Kelsen on Monism and Dualism

Abstract

Kelsen defends (a) monism, that is, the view that international law and the various state legal systems taken together constitute a unified normative system, and (b) the primacy of international law over state law within the monistic framework, arguing (c) that his analysis of the monism/dualism question undermines (what he refers to as) the dogma of sovereignty and removes a difficult obstacle to the efforts to further centralize international law. He argues in support of the (a)-claim (i) that only monism is compatible with the epistemological postulate, according to which cognition requires the unity of the object of cognition, (ii) that we must distinguish between norm contradictions and norm contrarieties, and that the norm conflicts that are said by critics of monism to undermine monism are harmless, since they are not norm contradictions, but only norm contrarieties, and (iii) that the doctrine of dualism collapses into solipsism due to its dependence on the doctrine of recognition. He argues in support of the (b)-claim (iv) that the idea of the primacy of international law within the monistic framework is necessary to account for the existence of states that are legally coordinated and separated from one another in their spheres of validity. And he argues in support of the (c)-claim (v) that there can be no sovereign states that are independent of international law and thus constitute an obstacle to the efforts to further centralize international law. In this article, I discuss claims (a), (b), and (c), and the arguments (i)-(v) that Kelsen adduces in support of these claims. I argue (I) that while Kelsen is right that only monism is compatible with the epistemological postulate, his claim is doubtful that the norm conflicts identified by the critics of monism and put forward as an obstacle to the theory of monism are only norm contrarieties, not norm contradictions. I also argue (II) that if Kelsen had been right that these norm conflicts are only norm contrarieties, the dualists should have been able to invoke the same defense against the critics of dualism and thus avoid the criticism of dualism via the criticism of the doctrine of recognition. Moreover, I argue (III) that Kelsen’s account of norm contradictions and norm contrarieties is problematic in more than one way, and that this affects the clarity and utility of this distinction. Next I argue (IV) that Kelsen’s claim is well founded that dualism collapses into solipsism due to its dependence on the doctrine of recognition, (V) that Kelsen’s defense of monism with international law primacy is plausible, but not without its problems, and (VI) that Kelsen’s claim is plausible that there can be no sovereign states that are independent of international law. Finally, I suggest (VII) that, contrary to what Kelsen appears to think, the epistemological postulate appears to rule out locating law in a higher realm of norms and values sharply separated from the world of time and space.

Full text

1Kelsen on Monism and Dualism
KELSEN ON MONISM AND DUALISM
Torben Spaak
Abstract: Kelsen defends (a) monism, that is, the view that international law and the
various state legal systems taken together constitute a unified normative system, and
(b) the primacy of international law over state law within the monistic framework. He
argues in support of the (a)-claim (i) that only monism is compatible with the
epistemological postulate, according to which cognition requires the unity of the object
of cognition, (ii) that the norm conflicts that are said by the critics of monism to
undermine monism are harmless because they are not norm contradictions, but only
norm contrarieties, and (iii) that the doctrine of dualism collapses into solipsism due
to its dependence on the doctrine of recognition. And he argues in support of the
(b)-claim (iv) that the idea of the primacy of international law within the monistic
framework is necessary to account for the existence of states that are legally coordinated
and separated from one another in their spheres of validity.
In this article, I discuss critically claims (a) and (b) and the arguments (i)-(iv) that
Kelsen adduces in support of these claims. I argue, more specifically, (I) that while
Kelsen is right that only monism is compatible with the epistemological postulate,
the claim is unpersuasive that norm contrarieties, as distinguished from norm
contradictions, are harmless from the point of view of this postulate, and (II) that
in any case not all the norm conflicts identified by the critics of monism are norm
contrarieties. I also argue (III) that if the above-mentioned claim (ii) had been
persuasive, the dualists, too, might have invoked it and avoided Kelsen’s criticism of
dualism via the criticism of the doctrine of recognition, and (IV) that Kelsen’s claim
is well founded that dualism collapses into solipsism because of the collapse of the
doctrine of recognition. Finally, I suggest (V) that contrary to what Kelsen appears
to think, the epistemological postulate rules out locating law in a higher realm of
norms and values that is sharply separated from the world of time and space.
1. Introduction
Kelsen defends (a) monism, that is, the view that international law and the various
state legal systems taken together constitute a unified normative system, and (b) the
primacy of international law over state law within the monistic framework, arguing
(c) that his analysis of the monism/dualism question undermines (what he refers to
as) the dogma of sovereignty and removes a difficult obstacle to the efforts to further
centralize international law. He argues in support of the (a)-claim (i) that only monism

is compatible with the epistemological postulate, according to which cognition requires
the unity of the object of cognition, (ii) that we must distinguish between norm
contradictions and norm contrarieties, and that the norm conflicts that are said by
critics of monism to undermine monism are harmless, since they are not norm
contradictions, but only norm contrarieties, and (iii) that the doctrine of dualism
collapses into solipsism due to its dependence on the doctrine of recognition. He
argues in support of the (b)-claim (iv) that the idea of the primacy of international
law within the monistic framework is necessary to account for the existence of states
that are legally coordinated and separated from one another in their spheres of validity.
And he argues in support of the (c)-claim (v) that there can be no sovereign states
that are independent of international law and thus constitute an obstacle to the efforts
to further centralize international law.
In this article, I discuss claims (a), (b), and (c), and the arguments (i)-(v) that
Kelsen adduces in support of these claims. I argue (I) that while Kelsen is right that
only monism is compatible with the epistemological postulate, his claim is doubtful
that the norm conflicts identified by the critics of monism and put forward as an
obstacle to the theory of monism are only norm contrarieties, not norm
contradictions. I also argue (II) that if Kelsen had been right that these norm
conflicts are only norm contrarieties, the dualists should have been able to invoke
the same defense against the critics of dualism and thus avoid the criticism of
dualism via the criticism of the doctrine of recognition. Moreover, I argue (III) that
Kelsen’s account of norm contradictions and norm contrarieties is problematic in
more than one way, and that this affects the clarity and utility of this distinction.
Next I argue (IV) that Kelsen’s claim is well founded that dualism collapses into
solipsism due to its dependence on the doctrine of recognition, (V) that Kelsen’s
defense of monism with international law primacy is plausible, but not without its
problems, and (VI) that Kelsen’s claim is plausible that there can be no sovereign
states that are independent of international law. Finally, I suggest (VII) that, contrary
to what Kelsen appears to think, the epistemological postulate appears to rule out
locating law in a higher realm of norms and values sharply separated from the world
of time and space.
I base my account and criticism of Kelsen’s discussion of monism and dualism
on the lucid presentation of these questions in the first edition of Reine Rechtslehre,
though I shall now and then refer also to the account in the General Theory of Law
and State. As far as I can tell, the discussions of monism and dualism in these two
works are very similar.
I begin by introducing the doctrines of monism and dualism and discussing
briefly a recent court case that illustrates them (Section 2). I then introduce the
epistemological postulate of the unity of the object of cognition and the distinction
between norm contradictions and norm contrarieties (Section 3) and consider
Kelsen’s claim that dualism collapses into solipsism due to its dependence on the
doctrine of recognition (Section 4). Having done that, I turn to consider Kelsen’s
argument in support of the idea of the primacy of international law within the
Torben Spaak2

monistic framework (Section 5) and add a few words about the political implications
of Kelsen’s analysis (Section 6). I then turn to a critical discussion of the distinction
between norm contradictions and norm contrarieties and the use to which Kelsen
puts it in his defense of monism (Section 7). The article concludes with some
thoughts on the relation between the epistemological postulate and Kelsen’s
ontological non-naturalism (Section 8).
2. Monism and Dualism
Kelsen explains that monism is the theory that all law – international law as
well as the various state legal systems – constitutes a unified system of law. The
idea, he points out, is that “one can conceive of international law together with the
state legal systems as a unified system of norms in exactly the same way as one is
accustomed to regarding the state legal system as a unity.”
1
Dualism, on the other
hand, is the theory that international law and state law do not constitute a unified
system of law, but exist instead independently of one another. As Kelsen puts it,
dualism sees “international law and state law as two different systems of norms,
independent of each other and reciprocally isolated because resting on two different
basic norms.”
2
The understanding of the ideas of monism and dualism on the part
of contemporary experts in international law, such as Antonio Cassese,
3
is very
much in keeping with that of Kelsen.
Moreover, it is common to distinguish within the framework of monism
between (a) monism with national law supremacy and (b) monism with international
law supremacy.
4
Whereas (a) takes its starting point in a national legal system, say,
Swedish law, and conceives of international law together with the other national
legal systems as part of this national legal system, (b) takes its starting point in
international law and conceives of the various national legal systems as subordinated
to, and in this sense as parts of, international law. As we shall see in Section 5, Kelsen
defends alternative (b).
There are at least two important legal consequences of the choice between
monism and dualism. First, under dualism, but not under monism, a state will not
be bound by international law, unless it has recognized it. Secondly, a state that
accepts dualism, but not a state that accepts monism, will have to transform the
norms of international law into state law by means of state legislation – if no such
transformation has taken place, the national courts will not be (legally) able to give
effect to the relevant norms of international law.
Kelsen on Monism and Dualism 3
1 H. Kelsen, Introduction to the Problems of Legal Theory, 1992, 111.
2 H. Kelsen, IPLT, 111.
3 See A. Cassese, International Law, 2nd. ed. 2005, 213–6.
4 A. Cassese, International law, 213–6.

Cassese points out that whereas the main political idea behind dualism is the
idea of state sovereignty, the main political idea behind monism with national law
primacy is a moderate nationalism, and behind monism with international law
primacy an ideal of internationalism and pacifism.
5
And even though his reasoning
centers on the above-mentioned epistemological postulate, it is clear that Kelsen is
drawn to the ideal of internationalism and pacifism. Indeed, as we shall see (in
Section 6), his criticism of dualism aims to show that (what he refers to as) the
dogma of state sovereignty does not pose an obstacle to efforts to further centralize
international law.
It appears, however, that contemporary legal practice does not conform clearly
to either the monist or the dualist view. As Cassese explains, although dualism has
been dominating for some time, it is no longer valid in its entirety, and some of
Kelsen’s ideas are coming to be accepted in the international community:
It [international law] no longer constitutes a different legal realm from the
various municipal systems, but has a huge daily direct impact on these systems.
It conditions their life in many areas and even contributes to shaping their
internal functioning and operation. In addition, many international rules address
themselves directly to individuals, without the intermediary of national legal
systems: they impose obligations on them (this chiefly applies to rules on
international crimes), or grant them rights (for instance the right to petition
international bodies). Those obligations must be fulfilled, and the rights may
be exercised, regardless of what national legal orders may provide. In short, in
many respects individuals have become international legal subjects, associated
to sovereign States.
6
The much-discussed decision by the European Court of Justice, the Kadi and
Al Barakaat case, is of some interest in this context. Whereas the Court of First
Instance of the European Communities appears to have adopted a monistic view,
7
the Court of Justice clearly adopted a dualist view of the relation between
international law, EU law, and the law of the member states of the European
Union.
8
Crudely put, whereas the Court of First Instance argued that central
principles of international law take priority over EU law and obligate the institutions
of the EU to act accordingly, the Court of Justice insisted that each EU institution
is obligated under EU law to check and see whether the action it intends to take
under international law is permissible in light of fundamental principles of EU law.
Torben Spaak4
5 A. Cassese, International Law, 214–6.
6 A. Cassese, International law, 216–7.
7 Case T-306/01, Ahmed Ali Yusuf and Al Barakaat International Foundation v. Council and Commission,
[2005] ECR II-3533, Case T-315/01, Yassin Abdullah Kadi v. Council and Commission, [2005] ECR II-
3649.
8 Joined cases C-402/05 and C-415/05, Yassin Abdullah Kadi and Al Barakaat Foundation v Council
of the European Union and Commission of the European Communities.

The central questions in the Kadi and Al Barakaat case(s) were whether the
European Council had the legal power (or competence) to enact a regulation that
provided for the freezing of funds of people or organizations that were suspected
of financing the activities of terror organizations, such as Al Qaeda, and, if so,
whether the enactment of this regulation violated the human rights of the people
involved.
The facts of the Kadi and Al Barakaat case(s) were as follows. On 15 October
1999, the Security Council adopted Resolution 1267 (1999), in which it condemned
the fact that Afghan territory was being used by terror organizations, such as Al
Qaeda, and demanded that the Taliban should without further delay hand over
Usama bin Laden to the appropriate authorities in a country where he could be
brought to justice. In paragraph 6 of the resolution, the Security Council established
an entity named the Sanctions Committee, whose responsibility it was to ensure
that the states implemented certain measures outlined in paragraph 4 of the
resolution. On 19 December 2000 the Security Council adopted Resolution 1333
(2000), demanding, among other things, “that the Taliban should comply with
Resolution 1267 (1999), and that they should cease to provide sanctuary and training
for international terrorists and their organizations and turn Osama bin Laden over
to appropriate authorities to be brought to justice.”
9
Paragraph 8(c) of Resolution
1333 provided that the states should freeze funds and other financial assets of
Usama bin Laden and people and organizations associated with him.
10
And in the
same paragraph the Security Council instructed the Sanctions Committee to
“maintain an updated list . . . of the individuals and entities designated as associated
with Usama bin Laden, including those in the Al-Qaeda organization.”
11
On 26 February 2001, the European Council, which took the view that action
by the European Community was necessary to implement Resolution 1333, adopted
Common Position 2001/154/CFSP concerning additional restrictive measures
against the Taliban, thus amending the previously adopted Common Position
96/746/CFSP. Article 4 of Common Position 2001/154/CFSP provided that
funds and other financial assets of Usama bin Laden and individuals or
organizations associated with him should be frozen and not made available to bin
Laden or the above-mentioned individuals or organizations.
12
On 6 March 2001,
the European Council adopted Regulation (EC) No 467/2001, which prohibited
the “export of certain goods and services to Afghanistan,” and repealed Regulation
No 337/2000. Article 2 of Regulation No 467/2001 provided, inter alia, as follows:
All funds and other financial resources belonging to any natural or legal person,
entity or body designated by the . . . Sanctions Committee and listed in Annex
I shall be frozen.
Kelsen on Monism and Dualism 5
9 Paragraph 14.
10 Paragraph 14.
11 Paragraph 15.
12 Paragraph 17.

No funds or other financial resources shall be made available, directly or
indirectly, to, or for the benefit of, persons, entities or bodies designated by the
Taliban Sanctions Committee and listed in Annex I.
Mr. Kadi and Al Barakaat, an organization based in Sweden, took the case to
the Court of First Instance of the European Community, seeking the annulment
of Regulation No 467/2001. In addition, Mr. Kadi also sought annulment of
Regulation No 2062/2001, and Al Barakaat sought annulment of Regulation No
2199/2001. I shall consider the Al Barakaat case only in what follows.
The Court of First Instance began its treatment of the question of the relation
between international law, EU law, and the legal systems of the member states of
the European Union with a consideration of the relation between, on the one hand,
international law, and, on the other hand, treaty law, including the EC Treaty, and
the legal systems of the member states of the United Nations. The Court took the
view that norms of the UN Charter take priority both over the norms of the legal
systems of the member states and over the norms of the EC Treaty and the
European Charter of Human Rights:
As regards, first, the relationship between the Charter of the United Nations and
the domestic law of the Member States of the United Nations, that rule of
primacy is derived from the principles of customary international law. Under
Article 27 of the Vienna Convention on the Law of Treaties, which consolidates
those principles (and Article 5 of which provides that it is to apply to ‘any treaty
which is the constituent instrument of an international organization and to any
treaty adopted within an international organisation’), a party may not invoke the
provisions of its internal law as justification for its failure to perform a treaty.
As regards, second, the relationship between the Charter of the United Nations
and international treaty law, that rule of primacy is expressly laid down in Article
103 of the Charter which provides that, ‘[i]n the event of a conflict between
the obligations of the Members of the United Nations under the present
Charter and their obligations under any other international agreement, their
obligations under the present Charter shall prevail.’
13
The Court concluded that not only the member states of the European
Community, but also the European Community itself, are bound by international
law, including the provisions of the UN Charter: “it must be held, first, that the
Community may not infringe the obligations imposed on its Member States by the
Charter of the United Nations or impede their performance and, second, that in
the exercise of its powers it is bound by the very Treaty by which it was established,
to adopt all the measures necessary to enable its Member States to fulfil those
obligations.”
14
This means that the Court arrived at the conclusion that the Council
Torben Spaak6
13 Al Barakaat, paragraphs 232–233.
14 Al Barakaat, paragraph 254.

could not review the regulation that it intended to enact (in order to implement the
resolution adopted by the Sanctions Committee) in light of fundamental norms of
EU law without violating its legal duty under the UN Charter. Hence, the Court
concluded, the Council acted in accordance with the law when it enacted the
regulation.
15
Al Barakaat (and Mr. Kadi) appealed the decision to the Court of Justice,
seeking annulment of the regulation in question. The Court found for the appellants
in so far as it set aside the judgments of the Court of First Instance and annulled
Council Regulation (EC) No 881/2002 in so far as it concerned Al Barakaat (and
Mr. Kadi).
The Court of Justice took the view that the Council was under a duty of EU
law to check whether the proposed regulation was in keeping with the fundamental
norms of EU law, and that the Court itself was duty-bound to review any proposed
act of the various institutions of the European Union in light of the fundamental
norms of EU law.
The Court began its analysis of the question under consideration by pointing
out that the European Community is based on the rule of law to the extent that
both its member states and its institutions must review the conformity of their acts
with the EC Treaty,
16
and that no international agreement can affect the autonomy
of the Community legal system.
17
The Court was, however, careful to point out
that a finding on the part of the Court that a Community act based on, say, a
resolution adopted by the Security Council would not entail any challenge to the
primacy of that resolution in international law.
18
It also notes later in the opinion
that “[t]he question of the Court’s jurisdiction arises in the context of the internal
and autonomous legal order of the community, within whose ambit the contested
regulation falls and in which the Court has jurisdiction to review the validity of
Community measures in the light of fundamental rights.
19
The repeated references
to community law as an autonomous legal order suggest that the Court here
adopted a dualist approach.
It is worth noting that neither court speaks of ‘monism’ or ‘dualism.’ What
they do is develop their respective view of the relation between international law,
EU law, and the law of the member states of the European Union, and they clearly
believe that one way of understanding this relation (or these relations) is legally
correct. So we may infer that, even though they do not speak of ‘monism’ or
Kelsen on Monism and Dualism 7
15 I find the Court’s strategy of citing the UN Charter in order to prove that the UN Charter takes
precedence over national law and international treaty law unconvincing, as it presupposes an affirmative
answer to the very question at issue, namely whether the UN Charter does take precedence. I shall
not, however, pursue this objection to the Court’s reasoning.
16 Kadi and Al Barakaat, paragraph 281.
17 Kadi and Al Barakaat, paragraph 282.
18 Kadi and Al Barakaat, paragraph 288.
19 Kadi and Al Barakaat, paragraph 317.

‘dualism’, they view these doctrines as first-order doctrines. That is to say, they think
of the doctrines as being part of the law and therefore as having normative content.
Kelsen, on the other hand, as we shall see, appears to view the concepts of monism
and dualism as second-order concepts, that is, concepts that legal scholars make
use of when analyzing the law, not first-order legal concepts that judges and others
use when stating the law, that is, when making first-order legal claims.
20
Note that
the distinction between first-order and second-order legal concepts is a corollary
to the well-established distinction between object-language and meta-language,
21
and that in the case of the law, the object language is normative, whereas the meta-
language is descriptive.
This difference between first-order and second-order legal concepts means that
legal scholars do not have to worry about the precise content given to a second-
order concept by a court, if the court were to use such a concept. For example, a
legal scholar may operate with a concept of a legal right, such that a person has a
legal right if, and only if, he has a legally respected choice in the sense explained by
H. L. A. Hart.
22
He can do this whether or not there are to be found in the case law
of the high court statements about the structure and import of the concept of a
legal right. He can take such statements lightly, should there be any, because his task
is to analyze the concept of a legal right in such a way that the analysis comports
with legal practice, not with competing theoretical views (even those expressed by
judges) about the structure and import of this concept.
3. The Unity of Cognition, Norm Contradictions,
and Norm Contrarieties
The reason why Kelsen espouses monism is that he accepts the epistemological
postulate of the unity of the object of cognition and believes that monism, but not
dualism, is acceptable from the point of view of this postulate.
23
The idea is that
cognition represents its object as a unity, and that this implies that there are no
contradictions or other inconsistencies in this representation. Kelsen reasons that
it follows from this postulate, when applied to the field of law, that one cannot
coherently maintain that two norms, whose content is mutually exclusive, are
simultaneously valid. Here is Kelsen:
Torben Spaak8
20 For an illuminating discussion of this (or a very similar) distinction, which the author refers to as
a distinction between J-Concepts (juridical concepts) and L-Concepts (law-concepts), see Å. Frändberg,
An Essay on Legal Concept Formation. In J. Hage and D. van der Pfordten, eds., Concepts in Law, 2009,
1–16.
21 On the distinction between object language and meta-language, see J. Lyons, Semantics 1, 1977,
10-3.
22 H. L. A. Hart, Bentham on Legal Rights. In A. W. B. Simpson, ed., Oxford Essays in Jurisprudence
(second series), 1973, 171-201.
23 H. Kelsen, IPLT, 113.

The negative criterion of this unity is non-contradiction, a logical principle that
also applies to cognition in the realm of norms. One cannot claim that a norm
with content A is valid, and claim at the same time that a norm with content
not-A is valid. One can claim—and no doubt must claim, in view of the facts—
that norms whose content is mutually exclusive actually are issued and, on the
part of the norm-addressees, are imagined and are obeyed or not obeyed. There
is no more logical contradiction in this statement, referring to natural facts, than
there is in the observation that two opposing forces are both efficacious. But
one cannot claim that two norms whose content is, logically speaking, mutually
exclusive are valid at the same time—that is, one cannot claim that A ought to
be and, at the same time, not-A ought to be, just as one cannot claim that A is
and, at the same time, not-A is.
24
Kelsen appears to view the epistemological postulate as more or less obvious
and unproblematic. And I am inclined to share this view. Like Kelsen, I find it
reasonable to assume that knowledge involves (or implies) unity of the knowledge
object at least in the minimal sense that this object can be described in a non-
contradictory manner, that one cannot describe in a non-contradictory manner a
situation in which A ought to be and, at the same time, not-A ought to be, and that
therefore one cannot have knowledge of what the persons involved ought to do in
this type of situation.
25
Kelsen notes, however, that critics of monism have objected that there exist
norms in state law and international law that conflict, and that this means that one
cannot coherently conceive of international law and state law taken together as a
unified normative system, as monists wish to do.
26
But, he points out, this objection
fails because these norm conflicts do not qualify as norm contradictions, but are
only norm contrarieties, that is, conflicts between norms on a higher and norms on a
lower level in the (unified) legal system, in the sense that the lower-level norm, N
2
,
although it was created on the basis of the higher-level norm, N
1
, has a content
that does not conform to the requirements in N
1
. As Kelsen puts it, “[t]he problem
posed here is that of ‘norm contrary to norm,’ namely, the unconstitutional statute,
the illegal regulation (the regulation that is contrary to statute), the judicial decision
or the administrative act that is contrary to statute and regulation.”
27
He adds that
norm contrarieties “simply represent[] either the invalidatability of the lower-level
norm, or the responsible authority’s liability to punishment.”
28
Kelsen on Monism and Dualism 9
24 H. Kelsen, IPLT, 112.
25 I assume here that knowledge implies truth: If you know that p, then it is true that p. But if the
description of p is contradictory, it cannot be true that p. Hence in such a case, you do not know that p.
26 H. Kelsen, IPLT, 114.
27 H. Kelsen, IPLT, 71.
28 H. Kelsen, IPLT, 118.

The idea on the part of Kelsen appears to be that positive law implicitly accepts
norm contrarieties, but not norm contradictions, because it does not declare them
to be null and void from the outset (ex tunc), but instead provides that the organ
that enacted an unconstitutional statute or an illegal regulation shall be punished,
or that a court may on appeal invalidate the statute or regulation in question (ex
nunc). Contemplating the case of the unconstitutional statute, he states the following:
If . . . an unconstitutional statute is possible . . . this can only be interpreted in
one way: the constitution aims not only for the validity of the constitutional
statute, but also (in some sense) for the validity of the ‘unconstitutional’ statute.
Otherwise, one could not speak of the ‘validity’ of the latter at all. That the
constitution does aim for the validity of the so-called unconstitutional statute
is shown in the fact that it prescribes not only that statutes should be created
in a certain way and have (or not have) a certain content, but also that if a statute
was created other than in the prescribed way or has other than the prescribed
content, it is not to be regarded as null and void, but is to be valid until it is
invalidated by the designated authority—say, a constitutional court—in a
procedure governed by the constitution.
29
Turning to consider the analogous case of the illegal regulation, he expresses the
same idea in somewhat different terms. The reason why there is no contradiction between
the illegal regulation and the higher-level norm that authorized the creation of the
illegal regulation, he explains, is that the higher-level norm includes alternative
provisions, and that the illegal regulation does not contradict the compound higher-
level norm, but only one of its alternative provisions. As he puts it, “. . . a contradiction
with the first of the alternative provisions into which the compound higher-level norm
divides is not a contradiction with the compound norm as a whole.”
30
Kelsen concludes his discussion of norm contrarieties by pointing out that
such norm conflicts do not threaten the cognitive unity of the system: “Normative
cognition tolerates no contradiction between two norms of the same system; the
possible conflict, however, between two valid norms at different levels is resolved
by the law itself. The unity in the hierarchical structure of the legal system is not
endangered by logical contradiction.”
31
4. Dualism and Solipsism
Since Kelsen espouses monism on the grounds that he believes that only
monism is compatible with the epistemological postulate, he must reasonably believe
that the main problem with dualism is precisely that it is not compatible with this
Torben Spaak10
29 H. Kelsen, IPLT, 72.
30 H. Kelsen, IPLT, 74.
31 H. Kelsen, IPLT, 75.

postulate. He notes, however, that dualists, too, accept the postulate,
32
if only tacitly,
and think they can solve the problem of norm conflicts by regarding their own
state legal system as the only system of legally valid norms.
33
He also notes that since
they realize that in practice it is more or less impossible not to view the norms of
international law and other state legal systems as valid (in the sense of binding),
dualists have introduced the doctrine of recognition in order to account for the existence
of other valid legal systems.
34
Under this doctrine, a legal system, B, qualifies as a
valid legal system from the point of view of another legal system, A, if, and only
if, A has recognized B. Hence B might exist from the point of view of A (because
A has recognized B), but not from the point of view of C (because C has not
recognized B).
But how can the doctrine of recognition be of any help to somebody who is
worried about norm conflicts in light of the epistemological postulate? The idea
appears to be that recognition by system A of system B will render the existence
of any conflicts between the norms of A and B unproblematic from the point of
view of A, on the grounds that A will then have accepted the existence of such norm
conflicts. This idea is similar to the one hinted at by Kelsen in his discussion of
norm contrarieties (see Section 3 above), namely that positive law has accepted the
existence of unconstitutional statutes and illegal regulations by not declaring them
to be null and void from the outset, but instead providing that the organ that enacted
an unconstitutional statute or an illegal regulation shall be punished, or that a court
may on appeal invalidate the statute or regulation in question, and that it has in this
way solved the norm conflict. But this assumes, of course, that the dualists are faced
with norm contrarieties, not with norm contradictions, and it is by no means clear
that this is so. Generally speaking, one cannot dissolve a contradiction by accepting
it. If one cannot coherently maintain that one or more persons ought simultaneously
to do A and not-A, one cannot coherently accept that they ought simultaneously to
do A and not-A either.
Kelsen also objects that the doctrine of dualism will ultimately collapse into
solipsism, that is, the view (crudely put) that the individual (in this case, the legal
system) is at the center of the world and the reason for the existence of every other
thing (in this case, every other legal system).
35
The problem with legal solipsism, he
explains, is that the legal solipsists (that is, the dualists) cannot explain how either
independent states or international law can exist independently of recognition. He
thus assumes that there is an objective legal reality, which features both independent
state legal systems and an independent system of international law and argues that
dualists cannot account for this reality:
Kelsen on Monism and Dualism 11
32 H. Kelsen, IPLT, 111.
33 H. Kelsen, IPLT, 114.
34 H. Kelsen, IPLT, 115.
35 I am relying here on Kelsen’s characterization of solipsism.

The subjectivist’s own ‘I’ is his point of departure in comprehending the world,
but despite extending it to the universe, the subjectivist cannot get past his own
sovereign ‘I’ to arrive at an objective world. He is incapable of comprehending
another subject, the not-‘I’, as an entity comparable to his own ‘I’, with the
same claim to sovereignty, the ‘you’ also claiming to be an ‘I’. Similarly, a
monistic construction based on the primacy of the legal system of one’s own
state is completely incompatible with the notion of a plurality of coordinate
states, equally ordered and legally separated from each other in their spheres
of validity. (And it is precisely this monistic construction that dualism, with its
bias for preserving the dogma of sovereignty, transforms itself into by way of
the doctrine of recognition.) Thus, the primacy of the state legal system implies
in the end not only the denial of the sovereignty of all other states, and thereby
their legal existence as states (in terms of the dogma of sovereignty), but also
the denial of international law.
36
The denial of international law, Kelsen explains, means that from the point of
view of state law, international law cannot perform its function of coordinating the
various states, “equally ordered and legally separated from each other in their spheres
of validity.” For if the validity of international law depends on the validity of state
law, specifically on the validity of the state constitution, and if the constitution can
be abrogated by a simple constitutional amendment, then this will undermine the
state’s recognition of other states, depending as it does on the validity of
international law. As Kelsen puts it, “[t]he theory of the primacy of the legal system
of one’s own state returns ultimately to its point of departure, namely, that one
acknowledge as law only the legal system of one’s own state.”
37
I find Kelsen’s critique of dualism persuasive. To assume, as Kelsen does, that
there is an objective legal reality that features independent state legal systems and
an independent system of international law is plausible, and to object that the legal
dualists cannot account for this reality is convincing. Like Kelsen, I thus believe that
dualism, due to its dependence on the doctrine of recognition, leads to absurd
results on the intellectual plane. This is not, of course, to say that dualism is
unworkable in legal practice.
5. Monism: the Primacy of International Law
Kelsen reasons that one who accepts the idea of the unity of cognition and
who therefore accepts monism will have to conceive of the relation between two
legal systems, A and B, either (i) as a matter of subordination, or (ii) as a matter of
coordination, in a way that presupposes the existence of a third system, C, to which
both A and B are subordinated:
Torben Spaak12
36 H. Kelsen, IPLT, 116.
37 H. Kelsen, IPLT, 117.

There are, in principle, two possibilities for a unified system. Two complexes
of norms, apparently different, form a unified system if one complex (or
system) proves to be subordinate to the other because the basis of its validity—
and so, its (relative) basic norm, the basic determinant of its creation—is found
in the other system, that is, in a norm of the other system. Or, alternatively, the
two systems appear as coordinates of one another, equally ordered, which really
means that they are separated from each other in their spheres of validity.
Coordinate systems presuppose a third, higher-order system that governs the
creation of both of the other systems, separating them from each other in their
spheres of validity, and thus, first and foremost, coordinating them.
38
He then maintains that one who wants to conceive of international law together
with the various state legal systems as a unified legal system will have to opt for the
primacy of international law, since this is the only way of accounting for the
existence of state legal systems that have the legal competence to enter into treaties
with one another, or to create customary law by acting in a certain way.
39
He argues,
more specifically, that it must be a norm of international law that confers such
competence on the states because it could not be a norm of state law, and it could
not be a norm of state law because such a norm could not confer competence on
another state:
If states are coordinates of one another, equally ordered, then each state can
impose obligations on, and grant rights to, only its own citizens. The competence
of a state does not reach beyond the sphere of validity of the state legal system.
And since the competence of one state and that of another cannot be added up
like mathematical quantities, then—unless a higher-order system delegates
appropriate powers—not even two states together are in a position to create norms
that, like norms created by state treaty, are valid for the territory of both states.
40
The idea appears to be that while state A can confer legal competence on the
organs and citizens of A and while state B can confer such competence on the
organs and citizens of B, A cannot confer such competence on B (or on the organs
and citizens of B) and B cannot confer such competence on A (or on the organs
and citizens of A). As Kelsen puts it in the quotation above, “[t]he competence of
a state does not reach beyond the sphere of validity of the state legal system.” What
is needed, then, is a norm that is independent of A and B and confers the necessary
competence on both A and B, and this would have to be a norm of international
law. What else could it be? My guess is that Kelsen has in mind here the principle
pacta sunt servanda, which he considers to be a fundamental principle of international
law.
41
On this analysis, then, the existence and legal competence of states presuppose
Kelsen on Monism and Dualism 13
38 H. Kelsen, IPLT, 112-3.
39 H. Kelsen, IPLT, 122-3.
40 H. Kelsen, IPLT, 122-3.
41 H. Kelsen, IPLT, 107.

the existence of international law, including the principle pacta sunt servanda, that
recognizes them and confers on them this competence.
This line of argument is not without its problems, however. First, if the
existence and legal competence of states presuppose the existence of international
law, how could international law have been created in the first place? But perhaps
Kelsen could argue that we must distinguish between a situation in which a legal
system is emerging and a situation in which it is already in existence, and that his
claim about the dependence of the states on international law concerns the latter
situation only. He might reason, in other words, that although he cannot explain
precisely how either international law or the state legal systems came into existence,
they do exist today, and that his claim is true in this situation. Doing this, he might
add, is no more (and no less) problematic than it is to maintain, as legal positivists
do, that a norm is a legal norm if, and only if, it can be traced back to a recognized
source of law, while admitting that a recognized source of law in turn is a source
of law that is recognized by (the courts of) a legal system. The positivists simply
assume that we can tell whether there is a legal system, LS, in existence in a certain
territory and proceed to argue that any proposed norm will be a norm of LS if,
and only if, it can be traced back to a source of law that is recognized by LS.
Secondly, why does Kelsen not consider the possibility that A can confer legal
competence on A and that B can confer such competence on B? Does he mean to
say that this alternative is ruled out on logical grounds, so that an entity simply
cannot confer competence on itself,
42
or does he mean to say that even if this is
possible, it is not enough? My impression is that he is primarily concerned with the
second alternative, arguing as he does (in the quotation above) that “not even two
states together are in a position to create norms that . . . are valid for the territory
of both states.”
43
I suppose he reasons that the very idea of states and state
competence will be meaningful only against a background of international law, since
only international law can recognize states and confer upon them competence to
change legal positions in the system of international law. And I am inclined to agree.
I conclude that Kelsen’s defense of the primacy of international law within the
framework of monism is plausible, though not without its problems. The main
problem, as I see it, is that one might object that Kelsen is simply assuming from
the outset that a (purported) state would not be a state and would not have legal
competence, if it had not been recognized by international law. But would there
really not be any states or any state legal competence if the system of international
Torben Spaak14
42 This would seem to be Alf Ross’s view. See A. Ross. On Self-Reference and a Puzzle in
Constitutional Law. Mind 78. 1969, 1–24.
43 As far as I can tell, he does not consider the first alternative. But he could have reasoned that
whereas the competence of citizens and organizations would have to be a competence to change
legal positions within the framework of the legal system of A (or B), the competence of A (or B) to
enter into a treaty with another state would have to be a competence to change legal positions within
the framework of the system of international law.

were to be dismantled? It seems to me that this question raises new questions, such
as “What is a state?” and “What is state legal competence?”.
6. State Sovereignty and the Centralization of International Law
Kelsen concludes his discussion of monism and dualism by pointing out that
he has undermined (what he refers to as) the dogma of sovereignty,
44
and that in
doing this he has removed a difficult obstacle to the efforts to develop international
law in the direction of further centralization:
The Pure Theory of Law opposes a line of argumentation that pronounces
such development incompatible with the nature of international law or with
the essence of the state—which is really to say, incompatible with all that the
concept of sovereignty is supposed to express. The Pure Theory of Law
exposes once and for all the attempt to use the concept of sovereignty to lend
to a purely political argument—which is always vulnerable to a comparable
counter-argument—the appearance of a logical argument, which would by its
very nature be irrefutable.
45
Kelsen’s idea, then, is that the discussion makes it clear that those who object
to further centralization of international law, on the grounds that such centralization
would be incompatible with the nature of international law and the idea of
sovereign states, are guilty of begging the question of whether there are, or can be,
any sovereign states that are independent of international law and thus constitute
obstacles to the further centralization of international law. As we have seen in
Section 5, Kelsen answers this question in the negative, arguing that there are no,
and cannot be any, sovereign states that are independent of international law,
because it was international law, and international law only, that recognized them as
states in the first place. So, on this analysis, any state that is recognized by
international law will only be as sovereign as international law allows it to be. He
therefore concludes that the objection under consideration to the efforts to further
centralize international law has been undermined. And, as I have argued in Section
5, I am inclined to agree that the existence and legal competence of the states do
depend on the existence of a system of international law, and that therefore Kelsen’s
analysis has indeed undermined the objection in question.
7. Norm Contradictions and Norm Contrarieties Revisited
Kelsen maintains, as we have seen, that norm contradictions are one thing and
norm contrarieties another, that norm contradictions, but not norm contrarieties,
violate the epistemological postulate, that the norm conflicts that the critics of the
doctrine of monism have identified and presented as obstacles to monism are not
norm contradictions, but only norm contrarieties, and that the crucial difference
Kelsen on Monism and Dualism 15

between norm contradictions and norm contrarieties is that positive law accepts
the latter but not the former.
I am not convinced by Kelsen’s argumentation, however. First, the account of
norm contradictions, and therefore of the distinction between norm contradictions
and norm contrarieties, is not as clear as it might be. To begin with, in his discussion
of norm contradictions, Kelsen clearly assumes that the laws of logic – such as,
e.g., the law of non-contradiction, modus ponens, and the usual tautological
implications and equivalences – apply to norms.
46
That this is so is not clear,
however. For the validity (as distinguished from the derivability) of an inference (or
an argument) is defined in terms of truth,
47
and many – including Kelsen himself
in his later years – hold that norms lack truth-value.
48
One might, however, argue
that whether or not norms have truth values, and whether or not the applicability
of the laws of logic to norms presupposes that norms have truth values, a norm
conflict is a norm contradiction just in case the norm subject(s) cannot satisfy the
two norms simultaneously.
49
If, however, one wishes to speak of norm
contradictions in this broader sense, one must be prepared to count as norm
contradictions norm conflicts where a person (or a group of persons) ought to do
A and simultaneously ought to do B, if doing A (B) means that, for practical reasons,
he (or they) cannot also do B (A). For example, Smith ought to teach a class in
Stockholm on Thursday afternoon (OA), but he ought also to be in Oslo on
Thursday afternoon and examine a doctoral dissertation (OB). But if he is in
Stockholm on Thursday afternoon, he cannot also be in Oslo on Thursday
afternoon, and vice versa. Hence, on this analysis, the existence of the norms (OA)
and (OB) amounts in this case to a norm contradiction. I shall therefore assume
that Kelsen’s discussion of norm contradictions may be understood in terms of
satisfiability.
Moreover, Kelsen does not say much about the number of norm subjects
involved in a norm contradiction. But what he does say above (in the first quotation
in Section 3) suggests that he contemplates not only situations in which one and
Torben Spaak16
44 H. Kelsen, IPLT, 124.
45 H. Kelsen, IPLT, 124.
46 He later gave up this view, however. See H. Kelsen, General Theory of Norms, 1991, 189–251.
For a useful list of what I call “the usual tautological implications and equivalences,” see, e.g., P. Suppes,
Introduction to Logic, 1958, 34.
47 One usually says that a logically valid inference (or argument) is such that necessarily, if the
premises are all true, then the conclusion is true. See, e.g., E. Lepore, Meaning and Argument. An
Introduction to Logic Through Language, 2000, 10–3; M. Sainsbury, Logical Forms, 2nd ed., 2001,
5–9, 19–23.
48 See, e.g., J. Jørgensen, Imperativer og logik. 4 Theoria , 1938, 183-90; S. Blackburn, Attitudes and
Contents. Ethics 98, 1988, 501–17. For helpful discussions of this problem, see C. E. Alchourrón & A.
A. Martino, Logic Without Truth. 3 Ratio Juris, 1990, 46–67.
49 This has been suggested by Albert Hofstadter and J. C. C. McKinsey, among others. See A.
Hofstadter and J. C. C. McKinsey, On the Logic of Imperatives. 6 Philosophy of Science, 1939, 446–57.

the same person (or group of persons) ought to do A and not-A, but also situations
in which one person (or group of persons) ought to do A and another person (or
group of persons) ought to do not-A.
This way of conceiving norm contradictions does not seem promising, however.
We can easily think of a situation in which one person (or group of persons) ought
to do A and another person (or group of persons) ought to do not-A that does not
strike us as problematic. Consider, for example, a situation in which one legal norm
requires car drivers to drive on the road (and only on the road) and another legal norm
requires cyclists to drive in the bicycle lane (and only in the bicycle lane).
50
Clearly, this
is a situation in which one group of persons (car drivers) ought to do A (that is, drive
on the road) and another group of persons (bicyclists) ought to do not-A (that is, not
drive on the road), but this does not strike us as problematic at all. Consider also a
situation in which there is a legal norm that requires the mother of a child to fill out
a certain form (concerning the health of the child) and the father of the child to fill
out the form just in case the mother fails to do so, that for some reason the mother
refuses to fill out the form, and that the mother and the father cannot both
simultaneously fill out the form. In this situation, the mother ought to do A (that is,
fill out the form), because she is required by the law to do so, and the father ought to
do A, because he is required by the law to do so if the mother refuses. But since they
cannot both do A, it holds true, from the point of view of the mother, that she ought
to do A and that the father ought to do not-A, while it holds true from the point of
view of the father, that he ought to do A and that the mother ought to do not-A.
51
Yet, the situation cannot be said to be problematic.
One might, however, speak meaningfully of norm contradictions in certain
situations where one person (or group of persons) ought to do A and another person
(or group of persons) ought to do not-A, namely in situations where these persons
(or groups of persons) cannot both simultaneously do A and not-A. For example,
if one legal norm requires one person (or group of persons) to allow a line of cars
to cross a bridge and another legal norm requires another person (or group of
persons) to stop the cars from crossing the bridge, we surely have a norm
contradiction. That is to say, the question of whether a situation involving the
existence of two legal norms that requires different persons (or groups of persons)
to do A and not-A qualifies as a norm contradiction will depend very much on
what type of action A stands for.
In addition, one might speak meaningfully of norm contradictions in certain
cases of conflicting obligations and permissions, namely (i) in situations where one and
Kelsen on Monism and Dualism 17
50 I owe this example to Jaap Hage.
51 The reason why, from the point of view of the mother, the father ought to do not-A is that the
mother ought to do A and that this presupposes that the father does not do A (=that he does not-
A), and the reason why, from the point of the view of the father, the mother ought to do not-A is
that the father ought to do A and that this presupposes that the mother does not do A (=that she
does not-A).

the same person (or group of persons) ought to do A and may do not-A and, and
(ii) in situations where one person (or group of persons) ought to do A and another
person (or group of persons) may do not-A, if the persons involved cannot
simultaneously do A and not-A. Even though Kelsen does not mention such norm
conflicts, they qualify as norm contradictions pursuant to the general characterization
of norm contradictions that he provides, since they involve “norms whose content
is, logically speaking, mutually exclusive”, provided, in the latter case, that different
persons (or groups of persons) cannot simultaneously do A and not-A.
For these reasons, I shall assume that, on Kelsen’s analysis, a norm conflict is a
norm contradiction if, and only if, it involves two norms that provide (i) that one person
(or group of persons) ought simultaneously to do A and not-A, (ii) that two different
persons (or group of persons) ought simultaneously to do A and not-A, if these
persons (or groups of persons) cannot simultaneously do A and not-A, (iii) that one
person (or group of persons) simultaneously ought to do A and may do not-A, or
(iv) that one or more persons (or groups of persons) simultaneously ought to do A
and may do not-A, if these persons (or groups of persons) cannot both or all
simultaneously do A and not-A. And I shall assume that a norm conflict is a norm
contrariety if, and only if, it involves two norms, a higher-level norm and a lower-level
norm, where the latter was enacted on the basis of the former and has a content that
does not conform to its (the former’s) requirements.
This means that, on Kelsen’s analysis, the differences between these two types
of norm conflict are (a) that whereas the norms involved in a norm contradiction
concern the same type of action (A), or else different but incompatible actions
(A&B), the norms that are involved in a norm contrariety never concern the same
type of action; and (b) that whereas the norms involved in a norm contradiction
concern one and the same person (or group of persons), and, in certain cases,
different persons (or groups of persons), the norms involved in a norm contrariety
never concern the same persons (or groups of persons).
Secondly, while Kelsen is justified in making a distinction between norm
contradictions and norm contrarieties, and in holding that norm contradictions, but
not norm contrarieties, violate the epistemological postulate, he is mistaken in
attaching so much importance to the (alleged) acceptance by positive law of norm
contrarieties. What is important from the point of view of the epistemological
postulate is not this kind of indirect “acceptance” by positive law,
52
but whether a
norm contrariety is necessarily such that the different persons involved (the
respective organs enacting the higher- and lower-level norms) cannot both perform
the relevant action(s), that is, enact the respective norm. And since it is clear in the
case of norm contrarieties that the different persons (or organs) can both perform
Torben Spaak18
52 As I said in Section 4, if one cannot coherently maintain that one or more persons ought
simultaneously to do A and not-A, one cannot coherently accept that they ought simultaneously to do
A and not-A.

the relevant actions, we have no good reason to think that norm contrarieties violate
the epistemological postulate.
Thirdly, I can see no good reason to accept Kelsen’s claim that each and every
alleged norm contradiction invoked by the critics of monism is nothing more than
a norm contrariety. Moreover, Kelsen does not consider the possibility that the
norm conflicts that the dualists aim to avoid by invoking the doctrine of recognition
are really nothing more than norm contrarieties. If Kelsen can argue that the norm
conflicts identified by the dualists and presented as obstacles to monism are nothing
more than norm contrarieties, why could not the dualists, too, argue that the norm
conflicts they aim to dissolve by invoking the doctrine of recognition are really
nothing more than norm contrarieties? In any case, if the dualists could convincingly
argue that they are confronted by norm contrarieties, not norm contradictions, they
would not need to invoke the troublesome doctrine of recognition and would
therefore not need to struggle with the problems of solipsism.
Fourthly, Kelsen never mentions in his discussion of norm contradictions an
obvious tool that one might use to dissolve norm contradictions, namely the well-
known conflict-solving maxims: lex posterior, lex superior, and lex specialis.
53
Kelsen’s
omission is noteworthy, because a norm contradiction could in many cases be
dissolved by the judge or the theorist by invoking one of the maxims. If we take these
maxims into account, we might say that each norm in a legal system is valid in the
sense that it ought to be obeyed or applied, provided that it should not be set aside by
reference to one of the maxims. Why, then, does Kelsen not consider the possibility
of invoking the maxims in order to get rid of at least some of the contradictions that
might arise? And why don’t the dualists adopt this strategy? Of course, if Kelsen were
right that the alleged norm contradictions are not really norm contradictions, but only
norm contrarieties, he would not need the maxims. But, as I have said, I do not think
he is right, and it is in any case clear that invoking the maxims would be an easier and
less controversial way of solving the problem than arguing that the alleged norm
contradictions are nothing more than norm contrarieties.
8. Hägerström on the Unity of the Study Object
54
Kelsen accepts, as we have seen, an epistemological postulate, according to which
cognition represents its object as a unity in the sense that there are no contradictions
or inconsistencies in this representation. The underlying idea, I take it, is that a true
Kelsen on Monism and Dualism 19
53 The conflict-solving maxims are part of the legal tool box that judges and legal scholars use to
interpret and apply the legal raw materials. On this topic, see T. Spaak, Guidance and Constraint: the
Action-Guiding Capacity of Theories of Legal Reasoning, 2007, 43–4. For use of lex specialis, see, e.g.,
Green v. Bock Laundry Machine Co., 490 U.S. 504, 524–5.
54 This Section can be found, more or less verbatim, in T. Spaak, Naturalism and Non-Naturalism
in Legal Philosophy: Hägerström on Kelsen (forthcoming in S. Eliaeson et al, Axel Hägerström and
Modern Social Thought).

description of reality cannot be contradictory. One may, however, wonder whether
Kelsen’s sharp distinction between the ‘world of the ought’ (where law is located) and
the world of time and space (where human beings carry on their activities) can be
squared with this postulate.
55
Does it not follow from the sharp separation of these
two realms, that even though one may be able to have knowledge of law as such, one
cannot have knowledge of the application of law to the behavior of human beings?
Indeed, Axel Hägerström has objected to Kelsen’s theory of law precisely that
the distinction introduced by Kelsen between the ‘world of the ought’ and the world
of time and space means that Kelsen’s theory becomes socially irrelevant. The
reason, Hägerström explains, is that the ‘world of the ought’ cannot be thought of
as even existing alongside the world of time and space:
A legal prescript is, in fact, for him [Kelsen] a judgment concerning a supernatural
existent, which nevertheless (at least in so far as his view is carried out
consistently) must be completely realized in the world of nature. But this is an
absurd idea. The supernatural juridical system cannot be thought of as even
existing alongside the natural order. For no knowledge of any reality is possible
except through relating its object to a systematically interconnected whole. But
the supernatural and the natural systems, as being different in kind, cannot be
co-ordinated in a single system. Therefore, so far as I contemplate the one, the
other does not exist for me. But, if the jurist as such must abstract from the
natural order, it is to be feared that the legal prescripts which he sets forth will
be far too empty. He cannot, e.g., talk of legal transactions as juridical facts, for
that becomes altogether meaningless if one may not assume any natural causal
nexus. Again, he cannot speak intelligibly of punishment, since a ‘punishment’
which led to no consequences by way of natural causal connexions could not
be called a punishment. He must simply be left gasping for air!
56
What Hägerström is saying is that we cannot even conceive of the two worlds in
question – the ‘world of the ought’ and the world of time and space – as existing side
by side, because it is necessarily the case that everything that exists is part of the one
(and only one) all-encompassing framework that he mentions. And since, on
Hägerström’s analysis, this framework is the framework of time and space, there can
be no place for any other framework, such as a non-natural world of norms and
values.
Hägerström’s critique of Kelsen depends on a certain theory of reality.
Unfortunately, Hägerström’s texts on the subject matter are quite difficult to
Torben Spaak20
55 On the distinction between the ‘world of the ought’ and the world of time and space, see, e.g.,
H. Kelsen, Das Problem der Souveränität, 1st. ed., 1920, 99 note 1; H. Kelsen, Reine Rechtslehre, 2nd
ed., 1960, 6. For a discussion of this question, see T. Spaak, Realism about the Nature of Law
(forthcoming in Ratio Juris).
56 A. Hägerström, Kelsen’s Theory of Law and the State. In K. Olivecrona, ed., Inquiries into the
Nature of Law and Morals. Trans. C. D. Broad, 1953, 257–298, 267 (originally published the journal
Litteris 5:20-40 and 81–99.)

understand. But following a well-known (in Sweden) interpreter of Hägerström, the
Swedish philosopher Konrad Marc-Wogau,
57
I shall spell out this theory in three main
theses, namely (1) that ‘reality’ means the same thing as ‘determinacy,’ (2) that there is
a certain all-encompassing framework in which everything real can be found and
which excludes any other framework, and (3) that this framework is the spatio-
temporal framework.
The first thesis – that ‘reality’ means the same thing as ‘determinacy’ – is meant
to elucidate our common concept of reality and to answer the question, “What
does it mean to say that something, X, is real?” Hägerström’s answer, then, is that
to say that X is real is to say that X is determinate. As Marc-Wogau points out, this
thesis is aimed at theories that hold that reality is in some sense contradictory or
indeterminate,
58
and it captures an idea that has fairly often been put forward in the
history of philosophy.
59
This thesis, in my view, catches the idea that underlies the
epistemological postulate, espoused by Kelsen.
The second thesis – that there is a certain framework in which everything real
can be found and which excludes any other framework – is, as Marc-Wogau points
out,
60
aimed at so-called two-world theories, such as Plato’s theory of the forms.
Marc-Wogau describes the import of this thesis in the following way:
Everything determined or real constitutes one connected whole – “das in toto gegebene
Wirkliche” – beside which nothing else is conceivable. This one connected whole
contains parts that are consciousnesses, as well as parts that are not
consciousnesses. All determined objects “are comprised” in that connected whole,
but they are included in it in different ways. Certain objects, such as what I have
dreamt or imagined, are included in the connected whole merely as contents of
the imagination; other things, such as what I have experienced in the waking state,
and even my fantasies as psychic reality are included in the connected whole, not
only as imaginative content. If an object A has this characteristic in relation to
the conception of A, that it is included in the connected whole of reality “together
with” the conception of A, we say that A “exists”.
61
Kelsen on Monism and Dualism 21
57 K. Marc-Wogau, Hägerströms verklighetsteori [Hägerström’s Theory of Reality]. In K. Marc-Wogau,
Studier till Axel Hägerströms filosofi [Studies in Axel Hägerström’s Philosophy], 1968, 85-112; K. Marc-
Wogau, Känslans plats i Hägerströms ontologi [The Location of Feelings in Hägerström’s Ontology]. In
K. Marc-Wogau, Studier till Axel Hägerströms filosofi, 1968, 113-133. But see also P. Mindus, A Real Mind.
The Life and Work of Axel Hägerström, 2009, 46-70; E. Pattaro, I Will Tell You about Axel Hägerström:
His Ontology and Theory of Judgment. Ratio Juris 23:1, 2010, 123-56; B. Petersson, Axel Hägerströms
värdeteori [Axel Hägerström’s Value Theory], 1973, 16-29; and J. Bjarup, Ought and Reality. Hägerström’s
Inaugural Lecture Re-Considered. In P. Wahlgren, ed., Scandinavian Studies in Law, vol. 40, 2000, 11–70,
18–25; J. Bjarup, The Philosophy of Scandinavian Legal Realism. Ratio Juris 18:1, 2005, 1–15, 2-4.
58 K. Marc-Wogau, Känslans plats, 114.
59 K. Marc-Wogau, Hägerströms verklighetsteori, 91–2.
60 K. Marc-Wogau, Känslans plats, 114.
61 K. Marc-Wogau, Hägerströms verklighetsteori, 101-2. Translated into English by Robert Carroll.
The Swedish original reads as follows: “Allt bestämt eller verkligt bildar ett sammanhang – ’das in toto

The third thesis, finally – that the framework mentioned in thesis (2) is the
spatio-temporal framework – is aimed at theories according to which reality is
spiritual. Marc-Wogau explains that Hägerström reasoned as follows.
62
If we
conceive of our own existence, we must think of ourselves as being independent
of anyone’s conception of us. Moreover, if we assume, as we do, that some
conceptions belong to one person, A, that some other conceptions belong to some
other person, B, and that yet other conceptions belong to a third person, C, etc., we
must be able to distinguish between these different persons, and the only way to do
that is to invoke space-time coordinates; and this, of course, is to presuppose that
there is a world of time and space.
I find Hägerström’s theory of reality plausible, especially the view that we
cannot allow ourselves to operate with two or more worlds that we cannot conceive
together in one conception, and this means that, like Hägerström, I am disinclined
to accept the view that law exists in the ‘world of the ought.’
But the important question in this context is whether thesis (1) and the
assumption that underlies the epistemological postulate are identical, or at the very
least similar. As I intimated in the beginning of this section, I think they are; and if
I am right about this, we must conclude that in espousing the postulate, Kelsen is
unwittingly raising doubts about the cogency of his own non-naturalism about the
existence and validity of legal norms.
Torben Spaak22
gegebene Wirkliche’ – vid sidan om vilket intet annat är tänkbart. Detta ena sammanhang innehåller som
led dels medvetanden, dels led som inte är medvetanden. Alla bestämda föremål ’ingår’ i detta
sammanhang, men de ingår däri på olika sätt. Vissa föremål, som t. ex. det jag drömt eller fantiserat,
ingår i sammanhanget blott som föreställningsinnehåll; andra ting, som t. ex. det jag i vaket tillstånd
upplevt och även min fantasiföreställning som psykisk realitet, ingår i sammanhanget inte blott som
föreställningsinnehåll. Om ett föremål A har denna egendomlighet i förhållande till föreställningen
om A, att det ingår i verklighetssammanhanget ’jämte’ föreställningen om A, så säger man att A
’existerar’.”
62 K. Marc-Wogau, Hägerströms verklighetsteori, 105–7.