empowerment is rather ambiguous.
On the one hand, he still thinks that it is, in law, the primary normative function
(Kelsen 1991: 97 [Kelsen 1979: 77]); and he defines it as “conferring on an
individual the power to posit or apply norms” (Kelsen 1991: 102 [Kelsen 1979:
82–4]). But this seems to conflict with his general definition of the Ought as the

56
This follows from the thesis that every norm includes and is determined by a command; Kelsen
1991: 233–4 [Kelsen 1979: 186–7].
 66 Philosophy of Law

meaning of an act of commanding. And so there is another explanation accord-

ing to which empowerment appears as a ‘conditional command’: If empower-
ment concerns the positing of norms, it ‘implies’ a command:

[B]y empowering the legislative organ to enact statutes binding on the

the subjects of the law, the constitution empowers the legislator to make
the subjects’ behaviour which does not agree with the statutes the
condition for sanctions and thereby make the statutes he enacts binding
on the subjects of the law. Thus the subjects are bound in the last
analysis by the constitution itself . . .. In other words, they are com-
manded by the constitution to comply with the statutes. (Kelsen 1991:
104 [Kelsen 1979: 83])

There seem to be three ways of interpreting this passage: The first interpretation
is that the legal system consists of just one command, contained in the constitution
(or even in the basic norm), saying that one ought to obey the legislator (or the
authors of the constitution); all other norms are simply ‘technical rules’ stating what
one has to do if one wants to obey the constitution (or the basic norm). Or, on
a second interpretation, we have a curious duplication of commands; it is not only
the lawgivers who command that, for example, theft ought to be punished, but also
the authors of the constitution. Both reconstructions seem to be slightly strange, but
they are possible, of course. In both cases, empowerment could be reduced to the
function of commanding. Yet, in neither case would it make sense to distinguish
between a subjective and an objective Ought. Besides, as commanding or prescrib-
ing, in law, is just the function of secondary norms, there seems to be a logical
circle: The normative function of primary norms, empowering, can be reduced to
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

the normative function of secondary norms, commanding; but secondary norms

can, in turn, be reduced to primary norms. The third interpretation would be to
return to the conception according to which empowerment ‘validates’ the Ought;
but then there seems to be, again, a primacy of the objective Ought, which grates
harshly with the command or will theory of Ought.

3.4.4 The Will Theory of Norms and Validity

A kindred problem is the relation between the will theory of norms and the
concept of validity. As shown in Section 2.3, the concept of validity was central
to the formation of normativism – and to Kelsen’s theory in the neo-Kantian
phase. Validity, in this last phase, is again defined as the “specific existence” of
the norm; it is analogous not to the truth of an assertion but to the existence of
a fact: “That a norm prescribing a certain behaviour is not valid means that this
norm does not exist. ‘A valid norm’ is a redundant expression. ‘An invalid
norm’ is a contradiction in terms. . . . Truth is a property of the statement.
 Hans Kelsen’s Normativism 67

Validity is not a property of the norm, but its existence” (Kelsen 1991: 171
[Kelsen 1979: 137]). And, as mentioned, Kelsen no longer avoids calling this
kind of existence “ideal.” It consists in that the norm should be complied with
or applied – independent of its recognition by the person to whom it is
addressed.57
All this is in accordance with most of what Kelsen said in the preceding
phases, but it is scarcely compatible with his strict will theory of norms. A norm
exists or is valid if one ‘really’ ought to act in the way it prescribes. This implies,
again, that the ‘objective’ Ought is the genuine Ought, while the subjective
Ought is just a purported or claimed Ought. In contradistinction, according to
the will theory the subjective Ought is fundamental; and the objective Ought is
a specially qualified subjective Ought.
This opaqueness is partly due to Kelsen not really offering a clear-cut concept
of an act of will; so this topic deserves a short digression. Three questions, in
particular, connected with the concept of will, are not answered clearly in any of
the phases of the Pure Theory:

(1) Is the act of will an internal psychic act, or is it the external expression of
someone to the effect that she wishes somebody to do something?
(2) Is the act of will necessary for the existence of the norm in the sense that the
norm only exists as long as there is an act of will as its ‘bearer,’ or is the act
of will just necessary to ‘trigger’ the existence of the norm, so that it is
possible that the norm exists while the act of will is no longer extant?
(3) Is the act of will part of the definition of the norm, or is it just a condition of
its existence?
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

The most plausible solution, at least for the sake of reconstructing law, is
that the act of will is an external act, and that it is necessary only to ‘trigger’
the norm. The problem is that Kelsen, in this phase, sees a much closer
connection between Ought and will. Even in the neo-Kantian phase, he had
an ambiguous understanding of the concept of will. Sometimes he doubted
whether there could be something like a will which is not a normative con-
struct (Kelsen 1922b: 241–5). So the will might be reducible to normativity. In
other places, he took the will to be a purely psychic category (Kelsen 1922a:
182). In that case, will and Ought would be conceptually independent from
each other. Then again, he defined the Ought as an impersonal objective
willing (Kelsen 1920: 9 fn.), so that the will seems to at least partly define
the Ought.

57
But not independent of recognition altogether, cf. Kelsen, 1991: 3, 50–1 (Kelsen 1979: 3, 39–40).
 68 Philosophy of Law

But if we take seriously Kelsen’s thesis from General Theory of Norms, that
‘Ought’ means that someone wants somebody else to act in a certain way, the
conclusion that Ought and willing are conceptually interwoven can hardly be
avoided. If we accept that, however, it would not make much sense to talk of
‘validity’ and of an ‘objective’ Ought; and the nature of empowerment would be
a mystery. Normativism, as propagated by Kelsen in the 1920s, would finally
have evaporated.

3.4.5 Prescriptive and Descriptive Ought; Imputation

The differentiation between a prescriptive and a descriptive Ought is slightly

modified in this last phase. Kelsen still maintains that Ought sentences can be
used both to describe and to posit a norm (an “imperative,” by contrast, can only
be used to command or to express a norm – to prescribe something) (Kelsen
1991: 150–1 [Kelsen 1979: 120–1]). Sometimes he comes close to holding that
the difference is less one of the semantics but of the pragmatics of Ought
sentences58 (which seems to be correct), and sometimes he seems to say that
a descriptive Ought sentence might be better formulated as a sentence about the
validity of a norm (not as “x ought to do a,” but as “the norm, that x ought to do
a, is valid”) (Kelsen 1991: 155 [Kelsen 1979: 125]).
Somehow, the concept of imputation has managed to make it into these
strange new surroundings:

In the linking of a condition and a sanction brought about by a general moral or

legal norm and described by ethics and legal science . . . we encounter a principle
which is different from the principle of causality expressed in the natural laws
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

formulated by the natural sciences. It is different, but analogous. I have suggested

calling it the Principle of Imputation. (Kelsen 1991: 24 [Kelsen 1979: 19])

As before, it is not quite clear whether imputation is a matter at the cognitive

level, a relation inside the norm, or part of normatively interpreted reality.

3.4.6 The Legal Hierarchy and the Basic Norm

Further problems, where the status of empowerment is involved, are the relation
between a “higher” and a “lower” norm in the legal hierarchy, and the status of
the basic norm. On the one hand, Kelsen abandons the conception of an
“indirect” logic of norms – a logic that is mediated by the logic of the assertions
describing the norms; on the other hand, both his will theory of norms and the
conception of the norm as a nonpropositional object given to cognition seem to

58
That is how I read the chapter concerning Sigwart, see Kelsen 1991: 151–3 (Kelsen 1979: 121–
3).
 Hans Kelsen’s Normativism 69

make it impossible that there can be logical relations between norms at all
(Kelsen 1991: 211, 217 [Kelsen 1979: 166, 171]).59
Kelsen describes the relation between a higher and a lower norm vaguely as
one of “correspondence.” His explanation of the relation between a general
norm and an individual norm is instructive. Imagine a general norm as follows:

If the competent judge has ascertained that somebody committed a theft, he

ought to decide that this person ought to be imprisoned for one to five years.

There are two individual norms which might be coordinated to it.

(1) After ascertaining that Jones has committed a theft, the competent judge
decides

Jones ought to be imprisoned for one year.

This is, as a judicial sentence, a classical example of an individual norm; it

cannot be deduced from the general norm but needs an additional act of will on
the part of the competent judge. It might be seen as the result of compliance with
and appliance of the general norm by the judge.

(2) But there is another individual norm involved. If the judge has ascertained
that Jones committed a theft, then it seems that she – or anyone – might at least
‘logically’ conclude, using modus ponens and the rule of substitution, that she
ought to decide that Jones ought to be imprisoned, so that there is a second
individual norm:

The judge ought to decide that Jones ought to be imprisoned for one to five
years.
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

The possibility of such a logical conclusion is, however, negated by Kelsen.

He holds that this individual norm is valid only if it is the outcome of an act of
will on the part of the judge which is identical with her recognizing the higher
norm. This recognition, again, seems, in turn, to condition the validity of the
higher norm (Kelsen 1991: 239–41 [Kelsen 1979: 191]).60
This conception of the relation between individual norm and general norm,
or “lower” and “higher” norm, has disastrous consequences. It contradicts

59
I cannot delve into the conception of the relation between norms and logic according to the
General Theory of Norms here. In short: Norms, in principle, are neither subject to the law of
noncontradiction nor to the rules of inference. But there can be logical relations between
concepts contained in norms, so that it is possible to derive a less general norm from a more
general norm. And hypothetical norms have a logical ‘if-then-structure’ (which apparently
doesn’t allow for the modus ponens to be applied; also, Kelsen is not quite clear on the point
whether this structure is part of the norm’s content or whether [the validity of] the norm itself is
conditioned). See Kelsen 1991: 249–50, 266–7 [Kelsen 1979: 201, 215–16].
60
For a depiction of this problem, see Heidemann 2000: 275–8.
 70 Philosophy of Law

what Kelsen said when explaining the concept of validity; in that context, he
maintained that the norm’s validity was independent of its acceptance by the
individual subjected to it. Besides, it is scarcely compatible either with
normativism or with the way jurists talk about law and ascertain its validity;
it is ultimately expressive of a recognition theory of law.61
Finally, the basic norm is probably what first comes to the mind of most
legal theorists when asked what radical change occurred in Kelsen’s theory
after 1960. For Kelsen abandons the idea that the basic norm might be
a hypothesis, resting on an act of thought, and instead maintains that it is
a fiction in the sense of Hans Vaihinger’s philosophy of “As if,”62
a pragmatic mental tool to enable the grasp of phenomena that cannot be
understood by employing the given conceptual apparatus. Judgments pre-
supposing the existence of a fictive object are false, but they can be justified
by their pragmatic function. A “full fiction,” according to Vaihinger, contra-
dicts reality and, besides, it is self-contradictory. Kelsen holds that the basic
norm contradicts reality because there is no such thing; his new will theory
of norms affords that any norm is based on a real act of will, and there is
nothing like that to support the basic norm. It is also self-contradictory
because it authorizes a highest legal authority while, as a norm, presuppos-
ing a higher authority on whose (empowered) act of will it rests (Kelsen
1991: 256 [Kelsen 1979: 206–7]).
In this context, it is not necessary to deal with Vaihinger’s highly idiosyn-
cratic philosophy. A few remarks must suffice. First, the change is less
serious than it might appear at first sight. In Kelsen’s third phase, the basic
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

norm somehow oscillated between the cognitive level and the object level;
now it is taken to be a ‘feigned object,’ which is a clarification. Yet even as
a fiction, the basic norm is a presupposition of legal thinking; and Kelsen
does not stop labelling it as a transcendental-logical condition of legal
cognition (Kelsen 1968d: 1976). Regarding it as a fiction was occasioned
by the need to accommodate the act of will necessarily connected with every
norm. That is not an important matter; the real upheaval took place when
Kelsen started seeing a necessary connection between any Ought and
a factual act of will. Second, it is only the act of will the basic norm is

61
To be fair, this problem is not just brought about by the strict will theory of norms. It is also
a product of the combination, existent even in the third phase, of the thesis that the validity of
each single norm is dependent on its efficacy, and the problem of how an empowerment norm
might be effective.
62
Hans Vaihinger (1852–1933) started as a neo-Kantian. He wrote Die Philosophie des Als Ob over
several decades. The book, published in 1911 was a success; ten editions were produced (the last
in 1927). Like many neo-Kantian texts, it is slightly quixotic (indeed, on the brink of dubious-
ness), combining pragmatism, positivism, and idealism.
 Hans Kelsen’s Normativism 71

supposed to rest upon which is not given in (natural) reality. So the (assump-
tion of the) basic norm contradicts reality only if the latter incorporates this
act of will in a way that it is part of the Is-and Ought-spheres at the same
time.63 Third, as norms, according to Kelsen in this phase, are not capable of
logical relations, how could the basic norm be self-contradictory? At best,
its content could be self-contradictory. But that would presuppose that
“being legitimized by a higher authority” was part of its content – which is
difficult to construct. It is more plausible to say that asserting the validity of
a basic norm contradicts normative reality64 and is self-contradictory,
because it entails that there is a norm which establishes a highest authority
while at the same time being legitimized by a yet higher authority. But even
in this interpretation, it is doubtful whether the basic norm is a full fiction.
For it just establishes the highest authority of positive law; nothing speaks
against it being theoretically validated by a yet higher nonlegal authority.

3.4.7 Relations between Is and Ought

The most relevant new element in Kelsen’s theory on the relations between
Is and Ought has already been discussed: the tight conceptual connection
between the norm and an act of will. Besides, Kelsen still maintains that the
validity both of legal systems and of individual legal norms is conditioned by
their efficacy (Kelsen 1991: 139–40 [Kelsen 1979: 112–13]). The theory of
the “indifferent substrate,” developed in the second phase and somewhat
neglected in the third phase, is elaborated in the final phase. Discussing
authors such as Edmund Husserl, Richard M. Hare, and Jørgen Jørgensen,
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

Kelsen maintains that this substrate is the meaning of an act of thinking that
is different from an assertion or a proposition. It is not modally formed and
must not be understood as being “descriptive” or “indicative.” An example
of a modally indifferent substrate would be a behavior that might actually be
the case (an Is-content) or which ought to be the case (an Ought-content)
(Kelsen 1991: 58-61 [Kelsen 1979: 44-7]): “‘Being-the-content-of-an-
Ought’ is the property of behaviour as modally indifferent substrate (just
as it is the property of a cherry-stone to be contained in a cherry” (Kelsen
1991: 61 [Kelsen 1979: 47]).

63
To be sure, if the act of will is not only a condition of the norm’s existence, but part of the concept
of a norm, then every norm would be a ‘chimera,’ a hybrid of an Ought-element and an Is-
element – something that Kelsen abhorred throughout his earlier writings.
64
This would, however, deviate from Kelsen’s conception which takes reality – in this phase, and
in the context of the basic norm – to be natural reality (Kelsen maintains that the basic norm
contradicts reality because the act of will connected with it is not given in reality).
 72 Philosophy of Law

4 Summary
It is time for a short summary. Legal normativism may be defined as the view
that law is exclusively a system of (normative) rules which are objective and
impersonal, and which cannot be derived from any factuality. It is based on the
dualism of Is and Ought. Hume might be regarded as the progenitor of general
normativism because he pointed to the logical independence of (subjective)
practical sentences from empirical or theoretical sentences. Kant followed suit,
but he took practical sentences to be subjected to reason and, therefore, to be
objective. In addition, he saw a primacy of practical reason over theoretical
reason, and introduced normativity as indispensable for the operations of
understanding, that is, for the cognition of objects. Lotze gave an ontological
touch to this kind of theoretical Ought by holding that there are valid truths from
which an Ought, or rules to think in a certain way, emanate. This interpretation
was adopted by Windelband who took normativity in the form of values to be
a basic philosophical element, and excluded any psychologistic explanation of
cognition; at the same time, Windelband rather neglected the moral or practical
Ought. Rickert systematized Windelband’s approach. Ultimately, however, his
theory foundered on the antagonism between immanent Ought and transcendent
value.
In Hauptprobleme der Staatsrechtslehre (1911), Kelsen presents his theory
as a methodology of legal dogmatics. He distinguishes Ought from Is and
takes law to be completely a matter of Ought: Law is norm, and only those
legal concepts which can be shown to be constructions on the basis of legal
norms are legitimate. Kelsen identifies just one level of legal norms, the
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

Rechtssätze, which are theoretical judgments expressing the constructed will

of the state to act in a certain way under certain conditions. The Ought is
somehow connected with the Rechtssatz, but Kelsen cannot yet determine in
which way it is related.
In the second phase of his writings, between around 1920 and 1935, Kelsen
introduces the theories of the legal hierarchy and the basic norm and explicitly
places himself into the neo-Kantian tradition. He takes his theory to be
a transcendental philosophy of law, that is, a theory of the necessary presup-
positions and elements of legal cognition as embodied in legal dogmatics. The
legal norm, the Rechtssatz, is the cognitive unit of legal science, a theoretical
hypothetical judgment connecting an exercise of coercion with certain condi-
tions by using the category of Ought, which Kelsen dubs “peripheral imput-
ation.” The Rechtssatz is the normative analogue of the causal law which
structures the sphere of Is; and peripheral imputation is the normative ana-
logue of causality. It is a purely theoretical category and has no addressee.
 Hans Kelsen’s Normativism 73

A second kind of theoretical Ought can be found in the concept of validity:

The validity of any logical judgment is given if one ‘ought’ to think in
accordance with it. The practical Ought, contained especially in the concept
of ‘duty,’ which has an addressee, is shunted off to the disposable level of
secondary norms.
The third phase, between around 1940 and 1960, is heralded by the aban-
donment of the idealistic neo-Kantian conception of cognition in favor of
a more or less implicit commonsense realism. Now, Kelsen takes the legal
norm to be “given” to legal cognition and “described” by the judgment of legal
science, the Rechtssatz. He avoids labelling it as an ideal or abstract entity and
tends toward legal realism. The sphere of Ought remains vague, and the
explanation of the dualism of Is and Ought dwindles to the thesis that it is
impossible that something which ought to be follows from something that is,
and vice versa. “Ought” is a portmanteau concept, covering very different
normative functions such as commanding, prohibiting, empowering, etc.
Peripheral imputation steps back as normative operator of the general legal
norm, and empowerment takes its place. All “practical” manifestations of the
Ought in law, that is, normative demands for a certain action, are once more
a matter of reducible secondary norms. Norms are taken to be the meaning-
contents of intentional acts; as they are not assertions, there is no direct logic
of norms. However, Kelsen introduces the dubious notion of an “indirect”
logic of norms – a theory of logical relations between norms which are
somehow a projection from the relations between judgments or statements
describing norms.
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

The writings of the fourth phase, lasting until Kelsen’s death in 1973, can be
seen as an attempt to resolve the inconsistencies of the third phase. They are
characterized by a strict will theory of norms. For the first time in his career,
Kelsen maintains that any Ought or norm is impossible without an act of will as
its bearer. The will even seems to be part of the definition of the Ought. But this
is at odds with Kelsen’s assumption that the Ought is objective. Furthermore,
partly because of his will theory of norms and partly because of his sticking to
the thesis that norms are objects “given” to cognition, Kelsen abandons the
notion that there may be any direct or indirect logical relations between norms.
Normativism, which flourished abundantly in Kelsen’s neo-Kantian period, is
practically gone.
Finally, a very short appraisal. I hope to have shown that the general philo-
sophical normativistic approach can be well argued for. As Kant demonstrated,
in respect of the aim of preserving objectivity of cognition, it is the least
demanding alternative to an untenable position of philosophical realism accord-
ing to which cognition deals with a “world-in-itself.” Furthermore, normativism
 74 Philosophy of Law

is able to thwart any naturalistic attempt to reduce cognition to physiological or

psychical processes. Lotze’s terse argument for this case given in Section 2.3.1
seems to be unanswerable; and it is still applicable in what seems to be
a reincarnation of the debate about psychologism – the current debate about
the role of cognitive sciences.
Kelsen’s version of legal-theoretical normativism, however, is too complex
to admit of a simple assessment. His theory began as a methodology of legal
dogmatics, as embodied in legal practice, and developed into a rational recon-
struction of the factum of legal science in his neo-Kantian phase. In this
program, normativism defines not only the institutionalized perspective of
lawyers, but also the view one has to take if law as a distinct object of cognition
is to be possible at all. This is largely plausible, and maybe the Pure Theory in its
neo-Kantian guise offers the best background for any attempt to reconcile the
normativity, the objectivity, and the positivity of the law. Yet, Kelsen’s norma-
tivism is obscured by his tendency to merge different aspects which should be
kept apart. He does not clearly distinguish between (1) the prescientific ‘prac-
tical’ normativity which is contained in ‘everyday’ legal discourse about rights,
duties, etc.; (2) the ‘theoretical’ Ought (be it the construed will of the state,
imputation, or empowerment) which is contained in the primary legal norm as
a product of cognitive activity and distinguishes law from other phenomena;
and (3) the even less substantial Ought making up the validity of any cognitive
judgment. And he does not succeed in determining the relation between Ought
and willing in an unequivocal way.
But perhaps it is exactly such ambiguities and inconsistencies which, together
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

with its critical attitude, make up the never-ending topicality and fertility of the
Pure Theory.
 References
Beiser, F. C. (2009). Normativity in Neo-Kantianism: Its Rise and Fall.
International Journal of Philosophical Studies (17), 9–27.
Beiser, F. C. (2013). Late German Idealism: Trendelenburg & Lotze. Oxford:
Oxford University Press.
Beiser, F. C. (2014). The Genesis of Neo-Kantianism, 1796–1880. Oxford:
Oxford University Press.
Bennett, K. and McLaughlin, B. (2018). Supervenience. In Stanford
Encyclopedia of Philosophy, https://plato.stanford.edu/entries/superveni
ence/ (accessed October 14, 2020).
Frege, G. (1986). Der Gedanke. In Gottlob Frege, Logische Untersuchungen,
ed. by G. Patzig, 3rd ed. Göttingen: Vandenhoeck & Ruprecht, 30–53.
Habermas, J. (1999). Wahrheit und Rechtfertigung. Frankfurt/Main: Suhrkamp.
Hart, H. L. A. (1961). The Concept of Law. Oxford: Clarendon Press.
Heidemann, C. (1997). Die Norm als Tatsache. Zur Normentheorie Hans
Kelsens. Baden-Baden: Nomos Verlag.
Heidemann, C. (1999). Norms, Facts, and Judgments. A Reply to S. L. Paulson.
Oxford Journal of Legal Studies (19), 345–50.
Heidemann, C. (2000). The Creation of Normative Facts. Law and Philosophy
(19), 263–81.
Heidemann, C. (2007). Noch einmal: Stanley L. Paulson und Kelsens
urteilstheoretischer Normbegriff. Archiv für Rechts- und Sozialphilosophie
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

(93), 345–62.
Heidemann, C. (2020). Das Faktum der Rechtswissenschaft bei Hans Kelsen. In
M. Jestaedt, R. Poscher, and J. Kammerhofer, eds., Die Reine Rechtslehre auf
dem Prüfstand. Hans Kelsen’s Pure Theory of Law: Conceptions and
Misconceptions. Stuttgart: Franz Steiner Verlag, 81–98.
Hoeschen, A. (1999). Das “Dostojewsky”-Projekt. Lukács neukantianisches
Frühwerk in seinem ideengeschichtlichen Kontext. Tübingen: de Gruyter.
Hume, D. (2007a). A Treatise of Human Nature. Oxford: Clarendon Press.
Hume, D. (2007b). An Enquiry Concerning Human Understanding. Oxford:
Oxford University Press.
Husserl, E. (1900). Logische Untersuchungen, Vol. 1. Leipzig: Veit & Comp.
Kant, I. (1992). The Jäsche logic. In I. Kant, Lectures on Logic, trans. and ed. by
J. M. Young. Cambridge: Cambridge University Press, 521–640.
Kant, I. (1998). Critique of Pure Reason, trans. and ed. by P. Guyer and
A. W. Wood. Cambridge: Cambridge University Press.
 76 References

Kant, I. (2002). Critique of the Power of Judgment, trans. and ed. by P. Guyer
and E. Matthews. Cambridge: Cambridge University Press.
Kant, I. (2004). Prolegomena to Any Future Metaphysics, trans. and ed. by
G. Hatfield. Cambridge: Cambridge University Press.
Kant, I. (2015). Critique of Practical Reason, trans. and ed. by M. Gregor,
revised ed. Cambridge: Cambridge University Press.
Kelsen, H. (1911). Hauptprobleme der Staatsrechtslehre. Tübingen: Mohr.
Kelsen, H. (1920). Das Problem der Souveränität und die Theorie des
Völkerrechts. Tübingen: Mohr.
Kelsen, H. (1922a). Rechtswissenschaft und Recht. Erledigung eines Versuchs
zur Überwindung der “Rechtsdogmatik.” ZöR (3), 103–235.
Kelsen, H. (1922b). Der soziologische und der juristische Staatsbegriff.
Tübingen: Mohr.
Kelsen, H. (1923). Hauptprobleme der Staatsrechtslehre, 2nd ed. Tübingen: Mohr.
Kelsen, H. (1925). Allgemeine Staatslehre. Berlin: Springer.
Kelsen, H. (1928a). Die philosophischen Grundlagen der Naturrechtslehre und
des Rechtspositivismus. Berlin: Pan-Verlag.
Kelsen, H. (1928b). Rechtsgeschichte gegen Rechtsphilosophie? Eine
Erwiderung. Vienna: Julius Springer.
Kelsen, H. (1934). Reine Rechtslehre. Einleitung in die rechtswissenschaftliche
Problematik. Leipzig and Vienna: Deuticke.
Kelsen, H. (1939). Die Entstehung des Kausalgesetzes aus dem
Vergeltungsprinzip. The Journal of Unified Science (Erkenntnis) 8, 69–130.
Kelsen, H. (1941). Vergeltung und Kausalität. The Hague: W. P. van Stockum.
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

Kelsen, H. (1957a). The Pure Theory of Law and Analytical Jurisprudence. In

H. Kelsen, What Is Justice? Justice, Law and Politics in the Mirror of
Science. Collected Essays by Hans Kelsen. Berkeley and Los Angeles:
University of California Press, 266–87.
Kelsen, H. (1957b). Why Should the Law Be Obeyed? In H. Kelsen, What Is
Justice? Justice, Law and Politics in the Mirror of Science. Collected Essays
by Hans Kelsen. Berkeley and Los Angeles: University of California Press,
257–65.
Kelsen, H. (1960). Reine Rechtslehre, 2nd ed. Vienna: Deuticke.
Kelsen, H. (1963). Die Grundlagen der Naturrechtslehre. ÖZöR (13), 1–37.
Kelsen, H. (1968a). Das Verhältnis von Staat und Recht im Lichte der
Erkenntniskritik. In H. Klecatsky, R. Marcic, and H. Schambeck, eds., Die
Wiener Rechtstheoretische Schule. Vienna: Europa-Verlag, 77–120.
Kelsen, H. (1968b). Die Rechtswissenschaft als Normwissenschaft oder als
Kulturwissenschaft. In H. Klecatsky, R. Marcic, and H. Schambeck, eds.,
Die Wiener Rechtstheoretische Schule. Vienna: Europa-Verlag, 31–77.
 References 77

Kelsen, H. (1968c). Die Problematik der Reinen Rechtslehre. ÖZöR (18), 143–84.
Kelsen, H. (1968d). Die Funktion der Verfassung. In H. Klecatsky, R. Marcic,
and H. Schambeck, eds., Die Wiener Rechtstheoretische Schule. Vienna:
Europa-Verlag, 1615–22.
Kelsen, H. (1979). Allgemeine Theorie der Normen, ed. by K. Ringhofer and
R. Walter. Viena: Manz.
Kelsen, H. (1991). General Theory of Norms, trans. by M. Hartney. Oxford:
Clarendon Press.
Kelsen, H. (1992). Introduction to the Problems of Legal Theory, trans. and ed.
by B. Litschewski Paulson and S. L. Paulson. Oxford: Clarendon Press.
Kelsen, H. (1998a). Foreword to Hauptprobleme der Staatsrechtslehre, trans. by
S. L. Paulson. In S. L. Paulson and B. Litschewski Paulson, eds., Normativity
and Norms. Oxford: Clarendon Press, 3–22.
Kelsen, H. (1998b) A Letter to Renato Treves, trans. by S. L. Paulson. In
S. L. Paulson and B. Litschewski Paulson, eds., Normativity and Norms.
Oxford: Clarendon Press, 169–76.
Kelsen, H. (2005). The Pure Theory of Law, trans. from the 2nd ed. by
M. Knight. Clark, NJ: The Lawbook Exchange.
Kelsen, H. (2006). Autobiographie (1947). In M. Jestaedt, ed., Hans Kelsen im
Selbstzeugnis. Tübingen: Mohr Siebeck.
Kelsen, H. and Klug, U. (1981). Rechtsnormen und logische Analyse. Ein
Briefwechsel 1959 bis 1965. Vienna: Deuticke.
Köhnke, K. C. (1993). Entstehung und Aufstieg des Neukantianismus.
Frankfurt: Suhrkamp.
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

Kubeš, V. (1980). Das neueste Werk Hans Kelsens über die allgemeine Theorie
der Normen und die Zukunft der Reinen Rechtslehre. ÖZöR (31), 155–99.
Lotze, H. (1841). Metaphysik. Leipzig: Weidmann’sche Buchhandlung.
Lotze, H. (1884). Logics, trans. and ed. by B. Bosanquet. Oxford: Clarendon Press.
McIntyre, A. C. (1959). Hume on “Is” and “Ought.” The Philosophical Review
(68), 451–68.
Merkl, A. J. (1917). Das Recht im Lichte seiner Auslegung. Deutsche
Richterzeitung (9), 162–76.
Nino, C. S. (1998). Some Confusions Surrounding Kelsen’s Concept of
Validity. In S. L. Paulson and B. Litschewski Paulson, eds., Normativity
and Norms. Oxford: Clarendon Press, 253–62.
Opałek, K. (1980). Überlegungen zu Hans Kelsens “Allgemeiner Theorie der
Normen.” Vienna: Manz.
Paulson, S. L. (1988a). Die Rezeption Kelsens in Amerika. In O. Weinberger
and W. Krawietz, eds., Die Reine Rechtslehre im Spiegel ihrer Fortsetzer und
Kritiker. Vienna and New York: Springer, 179–202.
 78 References

Paulson, S. L. (1988b). An Empowerment Theory of Legal Norms. Ratio Juris

(1), 58–72.
Paulson, S. L. (1990). Towards a Periodization of the Pure Theory of Law. In
L. Gianformaggio, ed., Hans Kelsen’s Legal Theory. A Diachronic Point of
View. Turin: Giappichelli editore.
Paulson, S. L. (2000). On the Puzzle Surrounding Hans Kelsen’s Basic Norm.
Ratio Juris (13), 279–93.
Paulson, S. L. (2012). A “Justified Normativity” Thesis in Hans Kelsen’s Pure
Theory of Law? In M. Klatt. ed., Institutionalized Reason. The Jurisprudence
of Robert Alexy. Oxford: Oxford University Press, 61–114.
Paulson, S. L. (2013). Hans Kelsen: Das Ende der Reinen Rechtslehre? Ein
Briefwechsel, ein Spannungsverhältnis und der Umsturz der Rechtslehre
Hans Kelsens. In St. Augsberg and A. Funke, eds., Kölner Juristen im 20.
Jahrhundert. Tübingen: Mohr Siebeck, 53–74.
Paulson, S. L. (2017). Hans Kelsen and Carl Schmitt: Growing Discord,
Culminating in the “Guardian” Controversy of 1931. In J. Meierhenrich
and O. Simons, eds., The Oxford Handbook of Carl Schmitt. Oxford:
Oxford University Press, 510–46.
Raz, J. (1974). Kelsen’s Theory of the Basic Norm. American Journal of
Jurisprudence (19), 94–111.
Rickert, H. (1909). Zwei Wege der Erkenntnistheorie. Kant-Studien (14),
169–228.
Rickert, H. (1921). Der Gegenstand der Erkenntnis, 4th and 5th ed. Tübingen:
Mohr.
Ross, A. (1957). Review of Hans Kelsen, “What Is Justice?”. California Law
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

Review (45), 564–70.

Schmitt, C. (1977). Über die drei Arten des rechtswissenschaftlichen Denkens.
In H.-J. Koch, ed., Die juristische Methode im Staatsrecht. Frankfurt/Main:
Suhrkamp, 366–98.
Thieme, K. (1887). Der Primat der praktischen Vernunft. Leipzig: Ackermann
& Glaser. https://archive.org/details/derprimatderprak00thie/mode/2up
(accessed August 20, 2021).
Walker, R. C. S. (2017). The Primacy of Practical Reason. In M. C. Altman, ed.,
The Palgrave Kant Handbook. London: Springer Nature, 191–209.
Weinberger, O. (1981). Normentheorie als Grundlage der Jurisprudenz und
Ethik. Berlin: Duncker & Humblot.
Willaschek, M. (2009). Rationale Postulate. Über Kants These vom Primat
der praktischen reinen Vernunft. In H. F. Klemme, ed., Kant und die
Zukunft der europäischen Aufklärung. Berlin and New York: de Gruyter,
251–68.
 References 79

Willaschek, M. (2010). Die “Spontaneität des Erkenntnisses.” Über die

Abhängigkeit der “Transzendentalen Analytik” von der Auflösung der
Dritten Antinomie. In J. Chotaš, J. Karásek, and J. Stolzenberg, eds.,
Metaphysik und Kritik. Interpretationen zur “Transzendentalen Dialektik.”
Würzburg: Königshausen und Neumann, 165–84.
Windelband, W. (1874). Review of Christoph Sigwart, Logik. Philosophische
Monatshefte (10), 33–42.
Windelband, W. (1907). Präludien, 3rd ed. Tübingen: Mohr.
Windelband, W. (1912). Prinzipien der Logik. Tübingen: Mohr.
Wittgenstein, L. (2009). Philosophical Investigations. German text with an
English trans. by G. E. M. Anscombe, P. M. S. Hacker, and J. Schulte, revised
4th ed. by P. M. S. Hacker and J. Schulte. Chichester, UK: Wiley-Blackwell.
Zalewska, M. (2020a). The Basic Norm at the Time of the Revolution. In
M. Belov and A. A. i Ninet, eds., Revolution, Transition, Memory, and
Oblivion. Cheltenham, UK: Edward Elgar Publishing.
Zalewska, M. (2020b). Does Hans Kelsen’s Pure Theory of Law Support Rule
of Law and Democracy? In H. Takikawa, ed., The Rule of Law and
Democracy. The 12th Kobe Lecture and the 1st IVR Japan International
Conference. Stuttgart: Franz Steiner, 203–16.
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press
 Elements in the Philosophy of Law

Series Editors

George Pavlakos
University of Glasgow
George Pavlakos is Professor of Law and Philosophy at the School of Law, University of
Glasgow. He has held visiting posts at the universities of Kiel and Luzern, the European
University Institute, the UCLA Law School, the Cornell Law School and the Beihang Law
School in Beijing. He is the author of Our Knowledge of the Law (2007) and more recently has
co-edited Agency, Negligence and Responsibility (2021) and Reasons and Intentions in Law
and Practical Agency (2015).

Gerald J. Postema
University of North Carolina at Chapel Hill
Gerald J. Postema is Professor Emeritus of Philosophy at the University of North Carolina at
Chapel Hill. Among his publications count Utility, Publicity, and Law: Bentham’s Moral and
Legal Philosophy (2019); On the Law of Nature, Reason, and the Common Law: Selected
Jurisprudential Writings of Sir Matthew Hale (2017); Legal Philosophy in the Twentieth
Century: The Common Law World (2011), Bentham and the Common Law Tradition,
2nd edition (2019).

Kenneth M. Ehrenberg
University of Surrey
Kenneth M. Ehrenberg is Reader in Public Law and Legal Theory at the University of Surrey
School of Law and Co-Director of the Surrey Centre for Law and Philosophy. He is the author
of The Functions of Law (2016) and numerous articles on the nature of law, jurisprudential
methodology, the relation of law to morality, practical authority, and the epistemology of
https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press

evidence law.

Associate Editor

Sally Zhu
University of Sheffield
Sally Zhu is a Lecturer in Property Law at University of Sheffield. Her research is on property
and private law aspects of platform and digital economies.

About the Series

This series provides an accessible overview of the philosophy of law, drawing on its
varied intellectual traditions in order to showcase the interdisciplinary dimensions of
jurisprudential enquiry, review the state of the art in the field, and suggest fresh
research agendas for the future.
Focussing on issues rather than traditions or authors, each contribution seeks to deepen
our understanding of the foundations of the law, ultimately with a view to offering
practical insights into some of the major challenges of our age.
 Elements in the Philosophy of Law

Elements in the Series

Hans Kelsen’s Normativism
Carsten Heidemann

A full series listing is available at: www.cambridge.org/EPHL

https://doi.org/10.1017/9781108993661 Published online by Cambridge University Press