Basic Norm

Ratio Juris. Vol. 14 No.
2 June 2001 (143±75)
Presupposing the Basic Norm*

UTA U. BINDREITER
Abstract. According to Hans Kelsen, the basic norm is the necessary presupposition
of the positivistic cognition of law, making possible both descriptive legal cognition
and Verbindlichkeit. The nature of the presupposition in question here has been a
subject of controversy ever since. Presupposing the basic norm gives rise to formu-
lations that are neither purely descriptive nor purely normative. The author contends
that Kelsen's doctrine of the basic norm was intended to apply to all jurists irre-
spective of function. Kelsen, without being aware of it initially, invested the verb ``to
presuppose'' with an ambiguity that suits precisely this purpose.
Introduction
What, exactly, does it mean to ``presuppose'' the basic norm? One way to
proceed, in considering this central issue in Hans Kelsen's legal theory, is to
take up each of the following, more specific questions:
1. Who presupposes the basic norm?
2. Is it possible to defend the presupposition in a way that is convincing?
3. Finally, what difference does the presupposition make?
In section I of the paper, I take up Kelsen's own view(s) on question 1.
I turn, in section II, to question 2, examining the essentials of two promin-
ent reconstructions of the presupposition, namely Aleksander Peczenik's
theory of the basic norm-transformation and Joseph Raz's doctrine of the
legal man. Section III is devoted to question 3; here I consider Kelsen's own
characterization of the significance of the presupposition as well as two
interpretations of details of his language. In the last section, I sum up
sections I±III and offer some concluding remarks.
* I should like to acknowledge my indebtedness to my highly esteemed teacher, Professor
Aleksander Peczenik, both for encouraging me to write this paper and for his guidance and
invaluable advice. Profoundly interested in my work and ever ready for discussion, Professor
Peczenik offered me a great many helpful comments and suggestions for which I am truly
grateful. Heartfelt thanks go to Professor Stanley L. Paulson for allowing me to draw on his
expert knowledge of Kelsen. Professor Paulson responded patiently and with the utmost
kindness to my numerous queries and, what is more, advanced both constructive criticism and
valuable suggestions. Having also profited extensively from his written work, I owe him much.
# Blackwell Publishers Ltd 2001, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA.
144 Uta U. Bindreiter
Specifically, I argue that Kelsen in speaking of the doctrine of the basic

norm uses the German verb voraussetzen (to presuppose) in two different
senses. This verb exhibits, in other words, an artificial ambiguity; as it turns
out, this ambiguity lends itself to the overall purposes of Kelsen's doctrine
of the basic norm.
I. Kelsen's View(s) on the Addressee of the Basic Norm

The basic norm is formulated in a number of different ways (see Paulson
1993).1 In the best-known formulation,2 it figures as a duty-imposing norm,
stipulating that the constitution ought to be `òbserved.''3 Observed by
whom? And in what capacity are those people functioning who actually
presuppose the basic norm? On this issue there are, on Kelsen's part, several
more or less precise statements. In the following I shall set out the most
important of these statements and say something about the context in which
they were made.
During the decade preceding his so-called classical phase,4 Kelsen's
production must be viewed partly in the light of his intense occupation
with the problematic surrounding the concept of ``state,'' partly in the light
of his antagonism towards natural law. From about 1920, however, he
1
The analysis of ten different ways of formulating the basic norm leads, in Paulson's study, to
four different groups of formulations, namely, formulations related to (1) the unity of the legal
system; (2) the validity of the system; (3) empowerment, and (4) a definition of the basic norm
(Paulson 1993, 58±70). According to Paulson, the basic norm must be seen as the core of a
comprehensive BegruÈndung of the law, for it constitutes the ultimate basis of the law in three
different ways: first, dynamically (as the ultimate ground of legal competence); second, statically
(as the ultimate ground of legal validity); and third, normatively (as the ultimate ground of legal
obligation). Consequently, Paulson endorses a multi-dimensional concept of the basic norm
(ibid., 57f.).
2
According to Kelsen 1934, the schematic formulation of the basic norm of a national legal
system is the following: ``Coercion is to be applied under certain conditions and in a certain
way, namely, as determined by the framers of the first constitution or by the authorities to
whom they have delegated appropriate powers'' (Kelsen 1934, par. 28, 57). According to Kelsen
1960, the formulation is as follows: ``Coercive acts ought to be performed under the conditions
and in the manner which the historically first constitution, and the norms created according to
it, prescribe. (In short: One ought to behave as the constitution prescribes.)'' (Kelsen 1960, par.
34(c), 201). Kelsen 1934 and Kelsen 1960, respectively, refer to the English translations of the
First and the Second Edition of the Pure Theory of Law. The occasional use of the German text is
set out by (RR 1) and (RR 2).
3
The verb ``to observe'' is actually misleading in that it hides the complex function of the basic
norm and, therefore, the fact that there are different sets of addressees, namely (1) the legislative
organs, authorized to issue general norms, and those individuals (officials and others) whose
actions are guided through these norms; only with respect to the latter is the verb ``to observe''
appropriate. On this issue, see Kelsen 1925, par. 19(c), 99; 1979, 83.
4
Paulson 1998a, 161, discerns three major phases in Kelsen's work, namely (1) a constructivist
phase up to and including 1911, followed by a transitional phase (1913±1922); (2) a classical
phase (1922±1960), divided into a neo-Kantian period (1922±1935) and a hybrid period (1935±
1960), and (3) a sceptical phase (after 1960). Bulygin (1990, 298±9), on the other hand, separating
the Kantian and the positivistic strands in Kelsen's thought, distinguishes (1) a phase where
Kantian and positivistic elements are seen to coexist (1911±1940); (2) a transitional phase (1940±
1960), and (3) a predominantly positivistic phase (after 1960).
# Blackwell Publishers Ltd 2001.

Presupposing the Basic Norm 145
was strongly influenced by two other factors, namely, by Adolf Julius

Merkl's doctrine of the hierarchical structure of legal norms (Stufenbaulehre;
see Merkl 1917),5 and by neo-Kantianism. The former factor accounts for the
dynamic and empowering character of the basic norm, whereas the latter
gives rise to the transcendental grounding of the Pure Theory, with the basic
norm figuring as the necessaryÐand necessarily hypothetical (see Kelsen
1925, 104; see also Paulson 1990, 168±78)Ðpresupposition of legal science.
In an important paper of 1914, Reichsgesetz und Landesgesetz nach
oÈsterreichischer Verfassung, Kelsen mentions the idea of a ``Grundnorm'' for
the first time,6 understanding, by it, a conceptual necessity in the shape of an
ultimate norm. Legal constructions in general, he says, necessarily have as
their point of departure an ultimate or highest norm or system of norms that
is ``presupposed'' as valid7; the validity of this ultimate norm or system of
norms is taken for granted qua presupposition of legal knowledge. Legal
science presupposes this ultimate norm, albeit extra-systemically.8
According to Allgemeine Staatslehre (1925), Kelsen's early treatise on what
we would call constitutional law, legal science presupposes a Grundnorm
(Kelsen 1925, par. 20(b), 104), a Grund- oder Ursprungsnorm (ibid., par. 19(c),
99; par. 20(b), 104)9 or an Ursprungsrechtssatz (ibid., par. 20(b), 104). These
terms refer to a fundamental and hypothetical norm that is not really a part
of positive law.10 Still, this hypothetical normÐan early basic norm, so to
speakÐis seen as grounding the validity of positive law qua authorizing
norm or, as Kelsen puts it, qua ``juridico-logical constitution'' (see ibid., par.
36, 249).11 This, to be sure, is less than clear.
5
Kelsen had already identified himself with Merkl's doctrine in 1920 (Kelsen 1920, par. 29,
118±20); his first complete statement of Stufenbaulehre, however, is found in Kelsen 1923±1924;
see also 1925, pars 32±6, 229±55. In the Foreword of Kelsen 1923, xv±xvi, Kelsen payed tribute to
Merkl's achievement. On Kelsen's initial adoption of Merkl's doctrine, see generally Paulson
1996; 1998b, xxviii±xxx.
6
Without, however, using the term ``Grundnorm.''
7
``The foundation of juristic constructions must be norms, conceived of as valid legal
statements [_] one must always proceed from some sort of highest norm (or system of norms)
that is itself presupposed as valid'': Kelsen 1914, 216±7; see Paulson 1997a, 223.
8
``Thus, the question of the validity of this ultimate norm, accepted qua presupposition of all
legal cognition, lies beyond legal cognition'': Kelsen 1914, 217; see Paulson 1997a, 223ff. For this
idea of `Ùrsprung'' in the shape of a fundamental norm, Kelsen was actually indebted to Walter
Jellinek: In a paper of 1913, Jellinek had alluded to a highest norm, a norm that cannot be
justified by appeal to a still higher norm and that is independent of reality, with its validity due
not to human action but to a conceptual necessity, Denknotwendigkeit (see Paulson 1997a, 222ff.).
9
The term `Ùrsprungsnorm'' figures already in Kelsen 1920, vii±viii; par. 24, 98 n., 100; par. 26, 107.
10
This, however, does not according to Kelsen endanger the positivity of the legal system: See
Kelsen 1925, 104.
11
As Kelsen points out, the term ``juridico-logical constitution'' is to a certain extent the
counterpart to a term of natural law, namely, Urvertrag (Grundvertrag), that is, the contract that
first constitutes a ``state.'' As far as the positive law is concerned, Kelsen says, there is the same
theoretical need as within natural law of a presupposition that would constitute the unity in
the plurality of all (municipal) legal actsÐthat is, normsÐforming a system (Kelsen 1925, 250).
The idea of a constitution in the juridico-logical sense stems from Alfred Verdross (see Kelsen
1923, xv). On the basic norm qua juridico-logical constitution, see Paulson 1993, 59, 65.

If a (higher-level) norm is thought to validate another (lower-level) norm,

it must then be seen, Kelsen contends, as having authorized the issuance of
the latterÐthus, if the highest-level positive norms of a legal system are to
be regarded as valid, the non-positive basic norm of that system must be
seen as having authorized the issuance of those norms. This, according to
Kelsen, will indeed be the case if the basic norm is seen as a constitution that
exists, in a temporal sense, before the positive constitution.12
Thus, assuming the validity of the positive constitution, its validity then
implies that the highest norm-issuing organ is eo ipso considered as having
been authorized or invested with the competence to issue valid legal norms
(see Kelsen 1925, par. 19(c), 99). In other words, the norms issued by the
highest legal authority have been ``qualified'' as legal norms (see ibid., par.
20(b), 104; par. 36(a), 249).13 Thus, the authorization to issue legal norms and
the term ``juridico-logical constitution'' turn out to be two sides of the same
coin (see Paulson 1993, 65).
Certainly, to invest an ultimate authority with the competence to issue
valid legal norms must be seen as one of the essential functions of a basic
norm. Empowerment alone, howeverÐor so Kelsen arguesÐis insufficient;
in addition, there must be a guarantee that the norms emanating from an
ultimate authority are issued in a way that makes sense.
Kelsen's short monograph, Die Philosophischen Grundlagen der Naturrechts-
lehre und des Rechtspositivismus (1928),14 marks a new development in this
and other respects. Hitherto, the basic norm had mainly figured as the
ground of formal unityÐthat is, as the sole criterion by means of which the
membership of legal norms (and, thus, the identity of the legal system as a
whole) can be established (see Paulson 1997b, 336±7, 342). In Philosophische
Grundlagen, the basic norm likewise figures as the ground of material unity;
as such, it ensures not only that law is that which is issued by an authority
but also that what the authority issues fits into a meaningful whole.15
At this stage of his work, that is, during the first half of Kelsen's classical
phase,16 the basic norm has neo-Kantian contours, purporting to function
qua ultimate and ``transcendental'' ground of positive law. In this paper,
12
On the problematic of the relation between the ex ante- and the ex post-perspective, see
Paulson 1993, 64.
13
By (legally) qualifying the constitution, the basic norm provides an Erzeugungszusammenhang
between the norms, that is, a dynamically unified system of norms: See Kelsen 1925, par. 36(a),
249.
14
Reprinted in Klecatsky et al. 1968, 281±350; trans. Kelsen 1945b, 391±446. I have used the
English text.
15
``The basic norm states that under certain conditions X a certain consequence A ought to take
place. It thereby states that, under like conditions X, non-A ought not to take place at the same
time. For the principle of non-contradiction must be posited in the idea of law, since without it
the notion of legality would be destroyed'' (Kelsen 1928, II.F, 406; see also Klecatsky et al. 1968,
par. 12, 299f.). The function of the basic norm is not only ``to recognize a historically given
material as law, but also to comprehend it as a meaningful whole'' (ibid. IV.B (d), 437; Klecatsky
et al. 1968, par. 36, 339); see also Kelsen 1960, par. 34(d); Paulson 1997b, 332ff., 343f.
16
On periodization, see note 4 above.

Kelsen, following Kant, first poses his own transcendental question (to be
fully elaborated in Kelsen 1960), namely: ``How is positive law, qua object of
cognitive legal science, possible?'' (Kelsen 1928, IV.B(d), 437). The answer to
this question is the hypothetical basic norm as suchÐthe necessary pre-
supposition of positivistic legal cognition.17
The basic norm, Kelsen says, is an expression of the hypothetical and
relative character of a positive norm system. As such, it is not only an
assumption (Annahme) of a specific legal theory (Kelsen 1928, I.D, 395). It is,
in addition, the formulation of the necessary presuppositions of all legal
cognition. That is, it brings to the surface and clarifies what lawyers, without
being conscious of it, do when in cognizing legal material they refuse on the
one hand to go back to natural law, and, on the other, understand the data
that are given not as facts, but as norms.18
Now the immediate object of legal cognition, the cognoscendum, is, for
Kelsen, the legal norm as such, not the legal proposition. He claims that
legal cognition presupposes that the principle of non-contradiction be
applied within the sphere of Sollen, that is, directly to norms.19 In his view,
principles of interpretation are in the main ``presuppositions'' of legal cog-
nition and must, therefore, be part of the sense of the basic norm (see Kelsen
1928, II.F, 407; Klecatsky et al. 1968, par. 12, 300). It follows that it is left to
the latter to guarantee that a legal system, even if less than fully consistent
(see ibid., 406f. and IV.B(d), 437; Klecatsky et al. 1968, par. 12, 299f. and par. 36,
339), must nevertheless be seen as constituting a meaningful whole.20
Reine Rechtslehre (1934), the First Edition of Kelsen's Pure Theory, shows his
effort to arrive at a ``middle way'' between the (moral) concept of validity
found in natural law and the (reductivist) concept of validity familiar from
traditional legal positivism (see Paulson 1993, 53±5). The basic norm is the
source of a dynamically21 grounded unity as well as the source of the
normativity of the law,22 transmitting, by means of delegation, the quality of
`òught'' (Sollen) to the other norms of the system (see Kelsen 1934; par. 29, 58).
17
For an exhaustive discussion on Kelsen's posing of the juridico-transcendental question, see
Paulson 1998b, xxxvff.
18
On the normativity problematic and Kelsen's ``middle way'' between the traditional theories,
see ibid.
19
``The principle of non-contradiction [_] applies equally to the normative (òught') and to the
factual (ìs') sphere. In both, the judgments À ought' and À ought not' are just as mutually
exclusive as À is' and À is not' '': Kelsen 1928, II.D, 402; see also Klecatsky et al. 1968, par. 10, 295.
20
Delegation alone cannot accomplish this: ``With the aid of the basic norm the legal materials
which have been produced as positive law must be comprehensible as a meaningful whole, that
is, they must lend themselves to a rational interpretation. The pure principle of delegation can-
not guarantee this'': ibid., II.D, 402; Klecatsky et al. 1968, par. 10, 295; see also ibid., II.F, 406±7;
III.A, 407f.; IV.B(d), 437; Klecatsky et al. 1968, par. 11, 299; par. 12, 300; par. 36, 339.
21
``The basic norm [_] is simply the basic rule according to which the norms of the legal system
are created; it is simply the setting into place of the basic material fact of law creation [die Ein-
Setzung des Grundtatbestandes der Rechtserzeugung]. This basic norm, the point of departure for a
process, has a thoroughly formal, dynamic character'': Kelsen 1934, par. 28, 56 (RR 1, 64).
22
``Rooted in the basic norm, ultimately, is the normative import of all the material facts
constituting the legal system'': ibid., par. 29, 58.

From this point henceforth, Kelsen argues that this basic norm is to be
understood in Kantian or neo-Kantian terms,23 and he sets himself the task of
working up an epistemic justification of his notion of the normativity of law:
Without the presupposition of the basic normÐthis is his argumentÐthat
which we know to be the case (namely, that law is normative) could not be the
case.24
Law is norm, not will (see Kelsen 1934, pars 5, 7, 8, 11(a), 15, 19, 29). To
say that law is valid means, for Kelsen, two things: first, that it was issued
(``posited'') by an authorized organ25 and, second, that it manifests the
quality of `òught,'' as drawn from the basic norm.26 Following Merkl's
doctrine of Stufenbau, Kelsen explains that to issue or ``posit'' a norm is at the
same time to apply a norm (see Kelsen 1934, par. 31(f), 70). The issuing of a
general norm, for instance, implies the application of a constitutional norm.
Thus, when the original, historical constitution of a legal system is passed,
one will want to say that that marks the application of a still higher-level
normÐthe basic norm. This, however, is not the case. Since the basic norm is
not a positive norm at all, not a norm issued by a still higher authority,
ordinary norm application (Vollziehung) is impossible (see ibid.).
How, then, is the ``hypothetical positing'' of the basic norm to be
understood? This is the question that remains. Kelsen asserts that the act of
``presupposing'' the basic norm has the character simply of norm-positing,
and he underscores this point graphically by writing ``Voraus-Setzung der
Grundnorm'' (Kelsen 1934 (RR 1), par. 31(f), 83). What this assertion seems to
amount to is the following.
Although not issued in the manner of other norms, the basic norm is
nevertheless voraus-gesetzt (with stress on ``vor''), that is to say, pre-posited
in the sense of `ònly hypothetically posited within juristic thinking.'' If it
were not misleading, one would be inclined to say that Kelsen's writing
``Voraus-Setzung'' indicates that this sort of positingÐnamely hypotheti-
callyÐis logically prior27 to the regular issuance of other, positive norms. It
seems clear that a norm that is ``pre-posited'' in this way cannot have any
normative consequences whatever. Kelsen, howeverÐat least during the
greater part of his long, classical phaseÐdid not appreciate this point.
Owing to the influence of Merkl (dynamic character of the norm hierarchy)
23
``The òught' designates a relative a priori category for comprehending empirical legal data'':
ibid., par. 11(b), 24f. Sollen is the specific meaning through which conditional and consequent
are tied together in the ``reconstructed'' legal norm, that is, the legal statement: See ibid., par. 29,
58.
24
On Kelsen's transcendental argument, see generally Paulson 1990, 1997c.
25
`À norm is valid qua legal norm only because it was arrived at in a certain wayÐcreated
according to a certain rule, issued or set according to a specific method'': Kelsen 1934, par. 28,
56.
26
See note 22 above.
27
One might argue that jurists, if they were capable of explaining the tacit knowledge
underlying their statements, would say that the hidden premises of these consist in implicit
assumptions about the law.

and the neo-Kantians (transcendental ``grounding'' of the Pure Theory28) the

verb ``to presuppose'' shows signs, in the Pure Theory of Law, of having
acquired a distinctly normative colouring.
Attributing normative consequences to the presupposing of the basic
norm was, however, a state of affairs that, sooner or later, would necessitate
a change of position with respect to the presupposing individuals. That is,
Kelsen over a long period of time entertained doubts as to the correctness of
having legal science presuppose the basic norm. He addresses this prob-
lematic in a paper of 1952, namely Was ist ein Rechtsakt? (Kelsen 1971).29 Here
the basic norm figures intra-systemically, that is, as a ground for or basis of
the constitution and, consequently, as a part of the constitution (see Kelsen
1971, 155; 1984, 201). As the meaning (Sinn) of an act of will, the basic
normÐand this is Kelsen's position from this point henceforthÐcannot be
said to be presupposed by legal science, for legal science lacks competence
to posit norms.
What Is a Legal Act? was written in reply to an essay by Alfred Verdross
(1951)30 on the question of how an act is legitimated as a legal act. Verdross
had claimed that there were acts that could not be legitimated in the normal
way (that is, by being accepted or recognized as ``legal'' by officials)Ðsuch
as, for instance, the norms issued by legal organs who have overstepped the
bounds of their authority (see Kelsen 1971, 160±1; 1984, 206±7), or norms
purportedly issued in the name of a revolution (see Kelsen 1971, 162±3; 1984;
207ff.). Verdross arrived at the conclusion that the question of the legitimacy
of a legal act was not, in the last analysis, answered by legal science and on
the basis of the legal system at all, but rather by sociology (see Kelsen 1971,
153; 1984, 200).
Kelsen begins by scrutinizing the question. In this connection he distin-
guishes a conceptual and a concrete component, namely, (A) the criteria of
``legal act,'' and (B) the question of who, in a concrete case, is to decide
whether a specific act is, indeed, a legal act (see Kelsen 1971, 153; 1984, 200).
In Kelsen's view, the answer to (A) is left to legal science, for it is the task
of legal science to define all legal concepts. Referring to the doctrine of
hierarchical structure (Stufenbaulehre), Kelsen shows that the definition of
``legal act'' is actually provided by the legal system itself, employing criteria
of a strictly formal nature. A legal act, in Kelsen's view, is an act by means of
28
Influenced by Merkl and neo-Kantian philosophy, Kelsen was anxious to provide his theory
with a ``transcendental'' grounding. The basic norm, intended to figure as ``transcendental-
logical presupposition'' of legal science (see Kelsen 1960, par. 34(d), 201ff.) is no longer the basic
norm of the transitional phase; rather, it is the ultimate ground of material as well as formal
unity, transmitting the sense of `òught'' to the lower-level norms of Stufenbau. On the neo-
Kantian dimension of the Pure Theory, see generally Paulson 1992c, 1994.
29
Appeared first 1952 in O È sterreichische Zeitschrift fuÈr oÈffentliches Recht and reprinted in Klecatsky
et al. 1968, vol. 2, 1381±93. I have used the English translation What Is a Legal Act? (Kelsen 1984).
The footnotes refer to Kelsen 1971 and Kelsen 1984, respectively.
30
Reprinted in Klecatsky et al. 1968, vol. 2, 1375±80.

which a legal norm is issued (``posited,'' gesetzt) or applied, provided that

this is done in conformity to those (higher-order) norms within the norm
hierarchy that govern the creation and the application of law within the
legal system (Kelsen 1971, 154; 1984, 200).
The answer to (B), on the other hand, is not a matter over which legal
science has any concern at all, for legal science lacks competence to decide
anything whatever. Rather, the legal quality of an act is established through
an authentic, constitutive, and legally binding decision on the part of the
law-applying organs, a decision to which the legal system attaches specific
consequences (Kelsen 1971, 157; 1984, 203).
Decisions, Kelsen contends, imply a function of the will, not of the
intellect, and decisions, therefore, are in fact individual legal norms (see
Kelsen 1971, 156±7; 1984, 202±3). From this it follows, on the one hand, that
decision-makingÐthe issuing of individual legal normsÐmust fall to the lot
of those individuals who have competence to issue norms, and it follows, on
the other, that these same individualsÐthat is, the courts and other law-
making bodies (see Kelsen 1971, 155ff.; 1984, 202ff.)Ðand they alone can
pre-``posit'' the basic norm.31
Within this context, Kelsen once again spells out his comprehensive
notion of ``legal positivism'' (see Kelsen 1928, II.E, 405±6 and IV.B(d), 437;
Klecatsky et al. 1968, par. 9, 294f. and par. 36, 339). If, he argues, legal posi-
tivism is understood to mean that positive norms, that is, ``posited'' or
issued norms, alone can be conceived as law, then legal positivism is pos-
sible only if a qualification is made, namely, that not only positive norms but
also the constitution, including the presupposed basic norm, fall under this
concept (Kelsen 1971, 155; 1984, 201±2). According to Kelsen, it is absolutely
necessary to conceive of a positive legal system in this way, for a positive
constitution cannot be regarded as binding without, at the same time,
presupposing a basic norm as underlying or ``founding'' it (see ibid.). If
constitutional norms are part of the legal system, then the basic norm,
the foundation of the constitution, must be conceived of as belonging to
this legal system, too.
Thus, the basic norm is not positive law. It is not an issued, merely pre-
supposed norm. Nevertheless, it is, as such, the (transcendental) presup-
position of all positive law. The officials who decide on the legal quality of
an act unconsciously presuppose the basic norm, whilst legal science con-
tents itself with describing and analyzing this process.
The paper What Is a Legal Act? clearly indicates the change in Kelsen's use
of ``presuppose'' and ``presupposition.'' At that time, that is, during the
31
In this context, it is of the utmost importance to scrutinize the German text: `Ès ist daher nicht
ganz richtig zu sagen, wie ich selbst wohl mitunter gesagt habe, dass die Rechtswissenschaft die
Grundnorm voraussetzt. So wie die Rechtswissenschaft keine Norm setzen kann, kann sie auch
keine Norm voraussetzen'': Kelsen 1971, 163; 1984, 209.

classical phase, the verb ``to presuppose'' (in German: voraussetzen) seems to
have been, for Kelsen, endowed with a certain quality which is linked to the
verb ``to posit'' (in German: setzen, that is, a norm) and can hardly be
described otherwise than as embracing the notion of bindingnessÐnamely,
the bindingness of ``posited'' law. And if this is so, it will explain Kelsen's
assertion that legal science does not ``presuppose'' the basic norm but rests
content instead with a statement to the effect that a constitution can be
conceived of as binding, and an act as a genuine legal act, only if the basic
norm is presupposed (namely, by those who apply the law).32
In about 1960, Kelsen changes his mind yet another time. As revealed
in the Second Edition of Reine Rechtslehre (Kelsen 1960 (RR 2), par. 34(d),
208±9), he now distinguishes between, on the one hand, the (purely cogni-
tive) presupposition of legal science and, on the other, the (normative)
presupposition of the basic norm on the part of the law-applying organs.
Owing to this distinction, Kelsen feels justified in asserting, once again,
that legal science can be said to presuppose the basic norm. The basic norm,
he now says (see ibid.), is presupposed by all who interpretÐor wish
to interpret33Ðthe subjective meaning of a legislative act as its objective
meaning, that is, as an objectively valid norm. In contrast, however, to
an interpretation provided by the norm-positing organs (see Kelsen 1960
(RR 2), par. 34(d), 207)Ðinter alia, the courtsÐthe interpretation by legal
science is purely cognitive in nature (see Kelsen 1960, par. 34(d), 204, n. 72).
Thus, Kelsen's doubts as to whether or not legal science can be said to
presuppose the basic norm seem, by 1960, to have been dispelled, for this
sort of presupposition represents the purely epistemological assumption of
an ultimate norm `ìn der BegruÈndung der objektiven Geltung der Rechtsnormen''
(Kelsen 1960 (RR 2), par. 34(d), 206), that is to say, in establishing the object-
ive validity of legal norms. As Kelsen is quick to add, however, from the fact
that the legal scientist pre-supposes the basic norm merely conceptually it
does not follow that he cannot presuppose it, at the same time, in the other
sense. It is not to say that he cannot presuppose it normatively, qua binding
norm.34 And this second possibility gives rise to the question: When would a
legal scientist be prompted to do so?
32
``Legal theory simply determines that ifÐand only ifÐthe basic norm is presupposed in the
proceeding subsequent to the act establishing the constitution, then the constitution can be
recognized as a binding legal norm and the act giving rise to the constitution can be recognized
as a legal act. Therein lies the hypothetical character of the basic norm'': Kelsen 1971, 163; 1984,
209.
33
On Deutung (interpretation), see Walter 1964, 33 n. 70.
34
``The science of law remains a merely cognitive discipline even in its epistemological
statement that the basic norm is the condition under which the subjective meaning of the
constitution-creating act, and the subjective meaning of the acts performed in accordance with
the constitution, are interpreted as their objective meaning, as valid norms, even if the meaning of
these acts is so interpreted by the legal science itself '': Kelsen 1960, par. 34(d), 204; my emphasis.

II. Is It Possible To Defend the Presupposition in a Way that Is

Convincing?
Kelsen's basic norm has attracted the attention of a host of scholars. Among
those whose research has promoted an understanding of its complex
character, Aleksander Peczenik and Joseph Raz, in particular, are worthy of
special attention. Peczenik and Raz think along quite different lines both
with regard to the status of the basic norm and the manner in which it is
presupposed, and both perspectives are highly instructive.
According to Peczenik, the basic norm is a duty-imposing norm, accord-
ing to Raz, it is a power-conferring rule. Peczenik's inquiries, focusing on
the basic norm qua basis for the validity of positive law, lead to the doctrine
of the law-creating transformation and the doctrine of the conditional basic
norm; Raz's examination of Kelsen's RechtssaÈtze leads to a reconstruction of
the basic norm theory and to the doctrine of the legal man, which has to be
regarded as an important albeit controversial step toward greater insight
into the modality of legal statements.
In the first part of the present section, I shall turn to the essential points of
Peczenik's basic norm-transformation, and in the second part, to an exam-
ination of Raz's doctrine of the legal man.
II.1. Aleksander Peczenik: Basic Norm-Transformation

According to Peczenik (1995, 162ff.), legal systems consist of two kinds of
rules, namely, constitutive rules and regulative rules. Constitutive rules are
characterized as setting the conditions for the existence of institutional facts,
whereas regulative rules are rules with respect to action and goal. Following
Peczenik, the Kelsenian basic norm is a rule of action, prescribing a duty,
namely the duty to apply and follow the constitution (ibid., 117).
Peczenik's research on the basic norm concentrates on two problems,
first, that of how the basic norm can be said to justifyÐthat is, legitimize
Ðpositive law, and, second, that of how the basic norm itself can be justi-
fied or, in other words, how itsÐmerely presupposedÐvalidity can be
established (ibid., 118f., 541f.). It may be said that both problems find their
solution within the doctrine of the basic norm-transformation, a doctrine
that contains the analysis of a basis or ``ground'' in the normative sense.35 On
one level, Peczenik argues that the basic norm may indeed be said to
constitute a ``ground'' of the (external) validity of law, and, on another
level, he contends that it is possible to legitimize the assumption of the
basic norm by adducing an additional normative premise. In the follow-
ing, I shall briefly take up the salient points of both these levels of
argumentation.
35
``[_] once the apex norm has been presupposed, the stage is set for using it as a `ground' for
legal validity'': Peczenik 1981, 282.

On the first or `èxternal'' level, the point of departure is what Peczenik

calls a ``normal'' society, that is, a system whose legal norms are neither
generally nor extremely immoral (ibid., 541). In a society of this kind, jurists
take for granted that the constitution of the system is indeed valid law and
that, being such, it ought to be followed. Peczenik points out in this con-
nection that the jurists, in taking such for granted, are in fact doing no more
than what is perfectly in accordance with Kelsen's basic norm. For the basic
norm says exactly the same thing, namely, that the constitution is valid (and,
therefore, ought to be `òbserved''; see ibid., 118f., 541f.; see also Kelsen 1934,
par. 16, 34±5).
Thus, to assume the basic norm in a ``normal'' societyÐwhere there is no
reason whatever to question the validity of the constitutionÐis entirely
unproblematic; the constitution is valid because the basic norm says it is.
Therefore, Peczenik maintains, it is perfectly appropriate to presuppose
Kelsen's basic normÐfor as long as the question as to why the constitution is
valid is not raised (see Peczenik 1995, 541).
What is more, Peczenik (1981, 281f.) contends, the activity of legal
reasoning actually implies taking for granted that the constitution is valid
and ought to be followedÐfor how would legal reasoning be at all possible
if it were not assumed, on the part of the jurists, that the highest positive
norms of the system are indeed valid? And once the basic norm has been
assumed conceptually, it can then be said to constitute the ``ground'' for the
validity of law.36
Given that jurists actually exhibit a tendency to take for granted that the
constitution ought to be applied and followed, then, Peczenik says (without,
however, adhering to this view himself), one might claim that there is a
social norm stipulating that the constitution ought to be followed. Thus, one
might argue that the basic norm `èxists'' within legal practice (see Peczenik
1995, 118f.). Peczenik nevertheless grants that it would be decidedly odd to
claim that a purely conceptual presupposition is the ground of the validity
of positive law37; to surmount this difficulty, he offers the theory of the basic
norm-transformation (see Peczenik 1981, 287f.; 1995, 543±5).
On this second or `ìnternal'' level Peczenik is investigating what might
well take place in the minds of jurists who `àssume'' the basic norm. Jurists,
he says, draw from certain premises the normative conclusion that the
constitution is valid and ought to be observedÐthat is to say, they in effect
draw out the basic norm as their conclusion. The premises state what jurists
know to be the characteristics of a valid legal system, and the inference can
then be described in the following way.38 A jurist is able to specify the facts
36
Since the lawyers in fact speak about the constitution as being valid law, ``they must speak
inconsistently unless they presuppose the apex norm'': Peczenik 1981, 282.
37
``This is a puzzle'': Peczenik 1981, 282.
38
The author would like to thank Professor Peczenik for discussing at great length the different
steps involved in this inference.

or the criteria for what is identified as valid law. In particular, these facts or
criteria can be specified by spurious normative sentences. Taking these as the
point of departure, genuine normative sentencesÐthat is, normsÐmay then
be inferred. The inference here is not, strictly speaking, a logical inference;
rather, it represents an intuitive step, which Peczenik calls the ``law-creating
jump'' (Peczenik 1995, 544).
The law-creating jump is a jump from certain observable criteria (of the
validity of a normative system) to legal validity. Among them is found,
for example, the criterion that the norms of the system are structured
hierarchically and contain constitutive rules (see ibid., 543±4). Thanks to the
jump, the jurist finds himself, as it were, ``within'' the legal system, he
is bound by it, and his statements are legally normative without being
norms.39
A jurist does not pause to consider why the constitution of a given
systemÐsay, of Swedish lawÐought to be observed; rather, he is intui-
tively capable of recognizing this system as one that ought to be fol-
lowed (see ibid., 544). This intuitive capability is owing to the upshot
of the jump, namely the so-called law-creating transformation by means of
which the jurist's knowledge of facts is transformed into knowledge of
valid law.
The import of Peczenik's law-creating jump is this. The jurist's
spontaneous inferenceÐthe jumpÐcan be turned into a logical inference
by means of the assumption (or presupposition) of a genuinely normative
premiseÐstrictly speaking, by assuming an additional hypothetical
premise, normative in character, which links certain facts or criteria to the
legal validity of the system and, thereby, to the duty of conforming to the
systemÐor, more precisely, to its norms (see ibid., 544±5).
As far as the doctrine of the basic norm is concerned, the additional,
presupposed premise simply says the following: If the norms of the system
in question are structured hierarchically, and if the system is coercive, and
if it is by and large effectiveÐthen the system in question is a valid legal
system whose constitution ought to be observed; then, in short, the basic
norm (saying exactly the same) can be presupposed. Owing to this add-
itional premise, the conclusion, that is, the norm to the effect that the con-
stitution ought to be observed, is transformed into a logically correct
conclusion (see ibid., 545).40
To summarize: Peczenik's doctrine of the basic norm-transformation is
illuminating, in my view, on the question of what, exactly, Kelsen under-
stood by the notion that the basic norm is presupposed in legal thinking. The
analysis of the so-called law-creating jump shows just how the basic norm,
by means of an additional normative premise, is inferred from those facts
39
The legal `òught'' of the conclusion, however, is not identical with the premises': See Peczenik
1981, 288.
40
On the concept of transformation in general see Aarnio, Alexy and Peczenik 1981, 137ff.

that, according to Kelsen, characterize valid lawÐhierarchical structure,

coercion, and effectiveness. Thus, the basic norm-transformation shows that
what the jurists unconsciously presuppose41 is the link between observable
facts and the validity of the legal system as such, and Peczenik is successful
in analysing this link. Once the basic norm has been presupposed, it can be
used as a (normative) ``ground'' for the validity of law.42
Among the many questions to which Peczenik's theory gives rise, at least
two ought to be mentioned in the present context; both concern the
sufficiency of Kelsen's three conditions for legal validity.
It was shown that in a ``normal'' society these three conditions are fully
sufficient, in Peczenik's view, for presupposing the basic norm. However,
Peczenik draws attention to the fact that even in a ``normal'' society, there
will be a discrepancy between the conditions for legal validity and the three
Kelsenian conditions in the sense that the norm hierarchy of the system,
according to Kelsen's theory, has to be perfect.43 Now if this is the case, only
a single transformation is necessary, that is, the basic norm-transformation,
and it yields the conclusion that the constitution is valid. If the norm
hierarchy, however, is not perfect, several (categorial as well as criteral)
transformations are needed, each of them yielding a separate normative
conclusion (see Peczenik 1981, 291±2).
What is more, as soon as the validity of the constitution is called into
doubt, the basic norm will reveal itself as, after all, a weak ``ground'' for
legal validity (see Peczenik 1995, 541). As Peczenik points out, the additional
hypothetical premise (which, as it stands, is comparatively simple) will then
probably have to make room for still other additional premises, namely,
premises that inter alia give expression to moral norms. In that case, the
jurist's normative conclusions will no longer be ``grounded'' in the Kelsenian
basic norm, but rather in a conditional basic norm (see ibid., 545ff.).
II.2. Joseph Raz: The Legal Man

In section II.1 the basic norm figured as a rule of action and species of a
social norm, the `èxistence'' of which may be said to be proved by the
41
`Ìn formulating the basic norm, the Pure Theory of Law is not aiming to inaugurate a new
method for jurisprudence. The Pure Theory 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 natural law as the basis of the validity of positive law, but nevertheless understand
the positive law as a valid system, that is, a norm, and not merely as factual contingencies of
motivation'': Kelsen 1934, par. 29, 58; my emphasis; see also Kelsen 1960, par. 34(d), 204±5.
42
In this way we broach the core of the issue at hand, namely, the question of what sort of
``ground'' the doctrine of the basic norm requires in order to make it work: See Peczenik 1981,
282.
43
This is the case if the legal system shows a number of levels and the validity of any legal norm
is seen to derive from valid norms of a ``higher'' levelÐthat is to say, if the norm has been
created in accordance with a valid norm of competence and a procedural norm of higher
standing.

existence of legal argumentation. The tacit assumption of this norm, on the

part of the jurist, leads to legally normative statements and the (legitimate)
use of normative terms.
Joseph Raz thinks otherwise. According to Raz (1980, 164±5; see also 1986,
79±97), a legal system contains two types of rules, namely, D-laws (duty-
imposing laws; see Raz 1980, 147ff.) and P-laws (power-conferring laws; see
ibid., 156ff.; see also Raz 1979, 27). Referring mainly to Kelsen's General
Theory of Law and State, he contends that the basic norm belongs to the latter
category: The basic norm, he says, is a rule of competence (see Raz 1980,
95)44 and, as such, a special rule of competence, namely, a so-called PL-law.
PL-laws are laws conferring power to legislate.45
The fact that Raz regards the basic norm as a rule of competence is by no
means a sign of his failure to appreciate its complex character. On the
contrary. In Raz's view, it is the case that laws conferring power to legislate,
that is, PL-laws, guide the behaviour of the (thus to be authorized) legislator
as well as that of the (potential) subjects of the latter's enactments and,
therefore, may be restated as D-laws (see Raz 1980, 167).46 However, formu-
lating the basic norm as a duty-imposing rule, as Kelsen does, is in Raz's
view unfortunate (see Raz 1980, 96f.); such a formulation, Raz believes,
cannot be squared with what the basic norm was essentially meant to
achieve, namely, to invest the first legislator with legal competence, thereby
endowing his legislative acts with the quality of being binding (see Kelsen
1925, par. 19, 99).
Precisely this claim, namely, to endow legislative acts with binding force,
has Raz turning with appreciation to the doctrine of the basic norm, not-
withstanding its problematic character as Raz sees it.
In general, Raz takes a dim view of basic norms, not least Kelsen's. In
Raz's view, basic norms have no genuine function: They are, he contends,
of no use in establishing the unity or identity of a legal system, and they
cannot justify legal norms eitherÐin short, Raz thinks that there is no reason
44
See also Raz 1980, 66 (here, Raz quotes Kelsen 1945, 120: ``[_] confers legal authority [_]'' and
114: ``qualifies a certain event [_] starting-point of a norm-creating process''); see also Raz 1979,
126f.
45
PL-norms are characterized by the fact that the reaction to the performance of the norm-act is
prescribed by other norms. These, however, do not exist when the PL-norm is createdÐrather,
they are created by the norm-acts of the PL-norm themselves: See Raz 1980, 163.
46
Raz explains this in the following way. The standard form of a PL-law is: ``X has legislative
powers to legislate laws by doing A in C.'' Whenever this statement is true, there is also a
statement (the standard description of a so-called `òbedience law''): ``Y has a duty to (or ought
to) obey X if he does A in C.'' This statement, then, is true, too, and it is entailed by the first
statement, and entails it. Therefore, if PL-laws, as Raz claims, are an explication of laws
authorizing legislation, so are obedience laws, and vice versa (see Raz 1980, 166); in other
words: ``[A]uthorizing X'' may imply `òrdering Y'' (see Kelsen 1979, 83) and `òrdering Y'' may
imply ``X is authorized to give orders'' (see ibid., 210). On authorization, see Spaak 1992. Spaak
contends that competence norms can logically be reduced to fragments of norms of action,
addressed to officials, and thus, lacking a specific type of normativity, ought not to have the
status of a distinct normtype: See ibid., 177±9.

whatever to accept the existence of basic norms (see Raz 1980, 138; 1979,
78±102, 127±9; see also Paulson 1997b, 342).47
Despite these serious reservations, Raz nevertheless acknowledges merits
in Kelsen's basic normÐthat is, merits in the form of laudatory, albeit
unfulfilled, ambitions where the nature of law-creating facts is concerned.
What the complex social practice fails to accomplish in Hart's theory,48
namely to endow the fundamental rules of the system with binding force, is
accomplished in Kelsen's theory by a conferralÐand this, in turn, as Raz
points out, is the purported function of that most fundamental competence
norm, namely, the basic norm. Given the ``purity'' of the Pure Theory, the
basic norm is the only norm conferring validity on the positive law and on
nothing else (see Raz 1986, 95±6).
Kelsen, as Raz readily grants, was right to emphasize that the relation
between the ultimate rules of a system and their sources is different from the
relation between ordinary rules and their sources.
Thus, Raz's negative view of basic norms in general does not stand in the
way of a sustained examination of Kelsen's basic norm and, specifically, of
the question `Ìn what sense is the basic norm presupposed?''
Raz was prompted to consider this question by certain puzzles that arise
from Kelsen's views on the modality of legal statements, in particular, the
problematic posed by Kelsen's normative concept of legal validity, that is,
validity qua bindingness (Verbindlichkeit).49 In short, Raz asks how binding
force, the legal obligation of the subject, can be given a juridico-theoretical
explication without any appeal to either morality or the facts and without
employing Kantian or neo-Kantian doctrines either (see Paulson 1992b;
1998b, xxxvff.).
According to Raz, the ``purity'' of Kelsen's Pure Theory of Law is seriously
endangered by Kelsen's views on the nature of legal statements. A legal
theory, Raz argues, is ``pure'' only if the propositions formulated according
to it refer neither to moral nor to empirical facts. For Kelsen, however,
normative statements, be they legal or otherwise, serve to express a practical
47
Raz grants, nevertheless, that there is a single argument on behalf of basic norms, namely, the
argument that every legal system must contain at least one non-positive norm, which confers
legislative power on the historically first legislator. This argument however is not embraced by
Raz himself, for it is based on an assumption that he holds to be mistaken, namely, the assump-
tion that legislative power necessarily originates with a norm. According to Raz 1980, 138ff., laws
can indirectly authorize their own creation; see Peczenik 1995, 168; see also Spaak 1992, 190.
48
Hart's views on the validity of law do not comprise the descent of normative force; rather, they
make a halt at the notion of ``criterion of validity'' as a test of membership. As Raz points out,
the complex social practice in Hart's theory that ties the ultimate rules of a legal system to social
reality does indeed succeed in establishing the membership of certain rules in this system; it
does not, however, succeed in transmitting the binding force of the law from one norm to
another. For Hart, there is no difference between the role of legislative acts and that of social
practice; see Raz 1986, 95±6.
49
To say that a norm is valid, means ``that it is binding [verbindlich]Ðthat an individual ought to
behave in the manner determined by the norm'': Kelsen 1960, par. 34(a), 193; cfr. 1925, par. 19(c),
99. On validity as a normative concept, see Bulygin 1990; Nino 1978.

attitude; that is, they are used to express a belief in the existence of a valid
norm, and a norm constitutes a value (see Raz 1986, 86±8).
If the bindingness of a legal statement is owing to its use in stating the
existence of a value, then, Raz concludes, it must be the case that legal
statements are normative after all, normative in the same sense that
moral statements are normativeÐthat they are, in short, fully normative,
``moral'' statements.50 If so, how can they then be pure?
Kelsen, Raz contends, must have found himself in a dilemma: If, on the
one hand, legal statements are normative in the same way that moral state-
ments are, then they are moral statements, which would undermine the
``purity'' of Kelsen's theory. If, on the other hand, legal statements are not
moral statementsÐand this is Kelsen's viewÐthen they are not normative
either, which would in turn undermine the claim of legal science to be a
normative science (see Raz 1986, 85, 88; see also Paulson 1992b, 158). How,
Raz inquires, is Kelsen able to assert that a valid norm is a binding norm that
ought to be obeyed51 and to offer, at the same time, his theory as free of
moral import?
Raz, pondering this question, turns to a reconstruction of the basic norm.
His point is that there is a reading of Kelsen's theory that leads him out of
the dilemma. Reconstructing the doctrine of the basic norm and introducing
a normative component in the form of the legal man, both the legal scientist
and the legal practitioner presuppose the basic norm in a sense that is
uncommitted or is ``detached,'' respectively (see Raz 1986, 89±91). Specific-
ally, then, Raz imputes three moves to Kelsen.
The first move represents Kelsen's manifesto, namely the line to be drawn
between legal science and sociology. Establishing and describing the exist-
ence and content of law does not necessarily require the use of normative
terms; instead, it may be carried out in a wholly descriptive way. The
outcome will be a description of what Hart calls a complex social practice.52
This type of descriptionÐa sociological description of lawÐwill certainly be
free of moral implications; it cannot possibly claim, however, that the social
facts thus described should have normative consequences. Precisely that,
however, is what the Pure Theory claims, namely, to be in a position to
describe the law as a normative system regardless of content.
The Pure Theory, Raz contends, can make this claim thanks to Kelsen's
second move, namely, his introduction of the basic norm: The basic norm
stipulates that the constitution is valid and ought to be observed. But for
whom, one can now ask, is the basic norm valid? Who ought to observe the
constitution?
50
On Raz's interpretation of Kelsen's concept of normativity, see Wilson 1982, 50±3, 54±5.
51
This is what Raz terms Kelsen's ``semantic antireductivism'': Raz 1986, 81. The opposite is the
thesis that legal statements are purely descriptive.
52
``[_] en empirical, though complex, question of fact'': Hart 1961, 245; see Raz 1986, 88±9.

At this point, the legal man makes his entry. According to Raz, the basic
norm can only be said to be valid for the legal man (Raz 1986, 95), for it is he
alone who can be said to presuppose it.
The legal man is a person whose moral beliefs coincide with the law. He
sincerely believes in the validity of all those legal normsÐand those legal
norms aloneÐthat belong to his legal system. This belief is no coincidence;
rather, it is the logical consequence of his belief in the moral authority of the
supreme legislative body. The legal man whole-heartedly accepts the basic
norm and all the norms that are traceable back to itÐand he accepts nothing
else (see ibid., 89).
This second move ascribed by Raz to Kelsen is decisive for the nature of
the statements made by legal science: Whereas the legal man really believes
in the basic norm and presupposes it in the sense of positing it as a genuine
juridico-moral norm, legal science neither believes in it nor presupposes it.
Legal science, Raz says,53 proceeds on the basis of a hypothesis: If the basic
norm is valid, then _ Or on the basis of the hypothesis: If the legal man is
correct, then _
In this way the presupposition of the basic norm by legal science is
reduced to a basis purely hypothetical in nature, yielding such hypothetical
normative statements as: `Ìf the basic norm is validÐor if the legal man is
rightÐX ought _'' By means of statements that are value-neutral and
conditional the science of law is in a position to establish what the law is if it
is valid (see Raz 1986, 90; 1979, 142±3). This, as Raz points out (Raz 1986, 90),
may be all that is required of the science of law for it to fulfil its task.
More, however, is required of legal practitioners: They do not say what
the law is if it is validÐinstead, they say that it is valid. Their statements are
categorical. On this reading, what Kelsen needed was a way in which even
the categorical statements of the jurist could appear value-neutral.
Consequently, Raz ascribes to Kelsen a third move (see ibid.), which con-
sists in having even the legal practitioner presuppose the basic norm, albeit
in a special way.54
According to Raz, the legal practitioner, by means of detached legal state-
ments (ibid., 91), speaks of the law as a valid system of norms. ``Detached''
legal statements are categorical statements made from a point of viewÐnot
the jurist's personal point of view but, rather, that of the legal man. The
jurist pretends, as it were, to embrace the view of the legal man. Raz
contends that this pretence represents what Kelsen meant by saying that the
jurist's presupposition of the basic norm amounted to a fiction (see ibid.),55
concluding that the so-called presupposition is no presupposition at all but,
53
Raz 1979, 142 prefers `àdopt'' to ``presuppose.''
54
To this special way, Raz contends, Kelsen was referring when saying that the basic norm was
being presupposed in the juristic consciousness: See Raz 1986, 90, referring to Kelsen 1945a, 116.
55
In my view, the legitimacy of using the term ``fiction'' in the present context may be called into
doubt since this term points to Kelsen's post-1960 phase.

rather, a ``professional and uncommitted adoption'' (Raz 1979, 143). To

understand both legal scientists and legal practitioners as ``presupposing''
the basic norm was in Raz's view the only option that offered itself to Kelsen
in his zeal to avoid ascribing a moral dimension to the law.
Without doubt, Raz's reconstruction of the doctrine of the basic norm is a
genuinely meritorious attempt to elucidate the significance of the doctrine
for Kelsen's conception of legal normativity, and there is much to learn from
Raz's insight into Kelsen's thinking.56 In my view, however, it is not so easy
to endorse the conclusion Raz draws from his doctrine, namely, that the
presupposition of the basic norm is identical with, as he says, the `ùn-
committed adoption'' of a moral attitude on the part of the jurist. It seems to
me that this adoption implies a necessary tie between law and morality
while nevertheless totally disregarding the content of law (see Paulson
1992b, 156).57
This issue aside, how can one adopt something in an `ùncommitted'' or
``detached'' way? Raz means, I believe, that both the legal scientist and the
practitioner are aware that for the sake of (professional) appearances, they
must pretend that law really is binding, and Raz contends that this pretence
makes it possible for them to use normative terms with a clear conscience.
However, this pretence (or ``detached'' adoption of the basic norm) only
seems to produce statements the modality of which is essentially unclearÐ
for what, exactly, does ``detached'' normativity mean?
In short, I am not convinced that the doctrine of the legal man can do
justice to Kelsen's point about having the basic norm presupposedÐnamely,
to underpin legal argumentation with a Grund or basis that would initiate
and transmit normativity without reference to morality, be it ``detached'' or
otherwise.58
III. What Difference Does the Presupposition Make?

Section II has shown that it is indeed possible to explain the presupposition
of the basic norm. While Aleksander Peczenik's explanation, based on his
theory of the basic norm-transformation, leads, as he puts it, to legally nor-
mative statements, Joseph Raz's explanation, based on his doctrine of the
legal man, leads to so-called ``detached'' legal statements.
How then, one asks, does Kelsen himself approach the issue at stake? And
what are his views on the character of the legal language?
56
For example, his observing the `òbscure'' distinction, as he puts it, that Kelsen makes between
positing and presupposing a norm: See Raz 1986, 90.
57
Besides being sceptical with regard to the type of legal normativity implied by the doctrine of
the legal man, Paulson (1992b, 159±61) points at a defect of the doctrine as such, namely, of being
inconsistent with the sources thesis.
58
On the issue of the type of legal normativity implied by Raz's doctrine of the legal man, see
Paulson 1992b, 156; Vernengo 1986; Wilson 1982.

When Kelsen, during his classical phase,59 speaks about how to achieve
a ``reconstruction'' of the legal normÐthat is, a recasting of the authentic
legal material that would reflect the peculiarly ``legal'' relation of impu-
tationÐhe is seen as using a variety of terms. Among these the most
frequent are Rechtssatz (to be distinguished from the technical notion asso-
ciated with this term), Sollsatz (`òught''-sentence), Soll-Urteil (`òught''-
judgement), andÐbeginning in about 1945Ð``rules of law in a descriptive
sense.''
The most intriguing of the above-named terms is Sollsatz. SollsaÈtze contain
an `òught,'' that is, a norm. They may do so either prescriptively (that is,
with imperative force) or ``descriptively.'' The linguistic form of the expres-
sions used may be exactly the same in the prescriptive and ``descriptive''
casesÐit is the use of the formulation that determines whether we are
expressing a norm or making a statement about a norm.
Consider the following example: ``You ought to park your car on this side
of the road.''60 On one interpretation, we are confronted with a norm-
formulation that is being used to issue a command; thus, `òught'' is
expressed prescriptively. On a second interpretation, we are confronted
with a norm-formulation that is being used to make a (true or false) state-
ment about a command; thus, `òught'' in this instance is being used
``descriptively.''61
In the following, I shall first briefly outline Kelsen's terminology and then
attempt to shed light on his notion of the ``prescriptive'' and ``descriptive''
readings of the legal `òught,'' drawing on the interpretations of Martin
Golding and Herbert Hart.
III.1. Kelsen's Terminology

Notwithstanding the fact that Kelsen kept on emphasizing the difference
between the function of the legal scientist and the legislator,62 he failed well
into the middle of his classical phase, as he himself grants (Kelsen 1960
(RR 2), par. 18, 83; see Weinberger 1982, 111), to mark the distinction
between legal norm and legal statement (see, for example, Kelsen 1911,
255f.; and 1925, par. 10, 47). None of the first three major worksÐnamely,
59
With respect to the periodization of Kelsen's work, I follow Paulson 1998a: See note 4
above.
60
This example is a slightly altered version of von Wright's (1963, 104f.). On the difference
between norm and proposition, see generally Bulygin 1982, 127±52; Paulson 1992a, 270±3; Raz
1979, 146±59; von Wright 1963, 93±106.
61
On ``descriptive,'' see Paulson 1992a, 265±74, 271.
62
`Ìn characterizing the law as norm, and in restricting legal science to the cognition of norms (a
function different from both making and applying the law), one separates the law from nature,
one separates legal science as a cognitive science of norms from all those cognitive sciences that
aim to explain natural events in terms of causal laws'': Kelsen 1934, par. 7, 13.

Hauptprobleme der Staatsrechtslehre (1911),63 Allgemeine Staatslehre (1925),64 and

Reine Rechtslehre (1934)Ðreflects any awareness on Kelsen's part that norm-
formulations may be used either way, that is, either to issue or to describe a
norm.
Although the systematic distinction between legal norm and legal state-
ment dates from the 1940s (see Kelsen 1941±1942),65 Kelsen does not set it
out in a fully systematic way until the Second Edition of Reine Rechtslehre
(Kelsen 1960, par. 16, 71ff.): Whereas Rechtsnorm is a meaning-content,
prescribing an `òught,'' (ibid., 71, 74) Rechtssatz is the (hypothetical) state-
ment formulated by legal science, describing an `òught.'' The hypothetically
formulated legal proposition corresponds to the (hypothetically formulated)
legal norm (ibid., 71; par. 18, 78±81).
Rechtsnormen areÐ`ìhrem Sinne nach,'' as Kelsen puts itÐprescriptions,66
issued by legal authorities, imposing duties upon and granting rights to the
legal subject (see ibid., 73). Rechtsnormen lack truth values; rather, they are
either valid or invalid (see ibid.).
Not so RechtssaÈtze in Kelsen's technical sense. Notwithstanding the
normative terms found in them, RechtssaÈtze are theoretical sentences: They
are, as Kelsen says, `òught''- judgements (Soll-Urteile) describing norms and,
as such, are either true or false just as `ìs''- judgements (Seins-Urteile) are
either true or false (see ibid.).
RechtssaÈtze are in no way binding. Instead, they provide information about
somethingÐnamely, about the existence of an `òught.'' And they do so,
according to Kelsen, by rendering the specific sense (``the specific meaning'':
ibid., par 18, 78) in which the legal norm is directed towards both legal
organ and legal subject; this being so, the Rechtssatz cannot be formulated
otherwise than as a hypothetical judgement; it states that, in accordance
with a given positive legal system and under certain conditions, a certain
(legal) consequence ought to take place (see ibid.).67
63
The term ``Rechtssatz,'' in Kelsen 1911, 70, 254, 385, is problematic because it is used in a
variety of ways, meaningÐinter aliaÐ``hypothetical judgement''; see also Kelsen 1928, 335.
Moreover, Kelsen's use of the term ``judgement'' is ambiguous, at least well into the first half of
the so-called classical phase. Paulson distinguishes no fewer than five different meanings of
Urteil in Kelsen's early work, namely (1) the hypothetically formulated norm; (2) the later
reading of Rechtssatz, i.e., the legal proposition; (3) the norm-formulation; (4) the judge's
decision; and (5) claims or statements in non-normative contexts. See Paulson 1998a, 163, n. 50.
64
Rechtsnormen oder RechtssaÈtze: Kelsen 1925, par. 10, 47.
65
However, Kelsen's awareness of the connection between legal propositions and the distinct
functions of the legal scientist and the legislator datesÐor so he would claim laterÐfrom early
times: Kelsen 1960 (RR 2), par. 18, 83; see also Kelsen 1919.
66
Prescriptions include commands and imperatives but even permissions and authorizations;
not in any circumstance, however, are they `ìnstructions'': Kelsen 1960, par. 16, 71.
67
According to the Pure Theory, the basic form of the rule of law (Grundform des Rechtssatzes) is
the following: `Ùnder conditions determined by the legal order a coercive act, determined by
the legal order, ought to take place'': ibid., 76; or, in other words: Given certain conditions, such
and such ought to take place.

The statements formulated by legal science are, Kelsen believes, `òught''-

statements (Soll-SaÈtze) and, indeed, must be `òught''-statements, for they
describe `òught''-norms (Soll-Normen; see ibid., par. 16, 74). Since, however,
`òught''-statements reproduce the `òught'' of legal norms descriptively (see
Kelsen 1945, 163), the `òught'' of the statements formulated by legal science
is exclusively descriptive (see Kelsen 1960, par. 18, 79).
At the beginning of the 1960s, Kelsen did not deem the linguistic
expression of the legal norm to be all that important. What really counted,
he thought, was the Sinn, as he puts it, of the act by means of which the
norm is issuedÐthe process of issuanceÐas distinct from the Sinn of the
product of issuance, in short, the norm issued.68 The outcome, he says, is a
``description'' of the normÐa description, however, in which the `òught''
has a quality that is, logically, different from the `òught'' of the norm.69
The `òught,'' Kelsen (1960, par. 18, 79±80) explains briefly, is the specific
Sinn in which condition and (legal) consequence are linked together by
means of the norm. This Sinn of the `òught'' is expressed by means of
the legal statementÐexactly how, Kelsen was unable to explain, lacking as
he did an adequate theory of meaning. However, that he insists on two
forms of the Sinn des Sollens suggests that he may well have been on the right
track.70
Kelsen's terminology and the question of appropriate renditions in English
remain a matter of considerable controversy (see for instance Raz 1986;
Paulson 1992b; Vernengo 1986). A related concern is the expressions Kelsen
employed in explaining his views. I turn briefly to two attempts at inter-
pretation, namely Martin Golding's and Herbert Hart's.
III.2. A Descriptive Mention of the `Òught'' (M. Golding)

As Kelsen (1945a) says, the function of the (hypothetical) `òught''-statements
of the science of law consists in ``descriptively reproducing'' the `òught'' of
the norms.71 In Anders Wedberg's translation, these statements are called
``legal rules'' or ``rules of law'' (Kelsen 1945a; 45)72; the original German used
by Kelsen is unavailable. In published writings, the technical term Rechtssatz
does not appear before 1953 (Kelsen 1953; Klecatsky et al. 1968, 612).
68
`Ìmportant, however, is not the linguistic form, but the meaning of the law-creating, norm-
positing act. The meaning of the act is different from the meaning of the law-describing rule of
law'': ibid., par. 16, 72.
69
At that time, Kelsen believed the distinction between these two kinds of `òught'' to be logical:
See Kelsen 1960 (RR 2), par. 18, 82.
70
Sollen is normally expressed by an imperative; see also Kelsen 1960 (RR 2), par. 16, 77f.,
referring to Christoph Sigwart.
71
``The ought-statements in which the theorist of law represents the norms have merely de-
scriptive import; they, as it were, descriptively reproduce the òught' of the norms'': Kelsen 1945a,
163; my emphasis; see also ibid., 45, 61.
72
This translation has been called into doubt: See Bulygin 1990, 35.

What Kelsen may have meant by ``rules of law in a descriptive sense''73

has been discussed at length (see, for example, Ofstad 1950). Of special
significance in this connection are Kelsen's expressions for the task of the
legal scientist and the function of legal statementsÐ``representation'' of the
positive law and ``descriptive reproduction'' of the `òught'' of the legal
norm, respectively. Although these expressions have been neglected by
most commentators, Golding (1971) is a noteworthy exception.
Golding (1971, 82ff., 90±1), though seriously critical of Kelsen's theory, is
one of the few scholars for whom Kelsen's mode of expression apparently
does not pose a problem. In Golding's opinion, the puzzle of how to
interpret the so-called ``descriptive òught' '' is tied to another question,
namely the status of the rational reconstruction of a normative system by
comparison with the original materials, that is, authoritative legal utterances
(see ibid., 77ff.). Golding suggests that there might have been less confusion
surrounding Kelsen's ``legal rules in a descriptive sense'' had he clearly
distinguished, as Golding puts it, between `ùsing and mentioning legal
utterances'' (ibid., 78; see also Paulson 1995, 130ff.): While legal utterances
under the appropriate conditions may count as issuance of a legal norm,
purporting to influence human behaviour, they may also occur embedded
in a formulation that might be understood as a statement in which the norm
itselfÐor, rather, the norm-formulationÐis mentioned and not used and, in
that case, contains more than was expressed by the original utterance.
Golding's interpretation stems from an insight into the purpose of the
Pure Theory, an insight that he acquired in the course of assessing Kelsen's
notion of ``legal system.''74 He argued that the key to the problem lay with
the concept of ``rational reconstruction.'' (Golding 1971, 80).75 Specifically,
Kelsen wanted to show how the rational reconstruction of a normative
system is to be carried out (see ibid., 71±3), and Golding suggests that the
English expression ``representation,'' appearing quite frequently in Kelsen's
writings,76 is best interpreted as a reference to the term of ``rational recon-
struction'' (ibid., 73, 75).
The major point of Kelsen's theory, as Golding understands it, was the elab-
oration of a conceptual apparatus that would provide for the construction
73
`Ìt is of importance that the term `legal rule' or `rule of law' be employed here in a descriptive
sense'': Kelsen 1945a, 45.
74
When Golding wrote Kelsen and the Concept of ``Legal System'', he did not yet have access to
Kelsen 1960.
75
Golding emphasizes the influence of logical positivism of the Vienna Circle (Wiener Kreis) on
Kelsen. The rational reconstruction of a certain branch of science means that the branch is being
reconstructed on the basis of rational principles. The scientific material as such is systematized
and explained by means of both accurate definitions and a reduction of the number of
undefined terms. This procedure is held to be justified by the obvious need for a conceptual
apparatus as it not infrequently happens that branches of science are developing in an
unsystematic and even casual way, without the scientists' being aware of their own pre-
suppositions or of the actual scope of their results; see Golding 1971, 71±3.
76
Golding 1971 quotes Kelsen 1945a, 45, 61, 163.

of a ``proper representation'' of legal norms. The Pure Theory, Golding

(1971, 73, 78) contends, is best seen as a method for carrying out a rational
reconstruction of any legal system, whatever its content.
As to the main function of a rational reconstruction, Golding claims that in
focusing on the linguistic expression of norms and in developing a kind of
standard formula corresponding to this (linguistic) expression,77 the Pure
Theory provides a standard means78 of formulating legal norms, namely, a
hypothetical construction, where the consequent clause, stipulating the
sanction, contains an `òught'' (see ibid., 88±9).
As to the status of a rational reconstruction, Golding distinguishes a
reconstruction from simple reproduction. It cannot possibly be the purpose
of a rational reconstruction, he contends, to produce a mere copy of the
original material (see ibid., 79). What is more, to produce such a copy would
not even be feasible, for there is no `òne-to-one correspondence'' (ibid.)
between the utterances of the law-creating agencies and the elements of a
rational reconstruction. And who would settle for a mere copy if the genuine
article were available? Golding (1971, 80) concludes that legal statements,
``reproducing'' in a descriptive sense the `òught'' of the legal norm, attain far
more than a simple reproduction of the norm (see Kelsen 1945, 163).
What sort of status does the descriptive reproduction of legal science
have by comparison with the legal norm? What is the difference between
the two? The difference, Golding says, is to be found both in the manner
of formulating legal norms and in the fact that the rational reconstruction
will contain more than the original material, that is, more in the sense of
including the norms assumed by officialsÐin particular, the basic norm (see
Golding 1971, 79f.; Kelsen 1945a, 156, 161f.).
Actually, Golding contends, it is quite simple to understand what Kelsen
meant by ``descriptive òught' ''Ðone need only to think of the difference
between using words or expressions and mentioning them. He explains his
meaning in the following way.
In issuing a legal norm with the words: `Ìf the situation S obtains, then
official A ought to direct that sanctions be applied'' a legal official is in fact
using the norm. In rendering what the official has done with the words: `` Ìf
the situation S obtains, then official A ought to direct that sanctions be
applied' is California law'' a legal scientist is, according to Golding (1971,
78), mentioning this norm.
Following the above scheme, one might then agree with Hart and argue
that all Kelsen meant by ``rules of law in a descriptive sense'' is that
statements describing the meaning of a legislative enactment will mention
certain SollsaÈtze or `òught''-statements as the equivalent in meaning of the
enactment (see Hart 1963, 714).
77
``[_] so that they wear their legal character on their faces, so to speak'': Golding 1971, 75.
78
``[T]he canonical form of legal norms'': ibid., 89.

Seen generally, statements in which words or expressions are mentioned

may be called paraphrases by contrast with statements in which words or
expressions are used. Let us consider, in this connection, the following
example.
If we exclaim ``Careful! That dog bites!'' we use the word ``dog''; our
intention is to warn somebody to stay clear, rather than, say, explaining the
meaning of the word. If, however, we utter the sentence `` `Dog' is a three-
letter word,'' we are then mentioning the word ``dog''Ðwith the intention of
giving some information about it.79
To conclude: Golding's interpretation of the so-called ``descriptive
òught' '' by way of a rational reconstruction can be appreciated in the light
of Kelsen's programme of raising the science of law to the level of a genuine
science. Seen within this context, Kelsen's term ``representation'' seems
appropriate; its connotations are richer than those of ``reproduction'' and,
indeed, it can be understood as a rough synonym for ``reconstruction.''
As has been shown above, Golding interprets the Rechtssatz as mentioning
the legal norm. Although this view has its followers,80 Kelsen's own view of
the matter raises doubts. For, as Hart (1963, 714f.) argues,81 Kelsen at their
debate in Berkeley vigorously repudiated the idea of legal statements as
paraphrases or second-order statements about the law whereby words are
merely mentioned and not used. Instead, Kelsen adhered to his termin-
ologyÐ``rules of law (or òught'-statements) in the descriptive sense.''
III.3. A Descriptive Use of the `Òught'' (H. L. A. Hart)

Kelsen's apparent shift on the normative nature of legal science, implied by
his speaking of ``rules of law in a descriptive sense,'' left Hart puzzledÐfor
if rules of law are ``descriptive,'' they are then evidently not normative (see
Hart 1963).
Legal statements qua ``legal rules employed in a descriptive sense''
constituted a major issue in the debate between Kelsen and Hart, held in
Berkeley in November 1961.82 As mentioned above, Kelsen at the debate
positively rejected Golding's interpretation, which Hart (1963, 713±15) had
set out in some detail. After the debate, Hart wrote Kelsen Visited, where
he presented a different explanation of Kelsen's mode of expression (see
ibid., 715).
79
See Herbert Hart's example in Hart 1963, 714.
80
`Ìn a Rechtssatz, therefore, we could say that norms are mentioned, or referred to, by the same
linguistic sentence that is used to apply them. Norms, we can say, appear in the RechtssaÈtze
obliquely'': Vernengo 1986, 101; see also ibid., n. 9.
81
Hart 1963 constitutes a record of the public encounter between Kelsen and Hart.
82
The other two major issues were Kelsen's view of legal definitions and the relationship
existing between positive law and morality. On the exchange between Golding and Hart, see
Paulson 1995, 130±5.

Hart's attempt to justify Kelsen's terminology is original. He draws

attention to an analogy between Kelsen's ``descriptive òught' '' and an
interpreter's translation of an order: The translation, he says, is a speech act,
which, as such, may be called an `òrder in the descriptive sense'' (ibid., 716).
In order to render this peculiarity of legal language, one must first examine
the relationship existing between law and legal statements. In Hart's view
this relationship is comparable to that existing between a person speaking a
foreign language and his English-speaking interpreter.
Hart illustrates the point by introducing the commander of a German
prison camp. The commander has to rely on an interpreter in order to make
himself understood by the English-speaking prisoners. Hart now asks how
the interpreter's yelling ``Stand up!'' ought to be understood by contrast with
the original German order ``Stehen Sie auf!''. What kind of statement is the
interpreter's speech act? It is certainly not a paraphrase that mentions the
original order: There is no explanation here to the effect that the German
words mean the same as ``Stand up!'', and it is not an order either. The inter-
preter lacks competence to issue orders; he is there simply to translate them.
Hart arrives at the conclusion that the interpreter's employment of the
imperative mode must be purely descriptive. The problem, however, is that
an intelligent interpreter, in Hart's view, is doing far more than merely
using the imperative mood descriptively: An interpreter, Hart says, will give
some sort of explanation of the original order and, if need be, sum up its
essentialsÐin short, he turns the original order into a rational reconstruction
(ibid.). What is more, the speech act is a reproduction of something: Since it is
clear to the prisoners that what is in question is an order, it is equally clear
that the interpreter was successful in reproducing it.
Thus, Hart's interpretation, like Golding's, is based on the idea of rational
reconstruction. If one compares Hart's interpretation with Golding's, how-
ever, it becomes clear that Hart is far more to the point in focusing upon what
Kelsen (1960, par. 18, 78) in fact declared to be the main task of legal science,
namely, to render the specific Sinn of legal norms. Hart, in contrast to
Golding, emphasizes that in so far as Kelsen's own view of legal statements is
concerned, it is a question of a special use of language, not merely its mention.
All considered, Hart thinks that the interpreter's utterance might reasonably
be called `àn order in the descriptive sense'' (Hart 1963, 715f.). By adducing
the example of the interpreter, Hart wants to show that the ``representation''
of law as seen by Kelsen is identical neither with statements about the mean-
ing of legal rules nor with paraphrases mentioning the `òught'' but not using
it. By calling legal language `à specific kind of use of language'' (ibid., 717),
one comes to appreciate that Kelsen was right after all.
Thus, while Golding contents himself with explaining Kelsen's ``rules of
law in a descriptive sense,'' putting them on a par with second-order
statements about the law in which words are merely mentioned, Hart arrives
at the conclusion that this view cannot do justice to what Kelsen meant by
``representation'' of the specific Sinn of the legal normÐnamely, a rendition

of the latter via a specific use of language. Hart's example of the interpreter is
significant in underscoring this point.
At first glance, Hart's analogy strikes one as genuinely illuminating. It
has, however, been termed inadequate. In his monograph on the legal
theories of Kelsen and Hart, Pawlik (1993, 189±91) argues that an inter-
preter's speech acts are not to be compared with legal statements as Kelsen
understands them. Focusing on the ``constitutive'' role of legal science in
Kelsen's theory, Pawlik (ibid.) contends that although the interpreter's
speech act adds nothing to the original material (see ibid., 190), the legal
scientist's interpretation gives rise to the quality of legal norms qua `òught''-
norms. The point of Pawlik's criticism is that this interpretation is made
possible, according to Kelsen, solely by means of an assumptionÐthat the
specific Sinn of law, that is, its bindingness, arises owing to the presup-
position of the basic norm (see ibid., 191).
To summarize: The issue of the present sectionÐwhat difference
presupposing the basic norm makesÐwas seen to turn on Kelsen's views
of the legal language and, specifically, his notion of the appropriate reading
of the legal `òught.'' This notion only emerges from a larger context.
Kelsen's views of the character of legal language date back to that phase in
his work where he aimed at a ``reconstruction'' of the legal norm, that is, a
recasting of the authentic legal material that would reflect the peculiarly legal
relation of imputation. This ``reconstruction'' Kelsen attempted by means of Ð
so-calledÐ`òught''-sentences. The specifically legal `òught,'' it would appear,
necessarily comprehends both ``prescriptive'' and ``descriptive'' elements.
It seems that the presupposition of the basic norm was intended, by
Kelsen, to evoke a specifically ``legal'' use of language. Since the linguistic
form of norm-formulations (used to issue norms) and of norm-statements
(used to ``describe'' norms) may well be exactly the same, the difference
between norm-formulation and norm-statement must lie in the illocutionary
force of the respective utterances.83
IV. Concluding Remarks

The issue of what it means to presuppose the basic norm, I am suggesting,
does not lend itself to a single, summary solution. Rather, it must be viewed
from different angles. This, in sections I±III, is what I have argued.
Section I was devoted to Kelsen's views on the addressee of the basic
norm. His statements in this connection, between 1914 and 1960, reveal an
attitude of considerable uncertainty about whether or not a presupposition
83
Any attempt to establish the nature of a specifically ``legal'' use of language must, then,
include proper attention to the role of the speaker. Evidently, the proper way to proceed, at this
point, would be to appeal to a theory of meaning. This, however, would reach well beyond the
scope of the present paper.

can be attributed to legal science. In my view, his ambivalence here leads

straightaway to one key to a solution, namely, to the meaning(s) of the verb
``to presuppose.''
In section II, it was shown that it is indeed possible to explain the
presupposition. While Aleksander Peczenik explains it as linking observable
facts and criteria to the validity of the system as such, giving rise to ``legally
normative'' statements, Joseph Raz interprets it as being an uncommitted
and professional `àdoption,'' as he puts it, of a virtually moral norm, the
outcome being ``detached'' legal statements.
I turned, in section III, to the Kelsenian legal proposition and to some of
the expressions Kelsen used to describe it. I focused on ``rules of law in a
descriptive sense'' and adduced two interpretations of the ``descriptive
òught','' namely, Martin Golding's and Herbert Hart's. It has emerged that
the ``presupposing'' of the basic norm must have been intended, by Kelsen,
namely, to evoke a specifically legal use of language.
In closing, I might recall to mind the questions I set out in the introduction:
1. Who presupposes the basic norm?
2. Is it possible to defend the presupposition of the basic norm in a way
that is convincing?
3. Finally, what difference does the presupposition make?
It has been shown that Kelsen was less than clear on the question of who
presupposes the basic norm (1), that the presupposition, on the part of the
jurists in general, is according to the received opinion both natural and
more or less explicable (2), and that the result of the presupposition, namely
the Kelsenian Sollsatz, reflects a peculiar use of language (3). Taken indi-
vidually, the respective answers to the three questions I posed at the outset
do not at first glance seem to have all that much in common. In fact,
however, they have a common denominator.
Specifically, I contend that the verb ``to presuppose'' has not only a
conceptual dimension but also a normative dimension, thus rendering the
doctrine of the basic norm applicable to all jurists, and that the expression
``presupposing the basic norm'' is adequate in the sense that it marks the
fused,84 that is, descriptive-normative nature of utterances made in what
may be called specifically ``legal'' speech-situations.
84
I am referring to the term ``fusion''Ða term coined by Eng 1998. What is ``fusing,'' is a
descriptive and a normative component of a categorical legal statementÐnamely, and on the
one hand, the jurist's view as to what his colleagues are likely to think the law will be in a
certain case (descriptive component) and, on the other, the jurist's own view as to what the law
ought to be in this case (normative component). According to Eng, the descriptive and the
normative components are so closely intertwined as to make it virtually impossible to
distinguish between them psychologically. The intertwining, or ``fusing,'' of the descriptive and
the normative component will result in a sentence the modality of which (``fused modality'')
expresses the jurist's view as to what, in the case in question, the law presumably is. I would
like to thank Professor Eng very warmly indeed for being so kindÐand at a time of great
stressÐas to discuss with me some of the salient points of his theory of ``fused modality.''

In the Pure Theory of Law, the basic norm is called the ultimate ``Grund''
as well as the ``BegruÈndung'' of the validity of positive law. This choice of
terms may strike one as unfortunate, but I want to suggest that it is in fact
highly instructive, throwing light on Kelsen's train of thought.85 Equally
instructive, I believe, is his choice of expression with respect to the basic
norm.
As I suggested above, one key to the issue of this paper is the verb ``to
presuppose.'' The other key, closely tied to the first, is the propensity of
jurists to use normative language. In Kelsen's view, the basic norm is pre-
supposed unconsciously in ``juristic'' thinking.86 Had he chosen to elaborate
on these words, he might have said something like the following. Notwith-
standing different roles and functions, jurists are participating in an activity
that is institutional and, consequently, underpinned by constitutive rules. By
invoking these rules, jurists necessarily commit themselves to them. In legal
reasoning, there is a rule to the effect that law is, objectively, valid and bind-
ing. Jurists are seen as accepting this rule. To be sure, they do not have to.
However, there will be such a thing as (descriptive) legal cognition as well
as juridico-normative bindingness (Verbindlichkeit) if (and only if ) they do.
As shown above, Kelsen's view on whether or not legal science can be
said to presuppose the basic norm was far from settled during a long
phase in his work. From the contexts in which ``presuppose'' is used, it may
be inferred that this expression, in connection with the basic norm, exhibits
two distinct senses.
Used in the one sense, ``presuppose'' (voraussetzen) means `àssume
(annehmen) epistemologically.'' What is being assumed, on the part of the
legal scientist, is a fundamental, ultimate norm. Owing to the merely
hypothetical nature of this norm, the act of assuming (or presupposing) is
entirely void of any normative import.
Used in the other sense, ``presuppose'' means, quite literally, ``to posit
(a norm) at an earlier stage'', to ``pre-posit (a norm).'' This meaning of
``presuppose'' has been coined, as it were, by Kelsen himselfÐcuriously
enough, the exact rendition of the German ``Voraus-Setzung'' (Kelsen 1934,
par. 31(f), 83) is ``preposition'' and not ``presupposition.'' That which is being
pre-supposed (or pre-posited) is the same ultimate and hypothetical normÐ
with, however, the important difference that the ``pre-supposing'' is carried
out not by the legal scientist but by the legal official, that is, by the one who
actually ``posits'' legal norms (in applying them). Thus, to presuppose the
basic norm in norm-positing contexts has, inevitably and unavoidably, a
normative dimension. Or so Kelsen would have us believe.
85
What I am suggesting is that Kelsen's use of ``Grund'' and ``BegruÈndung'' vis-aÂ-vis the objective
validity of law may be clarified by paying close attention to the etymology of those words.
86
Here, it is of the utmost importance to consider the German text: While Kelsen 1960 (RR 2),
par. 34(d), 209, speaks of `àlle Juristen,'' the translation mentions ``most legal scientists''; see
Kelsen 1960, par. 34(d), 204.

During the earlier years of the so-called transcendental or classical phase in

Kelsen's work, the basic norm is presupposed in the former sense, and by legal
science alone. This species of presupposition is tantamount to an epistemo-
logical assumption, the outcome being purely descriptive legal statements.
During the later years of the transcendental phase, however, the basic norm is
evidently being presupposed, by legal organs, in the latter senseÐa sense tied
to the very act of ``positing'' norms. This latter species of presupposition
serves to explain the (fully normative) legal utterances of the law-maker.
Thus, Kelsen's perplexities: Having the Rechtswissenschaftler presuppose
the basic norm in this latter, normative sense would imply that legal science
was usurping the task of the legislator.
Nevertheless, as Kelsen declared in 1960, legal science can very well be
said to presuppose the basic norm, for the basic norm constitutes the Sinn
not of an act of will but of thinking. This declaration of Kelsen's points to the
crux of the entire matter, for this declaration merely gives expression to his
initial viewÐthe basic norm qua epistemological assumptionÐand appar-
ently ignores the task of those actually engaged in applying the law. Only
apparently, howeverÐfor legal organs are still tacitly understood to be
``presupposing'' normatively.
Thus, and on the face of it, the expression ``presupposing the basic norm''
seems to be unfortunate. Nevertheless, Kelsen retained it. The explanation
may lie in the fact, simple enough, that he saw no other expression that
would better capture his purpose vis-aÁ-vis the basic norm. In my view,
however, his terminological bind can be accounted for in a different way.
There are, in the Pure Theory of Law, indicationsÐthe most conspicuous
of them being Kelsen's emphasis on the Standpunkt des Rechts (Kelsen 1960
(RR 2), par. 14, 72)Ðthat the doctrine of the basic norm was meant to apply
to both legal scientists and legal practitioners. Now a term used to refer to
distinct legal roles may well be systematically ambiguous. I have suggested
elsewhere that Herbert Hart may have consciously employed an ambiguous
verb, namely ``to recognize,'' in order to bridge the gap between the cog-
nitive and the normative.87 To my mind it is highly likely that `` `to pre-
suppose' the basic norm'' represents a similar case, with latent connotations
of the expression making it possible for Kelsen's theory to be applied even to
a category such as jurists, whose task lies somewhere between positing and
tacitly assuming norms and whose statements have, accordingly, a modality
that is not easily defined.
Notwithstanding the peculiar connection existing between the role of the
speaker and his respective way of presupposing the basic norm,88 it is, I
87
Here I refer to the chapter on Hart's doctrine of the Rule of Recognition in Bindreiter 2000.
88
This is, albeit indirectly, hinted at by Kelsen himself in his reply to Julius Stone (Kelsen 1964±
1965). In this paper, Kelsen contends that it is presupposed, in juristic thinking, as a norm, that
the basic norm, if it is presupposed, is at the apex of the norm hierarchy of any legal system
(ibid., 1141).

think, impossible to point to specific situations in which the Kelsenian legal

scientist would feel prompted to presuppose the basic norm normatively.
However, to presuppose itÐthat is, to presuppose it in a descriptive and a
normative senseÐqua point of departure within legal argumentation, is a
conditio sine qua non for ``legal'' knowledge and ``legal'' normativity. It would
be impossible to argue professionallyÐ``legally''Ðwithout tacitly regarding
the said rule as a (conditional) normative Grund: The legal `òught'' requires a
Grund-Norm.
It is not unreasonable, in my view, to assert that all jurists are uncon-
sciously employing Kelsen's basic norm all the time and in two different
waysÐon the one hand, in taking for granted that law is considered valid
and binding by the other members of their legal audience, and, on the other,
in being committed to this view themselves. In the first case, the basic norm
can be said to be assumed (or presupposed in a non-committal sense); in the
second, it can be said to be genuinely presupposed, that is, qua valid norm.
Indeed, legal reasoning is special in the sense that the jurist is seen to
move, unconsciously, from the descriptive to the normative, his utterances
thereby evoking a modality that Kelsen was unable to explain. It is safe
to say that it would not be feasible to reconstruct the legal language
rationally without ascribing to the jurist all of what goes under the aforesaid
`ùnconsciously.''
Characterizing the jurist's categorical statements as ``legal rules in a
descriptive sense,'' as Kelsen did, is, then, not so very far-fetched after all.
University of Lund
Faculty of Law
P.O. Box 207
SE-22100 LUND
Sweden
References
Aarnio, Aulis, Robert Alexy, and Aleksander Peczenik. 1981. The Foundation of Legal
Reasoning. Rechtstheorie 12: 133±58.
Bindreiter, Uta U. 2000. Why Grundnorm? A Treatise on the Implications of Kelsen's
Doctrine. Faculty of Law, University of Lund.
Bulygin, Eugenio. 1982. Norms, Normative Propositions, and Legal Statements. In
Contemporary Philosophy. A New Survey, vol. 3. Ed. Guttorm Flùistad, 127±52. The
Hague: Nijhoff.
ÐÐÐ . 1990. An Antinomy in Kelsen's Pure Theory of Law. Ratio Juris 3: 29±45.
Eng, Svein. 1998. U/enighetsanalyseÐmed saerlig sikte paÊ jus ùg allmenn rettsteori. Oslo:
Universitetsforlaget.
Golding, Martin P. 1971. Kelsen and the Concept of ``Legal System.'' In More Essays in
Legal Philosophy. Ed. Robert S. Summers, 69±100. Oxford: Blackwell. (1st ed. 1961.)
Hart, H. L. A. 1961. The Concept of Law. Oxford: Clarendon.
ÐÐÐ . 1963. Kelsen Visited. UCLA Law Review 10: 709±28.

Kelsen, Hans. 1911. Hauptprobleme der Staatsrechtslehre entwickelt aus der Lehre vom
Rechtssatze. TuÈbingen: Mohr.
ÐÐÐ . 1914. Reichsgesetz und Landesgesetz nach oÈsterreichischer Verfassung.
Archiv des oÈffentlichen Rechts 32: 202±45, 390±438.
ÐÐÐ . 1919. Zur Theorie der juristischen Fiktionen. Mit besonderer BeruÈcksichti-
gung von Vaihingers Philosophie des Als Ob. Annalen der Philosophie 1: 630±58.
ÐÐÐ . 1920. Das Problem der SouveraÈnitaÈt. TuÈbingen: Mohr.
ÐÐÐ . 1923. Hauptprobleme der Staatsrechtslehre entwickelt aus der Lehre vom
Rechtssatze. 2nd ed. with a new Foreword. TuÈbingen: Mohr.
ÐÐÐ . 1923±24. Die Lehre von den drei Gewalten oder Funktionen des Staates.
Archiv fuÈr Rechts- und Wirtschafts-philosophie 17: 374±408.
ÐÐÐ . 1925. Allgemeine Staatslehre. Berlin: Springer.
ÐÐÐ . 1928. Die philosophischen Grundlagen der Naturrechtslehre und des Rechtsposi-
tivismus. Charlottenburg: Pan.
ÐÐÐ . 1934. Reine Rechtslehre. Vienna: Deuticke.
ÐÐÐ . 1941±42. The Pure Theory of Law and Analytical Jurisprudence. Harvard Law
Review 55: 44±70.
ÐÐÐ . 1945a. General Theory of Law and State. Cambridge, Mass.: Harvard University
Press.
ÐÐÐ . 1945b. Natural Law Doctrine and Legal Positivism. Trans. Wolfgang H. Kraus.
In Hans Kelsen, General Theory of Law and State, 391±446. Cambridge, Mass.:
Harvard University Press.
ÐÐÐ . 1953. Was ist die Reine Rechtslehre? In Demokratie und Rechtsstaat. Festgabe
zum 60. Geburtstag von Zaccharia Giacometti, 143±61. Zurich: Polygraphischer
Verlag.
ÐÐÐ . 1960. Reine Rechtslehre. 2nd ed. Vienna: Deuticke.
ÐÐÐ . 1964±65. Professor Stone and the Pure Theory of Law. Stanford Law Review
17: 1128±57.
ÐÐÐ . 1971. Was ist ein Rechtsakt? In Internationale Festschrift fuÈr Alfred Verdross
zum 80. Geburtstag. Ed. ReneÂ Marcic, Hermann Mosler, Eric Suy and Karl
Zemanek, 153±65. Munich: Fink. (1st ed. 1952.)
ÐÐÐ . 1979. Allgemeine Theorie der Normen. Ed. Kurt Ringhofer and Robert Walter.
Vienna: Manz.
ÐÐÐ . 1984. What Is a Legal Act? The American Journal of Jurisprudence 29: 199±212.
ÐÐÐ . 1992. Introduction to the Problems of Legal Theory. Trans. Bonnie Litschewski
Paulson and Stanley L. Paulson. Oxford: Clarendon.
ÐÐÐ . 1998. Foreword. In Normativity and Norms. Critical Perspectives on Kelsenian
Themes. Ed. Stanley L. Paulson and Bonnie Litschewski Paulson, ch. 1. Oxford:
Clarendon.
Klecatsky, Hans R., et al. 1968. Die Wiener rechtstheoretische Schule. 2 vols. Vienna:
Europa.
Merkl, Adolf Julius. 1917. Das Recht im Lichte seiner Anwendung. Hanover: Helwing.
Nino, Carlos Santiago. 1978. Some Confusions around Kelsen's Concept of Validity.
Archiv fuÈr Rechts- und Sozialphilosophie 64: 357±77.
Ofstad, Harold. 1950. The Descriptive Definition of the Concept ``Legal Norm''
Proposed by Hans Kelsen. Theoria 16: 118±51, 211±46.
Paulson, Stanley L. 1990. LaÈût sich die Reine Rechtslehre transzendental begruÈnden?
Rechtstheorie 21: 155±78.
ÐÐÐ . 1992a. Kelsen's Legal Theory. The Final Round. Oxford Journal of Legal
Studies 12: 265±74.
ÐÐÐ . 1992b. Kelsen Without Kant. In O È ffentliche oder private Moral? Vom
Geltungsgrunde und der LegitimitaÈt des Rechts. Festschrift fuÈr Ernesto GarzoÂn ValdeÂs.

Ed. Werner Krawietz and Georg H. von Wright, 153±62. Berlin: Duncker &
Humblot.
ÐÐÐ . 1992c. The Neo-Kantian Dimension of Kelsen's Pure Theory of Law. Oxford
Journal of Legal Studies 12: 311±32.
ÐÐÐ . 1993. Die unterschiedlichen Formulierungen der ``Grundnorm''. In Rechts-
norm und Rechtswirklichkeit. Festschrift fuÈr Werner Krawietz zum 60. Geburtstag. Ed.
Aulis Aarnio et al., 53±74. Berlin: Duncker & Humblot.
ÐÐÐ . 1994. Kelsen and the Marburg School. Reconstructive and Historical
Perspectives. In Prescriptive Formality and Normative Rationality in Modern Legal
Systems. Festschrift for Robert S. Summers. Ed. Werner Krawietz, Neil MacCormick,
and Georg H. von Wright, 481±94. Berlin: Duncker & Humblot.
ÐÐÐ . 1995. Legal Knowledge versus Legal Interpretation? On Kelsen's Philo-
sophical Reconstruction and Its Limits. In Cognition and Interpretation of Law. Ed.
Letizia Gianformaggio and Stanley L. Paulson, 117±37. Turin: Giappichelli.
ÐÐÐ . 1996. On the Implications of Kelsen's Doctrine of Hierarchical Structure.
Liverpool Law Review 18: 49±62.
ÐÐÐ . 1997a. On the Early Development of the Grundnorm. In Law, Life and the
Images of Man. Festschrift for Jan M. Broekman. Ed. Frank Fleerackers et al., 217±30.
Berlin: Duncker & Humblot.
ÐÐÐ . 1997b. Kelsen's Early Work on Material and Formal Unity. In Justice, Morality
and Society. A Tribute to Aleksander Peczenik. Ed. Aulis Aarnio, Robert Alexy, and
Gunnar Bergholtz, 331±45. Lund: JuristfoÈrlaget.
ÐÐÐ . 1997c. On the Kelsen-Kant Problematic. In Normative Systems in Legal and
Moral Theory. Festschrift for Carlos E. AlchourroÂn and Eugenio Bulygin. Ed. Ernesto
GarzoÂn ValdeÂs et al., 197±213. Berlin: Duncker & Humblot.
ÐÐÐ . 1998a. Four Phases in Hans Kelsen's Legal Theory? Reflections on a
Periodization. Oxford Journal of Legal Studies 18: 153±66.
ÐÐÐ . 1998b. Introduction. In Normativity and Norms. Critical Perspectives on
Kelsenian Themes. Ed. Stanley L. Paulson and Bonnie Litschewski Paulson,
pp. xxiii±liii. Oxford: Clarendon.
Paulson, Stanley L. and Bonnie Litschewski Paulson, eds. 1998. Normativity and
Norms. Critical Perspectives on Kelsenian Themes. Oxford: Clarendon.
Pawlik, Michael. 1993. Die reine Rechtslehre und die Rechtstheorie H. L. A. Harts. Ein
kritischer Vergleich. Berlin: Duncker & Humblot.
Peczenik, Aleksander. 1981. On the Nature and Function of the Grundnorm. In
Methodologie und Erkenntnistheorie der juristischen Argumentation. Ed. Aulis Aarnio
et al. Rechtstheorie Beiheft 2: 279±96. Berlin: Duncker & Humblot.
ÐÐÐ . 1995. Vad aÈr raÈtt? Om demokrati, raÈttssaÈkerhet, etik och juridisk argumentation.
Stockholm: Fritzes.
Raz, Joseph. 1975. Practical Reason and Norms. London: Hutchinson.
ÐÐÐ . 1979. The Authority of Law. Oxford: Clarendon.
ÐÐÐ . 1980. The Concept of a Legal System. 2nd ed. Oxford: Clarendon.
ÐÐÐ . 1986. The Purity of the Pure Theory. In Essays on Kelsen. Ed. Richard Tur and
William Twining, 79±97. Oxford: Clarendon.
Spaak, Torben. 1992. The Concept of Legal Competence. An Essay in Conceptual Analysis.
Trans. Robert Carroll. Aldershot: Dartmouth.
Verdross, Alfred. 1951. Eine Antinomie der Rechtstheorie. Juristische BlaÈtter 73:
169±71.
ÐÐÐ . 1968. Eine Antinomie der Rechtstheorie. In Die Wiener rechtstheoretische
Schule, vol. 2. Ed. Hans R. Klecatsky et al., 1375±80. Vienna: Europa.
Vernengo, Roberto J. 1986. Kelsen's RechtssaÈtze as Detached Statements. In Essays on
Kelsen. Ed. Richard Tur and William Twining, 99±108. Oxford: Clarendon.

von Wright, Georg Henrik. 1963. Norm and Action. A Logical Enquiry. London:
Routledge.
Walter, Robert. 1964. Der Aufbau der Rechtsordnung. Eine rechtstheoretische Untersu-
chung auf Grundlage der Reinen Rechtslehre. Graz: Leykam.
Weinberger, Ota. 1982. Kelsens These von der Unanwendbarkeit logischer Regeln
auf Normen. In Die Reine Rechtslehre in wissenschaftlicher Diskussion. Schriftenreihe
des Hans Kelsen-Instituts 7: 108±21. Vienna: Manz.
Wilson, Alida. 1982. Joseph Raz on Kelsen's Basic Norm. The American Journal of Juri-
prudence 27: 46±63.

Basic Norm

Uploaded by

Copyright:

Available Formats

Basic Norm

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Basic Norm

Uploaded by

Copyright:

Available Formats

Ratio Juris. Vol. 14 No.

2 June 2001 (143±75)

Presupposing the Basic Norm*

Specifically, I argue that Kelsen in speaking of the doctrine of the basic

I. Kelsen's View(s) on the Addressee of the Basic Norm

# Blackwell Publishers Ltd 2001.

was strongly influenced by two other factors, namely, by Adolf Julius

# Blackwell Publishers Ltd 2001.

If a (higher-level) norm is thought to validate another (lower-level) norm,

# Blackwell Publishers Ltd 2001.

# Blackwell Publishers Ltd 2001.

# Blackwell Publishers Ltd 2001.

and the neo-Kantians (transcendental ``grounding'' of the Pure Theory28) the

# Blackwell Publishers Ltd 2001.

which a legal norm is issued (``posited,'' gesetzt) or applied, provided that

# Blackwell Publishers Ltd 2001.

# Blackwell Publishers Ltd 2001.

II. Is It Possible To Defend the Presupposition in a Way that Is

II.1. Aleksander Peczenik: Basic Norm-Transformation

# Blackwell Publishers Ltd 2001.

On the first or ``external'' level, the point of departure is what Peczenik

# Blackwell Publishers Ltd 2001.

# Blackwell Publishers Ltd 2001.

that, according to Kelsen, characterize valid lawÐhierarchical structure,

II.2. Joseph Raz: The Legal Man

# Blackwell Publishers Ltd 2001.

existence of legal argumentation. The tacit assumption of this norm, on the

# Blackwell Publishers Ltd 2001.

# Blackwell Publishers Ltd 2001.

# Blackwell Publishers Ltd 2001.

# Blackwell Publishers Ltd 2001.

rather, a ``professional and uncommitted adoption'' (Raz 1979, 143). To

III. What Difference Does the Presupposition Make?

# Blackwell Publishers Ltd 2001.

III.1. Kelsen's Terminology

# Blackwell Publishers Ltd 2001.

Hauptprobleme der Staatsrechtslehre (1911),63 Allgemeine Staatslehre (1925),64 and

# Blackwell Publishers Ltd 2001.

The statements formulated by legal science are, Kelsen believes, ``ought''-

III.2. A Descriptive Mention of the ``Ought'' (M. Golding)

# Blackwell Publishers Ltd 2001.

What Kelsen may have meant by ``rules of law in a descriptive sense''73

# Blackwell Publishers Ltd 2001.

of a ``proper representation'' of legal norms. The Pure Theory, Golding

# Blackwell Publishers Ltd 2001.

Seen generally, statements in which words or expressions are mentioned

III.3. A Descriptive Use of the ``Ought'' (H. L. A. Hart)

# Blackwell Publishers Ltd 2001.

Hart's attempt to justify Kelsen's terminology is original. He draws

``representation'' of the specific Sinn of the legal normÐnamely, a rendition

IV. Concluding Remarks

# Blackwell Publishers Ltd 2001.

can be attributed to legal science. In my view, his ambivalence here leads

# Blackwell Publishers Ltd 2001.

# Blackwell Publishers Ltd 2001.

During the earlier years of the so-called transcendental or classical phase in

# Blackwell Publishers Ltd 2001.

think, impossible to point to specific situations in which the Kelsenian legal

# Blackwell Publishers Ltd 2001.