Some Confusions around Kelsen's Concept of Validity

Author(s): CARLOS SANTIAGO NINO

Source: ARSP: Archiv für Rechts- und Sozialphilosophie / Archives for Philosophy of Law and
Social Philosophy, Vol. 64, No. 3 (1978), pp. 357-377
Published by: Franz Steiner Verlag
Stable URL: http://www.jstor.org/stable/23679244
Accessed: 30-11-2015 19:42 UTC

REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/23679244?seq=1&cid=pdf-reference#references_tab_contents

You may need to log in to JSTOR to access the linked references.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.

Franz Steiner Verlag is collaborating with JSTOR to digitize, preserve and extend access to ARSP: Archiv für Rechts- und
Sozialphilosophie / Archives for Philosophy of Law and Social Philosophy.

http://www.jstor.org

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
 Some Confusions around Kelsen's
Concept of Validity*

BY CARLOS SANTIAGO NINO, BUENOS AIRES

The object of this article is to clarify some aspects of Kelsen's concept of

legal validity and to show how it has been misinterpreted by some theo
rists, generating confusions that still pervade much of the discussions on
the subject.

The current assumption is that Kelsen's notion of validity is primarily

related to issues such as those of the identity of a legal system, the mem

bership of particular norms to it, the internal consistency of the system,

and that, consequently, it has no affínitywith the idea of validity prevail

ing in traditional legal philosophy, which was mainly connected with the
justification of the law. I think that this assumption is wrong and has led
its supporters to disregard important aspects of Kelsen's theory and to

formúlate unwarranted and tortuous interprétations of it in order to

show how it deals with issues with which its notion of validity is errone
ously supposed to be primarily connected.

Undoubtedly, the peculiarities of Kelsen's theory explain how this

confusion was generated, since its strong positivistic bent makes the

attempt to associate its notion of validity with that of traditional out

looks, of natural law inspiration, to seem implausible, and, furthermore,
the theory does deal with problems such as that of membership in a way
which deceptively evokes an identification with the problem of validity.
I think that one can reasonably summarize the récurrent features of
the concept of validity which was employed in such heterogeneous works
as those of A quinas1, Suarez2, Puchta3, etc., in connection with the

This article is part of a larger work made possible by a Guggenheim Fellowship.

See Summa Theologiae, Blackfriars 1966, Primae-Secundae, question 9,
question 95 art 2, question 96 arts 4 and 5.
See Francisco Suarez, De Legibus, Madrid 1968, Libro VI. For a very interest

ing analysis of Suarez' notion of validity and a comparison of it with that of

Kelsen and other theorists, see E. Garzón Valdes, "Algunos modelos de validez
normativaRevista Argentina de Filosofía, December 1976.
See Outlines of the Science of Jurisprudence, translated by W.Hastie, Edin

burgh 1887, p. 41.

ARSP (Archiv für Rechts- und Sozialphilosophie), Bd. LXIV/3 (1978)

© Franz Steiner Verlag GmbH, D-6200 Wiesbaden

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
358 Carlos Santiago Nino

justification of the law, with the following very general propositions: (a)
To predícate validity of a legalsystem, or a particular legal rule, was to

assert that it had binding force; that its prescriptions constitute conclu
sse reasons for action, (b) When the validity of a legal System, or of a
legal rule, was contested, this was tantamount to denying its existence,
since the lack of validity was taken to imply that the system, or the rule,
did not have the normative conséquences that it stipulâtes (for instance,
if the invalid rule prohibits a certain action, the action would not be,
thereby, prohibited; the rule would be as ineffectuai for establishing
normative relationships as if it had not been enacted). (c) Even when
the ascription of validity might be dépendent on certain facts (for some
conceptions, the effîcacy of the system was a necessary or even a suffi
cient condition for that ascription) the meaning of "validity" was not
descriptive but normative; to say that a system or a particular legal rule
was valid was to endorse it, to maintain that its application and observ
ance was obligatory and justified (this feature of the traditional notion
of validity permits us to hold that the différent authors whose views I
am taking into account, employed the same concept of validity, in spite
of their wide discrepancies on the criteria for ascribing validity).
It seems clear to me that these features are présent in Kelsen''s notion
of validity. The following paragraphs are directed to justify this con
tention, showing that Kelsen's theory, like traditional philosophical
accounts of law, equates the validity of the law with its binding force
and its existence, and that it conceives of the ascription of validity as a
normative judgement.

(a) Validity as Equivalent to Binding Force

On p. 30 of his General Theory of Law and State4 Kelsen states: "To

say that a norm is valid is to say that we assume its existence or — what
amounts to the same - we assume that it has force' for
thing 'binding
those whose behaviour it régulâtes." On the next page Kelsen distin
guishes legal norms from commands, asserting that only legal norms,
enacted by organs authorized to do so, oblige the individuáis to whom
they are directed.
On p. 8 of the Pure Theory ofLaws Kelsen says: "The legislative act,

Edition of 1961, New York.

Second édition, 1967.

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
 Some Confusions aroundKelsen's Concept of Validity 359

which subjectively has the meaning of ought also has the objective
— that is the of a valid norm — because the constitu
meaning meaning
tion has conferred this meaning upon the legislative act. The act whose
meaning is the constitution has not only the subjective but also the ob
jective meaning of ought, that is to say the character of a binding norm;
if — in case — it is the historically firstconstitution, we présupposé in our
juristic thinking that we ought to behave as the constitution prescribes."
In the same work, on p. 15,Kelsen writes: "To say that the behaviour of

an individual is commanded by an objectively valid norm amounts to the

same as saying that the individual is obliged to behave in this way."
Finally, on p. 46, Kelsen says: "It was observed earlier that the validity
of a norm (which means that one ought to behave as the norm stipulâtes)
should not be confounded with the effectivityof the norm ..."

(b) Validity as Equivalent to Existence

The first quotation of the foregoing section shows also this feature of
Kelsen's notion of validity; even more clearly, that phrase is preceded by
the following remark: "By validity we mean the specific existence of
norms." On p. 48 of the same work (the General Theory of Law and
State) Kelsen asserts: "The existence of legal norms is its validity; and
the validity of a legal norm, although not identical with certain facts, is
conditioned by them." Later on (p. 170) Kelsen says: "... normative
jurisprudence asserts the validity of a norm, and that means its existen

ce."
In the Pure Theory ofLaw there are similar remarks, but there is also
an interesting explanation of what is involved in the identification be
tween the binding force of a norm and its existence: "It is, however,

necessary to distinguish the subjective and the objective meaning of the

act [of will], Ought is the subjective meaning of every act of will directed
at the behaviour of another. But not every such act has also objective this

meaning, and only if the act of will has also the objective meaning of an
ought, is this ought called a 'norm'. If the ought is also the objective
meaning of the act, the behaviour at which the act is directed is regarded
as something that ought to be not only from the point of view of the in
dividual who has performed the act, but also from the point of view of
the individual to whose behaviour the act is directed, and of a third indi
vidual not involved in the relation between the two" (p. 7).

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
360 Carlos Santiago Nino

(c) Validity as a Normative Concept

This is the crucial point for the understanding of Kelsen's concept of

validity, for, until the normative character that Kelsen ascribes to this

concept is not fully grasped, interprétations which identifies it with, for

instance, the membership of a norm to a legal System can still prospere,
under the assumption that, when Kelsen equates validity with binding
force, he is using "binding force" in a special sense (to refer, for instance,
to the circumstance that other norm of the System prescribes to obey the
norm in question), and that, when he equates validity with existence,
"existence" just means here membership of a norm to a legal system.
That the ascription of validity to a legal norm does not exhaust itself
in the description of certain facts is repeatedly asserted by Kelsen. For

instance, on p. 193 of the Pure Theory Kelsen says: "A norm referring to
the behaviour of a human being is 'valid' means that it is binding - that
an individual ought to behave in the manner determined by the norm. It
has been pointed out in an earlier that the question why a norm
contextj
is valid, why an individual ought to behave in a certain way, cannot be
answered by ascertaining a fact, that is by the statement that something
is, that the reason for the validity of a norm cannot be a fact. From the
circumstance that something is cannot follow that something ought to

be; and that something ought to be, cannot be the reason that something
M."

But perhaps this feature of the Kelsenian concept of legal validity

can be more firmly established if instead of relying on explicit remarks
of Kelsen we show that it is imposed by the general structure of his

theory.
In Kelsen's theory the validity of a legal rule requires that its enact
ment must be authorized by another rule that is, itself, valid. The judge
ment that a certain raie is valid présupposés, therefore, a judgement that
another raie, which authorizes the création of the former, is valid. The
juristic judgements of validity, about raies which stand in a certain
relationship of "dérivation" between themselves (constituted by the fact
that one authorizes the enactment of other), stand also in a certain

relationship of "dérivation" between themselves. While a rale "derives"

from another when the latter authorizes the création of the former, a

juristic judgement that a raie is valid "derives" from the judgement that
another raie is valid when the former présupposés the latter. This can be
illustrated as two parallel chains of "dérivation", one of raies and the

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
 Some Confusions around Kelsen 's Concept of Validity 361

other of judgements of validity about those rules, formulated typically

by jurists:

Rules Judgements of validity

Rule x: 'The organ A is authorized 'Rule jc is valid'

to enact rule y' I

i i
Rule y: 'If somebody does <pit 'Rule y is valid'
ought to be a sanction'

when it is the case that the rule referred to by a judgement of

Now,

validity is the upper positive norm of a legal system, Kelsen states that its
validity or binding force is presupposed and that the formulation of that
presupposition is the basic norm of the system (p. 115 of the General
He also that "since the reason of the validity of a norm can
Theory). says
be another norm, the must be a norm; not one
only presupposition

posited (i.e. created) by a legal authority, but a presupposed norm."

After this remark Kelsen offers the well-known formulation of the basic

norm as follows "one has to behave as the constitution prescribes."

From these assertions it is easy to infer that the basic norm is not one

which belongs to the chain of dérivation of rules, but one which is part
of the chain of dérivation of judgements of validity. In other words, the
basic norm is a judgement of validity; it is the primitive juristic judge
ment of validity on which all other judgements about the validity of
rules, whose création is directly or indirectly authorized by the constitu

tion, must be based. The obvious conclusion is that, if that primitive

— the basic
judgement of validity is itself, according to Kelsen, a norm
norm - and if from norms but norms can be derived, all further
nothing
of about mies other than the constitution must be
judgements validity
themselves norms. Judgements of validity (including the basic norm)
that what the mies, referred to by the judgements, stipulate
prescribe
to be done. If the mle in that some act is
ought question stipulâtes
to State that
obligatory, to predícate of that mle that it is valid implies
there is an obligation to do that act (which is not the same as to say
but entails
merely that there is a mle that prescribe such an obligation,
the Statement that that mle succeeds in creating the obligation it pre

to be
scribes). The ascription of validity to a rule (which comes, thereby,
conceived as a norm) habilitâtes, in Kelsen's theory, the passage from

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
362 Carlos Santiago Nino

an "inverted commas" normative language for describing the content of

the law to a direct normative language6.

II

As I have said, many contemporary theorists disregard the normative

nature that Kelsen assigns to his concept of validity and explicitly or
implicitly assume that it describes some features of a legal system or of a

particular legal norm, mainly features connected with the criteria for

identifying a legal system, distinguishing it from others, and for establish

ing that a legal norm belongs in a certain legal system. It is true that Kel
sen relates the question of the validity, or binding force, of a legal system
or norm to those of the identification of legal Systems and the member

ship of norms to them. He says: "The legal order is a system of norms.

The question then arises: what is that which makes a system out of a
multitude of norms? When does a norm belong to an order? This ques
tion is in close connection with the question as to the reason of validity
of a norm" (General Theory of Law,p. 110). But this assertion of the
existence of a "close connection" between validity, on the one hand, and
identification and membership, on the other, does not authorize to con
clude that, in Kelsen's theory, questions of validity are reducible to

questions of identification and membership.

This impression has undoubtedly been caused by the fact that the
answers to both kinds of problems, given by Kelsen, can be stated in a

superficially similar way. Kelsen's answer to the question about what

makes a legal system valid can be taken to be "the basic norm", and the
same answer seems to be Kelsen's response to the problem of what makes
a set of norms to constitute an unitary legal system différent from
others. However, it is easy to realize that both answers are substantially
différent. Kelsen's solution to the problem of the validity of a legal
system can be paraphrased as follows: "A legal system is valid when we

présupposé a basic norm which prescribes that it ought to be done what

its rules stipula te." A similar answer to the problem of the identity of a
legal system would clearly be absurd and incomprehensible (and not
merely mistaken). In fact Kelsen's solution to this latter problem is one
which can be paraphrased in this way: "A set of norms constitute a uni

See the distinction between "inverted-commas" and "direct normative language"

in: R. M. Hare, The Language of Moráis, Oxford 1972, p. 124.

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
 Some Confusions around Kelsen 's Concept of Validity 363

tary legal system when we ascribe validity to them on the basis of one
and the same basic norm." Of course, the phrase between quotation
marks is not a paraphrasis of the basic norm or of any other norm, it
formulâtes a criterion of identification which points to the circumstance
that the same basic norm is presupposed when we ascribe validity to all
the norms of the same legal System. Therefore, authors like Aichourron
and Bulygin1 are mistaken when they criticize Kelsen'% basic norm on
the ground that what is needed for the identification of a legal system is
a criterion and not a norm. Kelsen's basic norm establishes the oblig
atoriness of a legal system; its identity is determined by a criterion which
takes into account the fact that the same basic norm is presupposed
when ascribingobligatoriness to all the norms of that system. However, as
a criterion of identity, the foregoing is vacuous, since the content of each
basic norm (and, consequently, their own identity) cannot be established,
in the context of Kelsen's theory, before circumscribing the norms

that belong in the legal system to which the basic norm ascribes binding
force8.

Kelsen also seems to résolve the question of the validity of a particular

norm and the question of its membership to a legal system in an identical
way. His answer to both questions can be stated as follows: "A norm is

valid/belongs in a legal system when it derives from a valid norm of that

system." However, this way of presenting his solution to those questions
is misleading; the foregoing answer can be unfolded into two parts, each
of which deals separately with the two questions at stake: (a) "A norm is
valid when it derives from another valid norm";(b) "A norm belongs in a
legal system when it derives from another norm which belongs in that

system". The latter Statement endorses a "genetic" criterion of member

ship which is not very much in discussion in modem legal theory (of
course it should be completed with a clause dealing with the membership
of the primitive norms of the system; this clause is not successfully pro
vided by Kelsen). This genetic criterion of membership is quite independ
ent of the validity of the norms in question; nowhere in Kelsen's works

it is suggested that it applies only to valid legal norms.

In spite of the fact that the criterion of membership of non-primitive
norms and the criterion for "transmitting" validity from the constitution
referred to by the basic norm to norms which are not directly referred to

7
See Normative Systems, Wien-New York,1971, p. 73 ff., 2.
8
See this criticism in J. Raz, The Concept of a Legal System, Oxford,p. 102.

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
364 Carlos Santiago Nino

by it, are mutually independent, both are based on the same notion of

"dérivation", according to which one norm "derives" from another when

its enactment is authorized by the latter. This produces the resuit that
when the primitive norms of the system are held to be valid, there is a
coextensivity between the class of norms which are valid according to a
certain basic norm and the class of norms which belongs in the system to

which that basic norm applies. Kelsen is, obviously, much in favour of
maintaining this coextensivity (that which explains why many authors
have confused it with identity), and it is this feature of his theory which
genera tes many unwelcome results. The fact is that, while,in relation to

membership, the dérivation from one norm from another, in the sense

explained, seems a prima facie sound criterion, with regard to the "trans
mission" of validity, it is clearly insufficient. If we accept the validity
of certain set of legal norms we are committed to accept the validity of
some other norms, not only when their création was authorized by the
former but also when the norms, whose validity is accepted, "recognize"
or establish the duty to obey those other norms. Thus, it has been
alleged9 that a legal system could recognize as valid the rules of other

legal Systems or the rules of private associations without this implying

that those rules come to be part of the legal system in question. But the

insufficiency of Kelsen''s criterion for "transmitting" validity is best

manifested in his treatment of another problem: that of voidable norms.
Norms which are enacted in violation to the conditions set up by valid
norms of the system are, in some circumstances, considered to be oblig

atory until some compétent organ declare their invalidity. According to

the genetic criterion of membership, these norms are not part of the legal

system in question. Whether this is a sound solution or not may be in

dispute, but if validity were équivalent to membership, there would not

be any inconsistency whatsoever in the proposition that a norm which
is not valid in a legal system is, nevertheless, obligatory according to
that legal system (just as in the case of norms of other legal Systems
which are deemed obligatory by rules of conflicts of laws)10. The fact
that Kelsen does worry about this problem suggests again that he identi
fies validity not with membership but with obligatoriness or bindingness.
Given this identification a norm which is obligatory must obviously be

See J. Raz, Practical Reason andNorms, London 1975, pp. 152-54.

See this analogy in E. Bulygin, "Sentenza goudiziariae creazione di diritto''
Rivista Internazionale di Filoso fia del Diritto 44, 1967, pp. 164-80.

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
 Some Confusions around Kelsen 's Concept of Validity 365

valid. But Kelsen considere a norm to be valid only when its enactment is
authorized by another valid norm and, so, he is compelled to adopt the
absurd thesis that voidable norms are, in fact, authorized by the perti
nent superior norms, since they have, besides its explicit content which

the voidable norm violâtes, a tacit clause which authorizes the création of

any norm with whatever content and enacted by whatever procédure11.

Of course, the evident way out to the problem would be to recognize

that validity or obligatoriness is transmitted from some norms to others

not only when the former authorize the création of the latter but also

when the former impose the duty to obey or apply the latter. A voidable

norm would, thus, be a norm which, not having been authorized by valid
norms of the system in question (and, therefore, not belonging in that

system), is held by other norms of the system to be obligatory, and,

therefore, valid, until some especial procédure of nullification is com

pleted. This solution would negate the coextensivity between validity

and membership to a system, but this is all to the better since the main

tenance of that coextensivity is one of the most remarkable weaknesses

of Kelsen's theory and the source of the current confusion, among inter

preters of that theory, between one and the other concept12.

III

Some authors, like Hart13 and Aichourron and Bulygin14, assume that

Kelsen s notion of validity is équivalent to membership without much

dicussing the matter. Other authors have considered the possibility that

that notion be related to that employed by traditional legal philosophers

in connection with the justification of the law, but have eventually

rejected this interprétation. Among these latter analyses it is worth

See General Theory of Law and State, op. cit., pp. 153-162.
See a more extensive treatmcnt of the problem of confliet of norras of différent
levels in Kelsen's theory in my "El concepto de validez y el problema del con
flicto entre normas de diferente jerarquía en la Teoría Pura del Derecho", in:
Derecho Filosofía y Lenguaje. Homenaje a A. L. Gioja, edited by G. P. Carrio,
Buenos Aires 1976.
See The Concept of Law, Oxford 1961, p. 245 (notes) where Kelsen's basie
norm is compared with Hart's own rule of récognition assuming tlrat they have
the same role.

Sec, op. cit., pp. 72 et seq.

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
366 Carlos Santiago Nino

considering those of J. Raz in his book The Concept of a Legal System15

(he has lately changed his interprétation of Kelsen, as we shall see in the
next section) and J. W. Harris in the article "When does the Grundnorm

Change"16.
Raz points out certain features which, in Kelsen's theory, are charac
teristic of norms, ail of which are linked with the double character of
norms as guides for conduct and as patterns for evaluating behaviour. But
Raz realizes that ail these features are common to norms and ordinary
commands, so he asks himself what would be the property on which, in
Kelsen's theory, the distinction between norms and commands is based.

Taking into account some passages of Kelsen's books (some of which

have been quoted in section 1),Raz draws five possible interprétations of
the idea that norms differ, in Kelsen's theory, from commands in that
the former not only serve to justify actions but are themselves justified.
These interprétations are, in my own words, the following:

(a) That every legal norm is justified.

(b) That, although legal norms are not necessarily justified, when we
use "oughf'-statements to describe them, we mean both that a norm
exists and that it is justified.
(c) That "oughf'-statements describing the law do not convey cate

gorial judgements that the norm in question is justified but mere hypo
thetical judgements that it is justified if the basic norm of the System
is justified.
(d) That "oughf'-statements do not mean that the norms to which

they refer are, either categorically or hypothetically justified, but when

we use them we are implying that the norms in question are regarded as

justified by the bulk of their addresses.

(e) That "oughf'-statements do not mean that the norms to which

they refer are justified but when we use them we imply our belief that

they are so.

Raz rejects rapidly the first interprétation on account of Kelsen's

positivistic outlook. Then he rejects together the second, fourth and fifth
interprétations in a not very clear passage which seems to contain mainly
these two arguments: (i) that Kelsen is clear in to
ascribing "oughf'
statements a descriptive — the second is
only meaning so, interprétation
false — and (ii) that Kelsen is not equipped with a linguistic theory which

See, op. cit., chapter VI, section 'The Basic Norm and Dynamic Justification'
In Cambridge Law Journal 29 (1971).

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
 Some Confusions around Kelsen 's Concept of Validity 367

would enable him to assert that when we make an Statement we could

not covered the of the Statement — the

imply things by meaning so,
fourth and fifth interprétations are also false. At this point Raz says that

the foregoing considérations suggest that Kelsen only thinks of the basic

norm as a common link between all the norms of a legal System (serving
as a ground for resolving problems of identification and membership)

having nothing to do with the justification of the law. Some passages of

Kelsen, Raz says, seem to give support to the third of the interprétations

abovementioned, but this is only because of his unfortunate confusion

between the conditions for the existence of norms and the use that we

can make of those conditions when justifying them. Such interprétation

(the third one) cannot reflect Kelsen's posture, since it would imply,

given the équivalence between validity and existence, that the existence

of legal norms is conditional to the assumption of the justification of the

basic norm, that which would contradict Kelsen's denial that legal norms,
like idéologies, have a mere conditional existence.

I believe, as Raz himself seems to think at présent, that, though none

of the foregoing interprétations of the Kelsem&n notion of a valid norm

is some of them — the third

wholly correct, nevertheless, especially
and fifth - reflect some of that notion. This will be
partial aspects
substantiated in the last section; for the moment it is worth seeing that

the arguments that Raz addressed against those interprétations are not

fully convincing. First, it is true that Kelsen conceives of the meaning of

"oughf'-statements as purely descriptive and that he is not equipped

with a linguistical theory allowing him to distinguish between the mean
ing of a Statement and what could be implied in the formulation of that

Statement. But in lieu of such a linguistical theory Kelsen is equipped

with an epistemological theory, of Kantmn inspiration, which allows him

to maintain that a certain kind of knowledge présupposés some norma

tive assumptions, and, therefore, to contend that some sort of descriptive

cannot be formulated without adopting a normative point of

proposition
view. As it is well-known, Kelsen thinks that, in the same way that the

of causality is a necessary epistemological presupposition for

principie
our knowledge of the natural world, the principie of "imputation", one
of whose applications is the basic norm, is an epistemological require
ment for our knowledge of normative reality. Secondly, the epistemolo

gical outlook just referred to gives some ground for the idea that, for
Kelsen, the existence of the norms is conditional, in certain sense of

"conditional". In fact, Kelsen says things like this: "[Epistemologically]

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
368 Carlos Santiago Nino

the science of law 'créâtes' its object in so far as it comprehends the ob

ject as a meaningful whole"17. (t is not surprising that under the Kant

ien outlook that Kelsen adopts, he would be ready to admit that condi

tion for the knowledge of legal norms is also a condition for their exist

ence. A condition for kpowing legal norms as such would be to assume

their justifiability (that which is done when we présupposé a basic norm

which establishes their obligatoriness) ; therefore, the existence of legal

norms would be conditional to that assumption.

'
Now I want to deal with Marris analysis of the topic. He is concerned

in the article mentioned with the problem of the change of the Kelsen

ian basic norm, that which he connects with the problem of whether

judges are justified, after a révolution has taken place, to recognize the

new system as a legal one. His interesting argumentation follows basically

this route:

(i) Kelsen maintains that his basic norm has these two functions: a) it
enables us to regard norms as objective "ought" as différent from simple
commands, and b) it permits us to interpret ail valid norms as a coherent

field of meaning.
(ii) But the first of the above functions is in fact not performed by the
basic norm, nor does it need to perform that function. For a norm can

be distinguished from a command before we présupposé the basic norm,

by merely taking into account the différences between the contexts in

which one and the other are formulated.

(iii) Kelsen ascribes the first function to the basic norm as a result of
a logical error he commits; he confuses two différent senses of the word

"validity". This word can be used to predícate at least four différent

things of norms: (a) that a norm conforms to a higher norm; (b) that a
norm is part of a consistent field of meaning; (c) that a norm is effica

cious; (d) that a norm has "an inherent claim to fulfilment". In defend

ing a pure science of law, Kelsen rejects (d); and when he distunguishes
between validity and efficacy, Kelsen also rejects (c). Kelsen uses "validi
ty" sometimes with the meaning (a) and sometimes with the meaning
(b), and frequently he confuses one with the orther. The sense (a) is
relevant to distinguishing between a command and a legal norm but it has

nothing to do with the basic norm. The converse is true with regard to

This quotation is madc by Harris in the article referred to in note 16; sec foot
note 16, p. 106.

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
 Some Confusions around Kelsen 's Concept of Validity 369

sense (b), which is in fact the central meaning of "validity" in Kelsen's

theory.

(iv) Legal science arranges its material as a consistent field of meaning,

that is to say, it requires that norms of the same level should not be

contradictory and that lower norms should derive from higher ones.

(v) To fulfíl this aim of legal science something like the basic norm
must be accepted.

(vi) If legal science is to be a useful activity, it can only apply a basic

norm to efficacious Systems, so as to interpret them as a consistent field
of meaning. Therefore, when a révolution succeeds, legal science must

présupposé a basic norm in relation to the résultant System, lest it ceases

to fulfil its useful function. •

(vii) Judges are not legal scientists and, therefore, the reason why legal
science must présupposé the basic norm in relation to a post-revolution

ary System does not directly apply to them. But, if judges do not follow
the solutions of legal science, this would become ineffective, and, as legal
science must be useful, then judges should présupposé the basic norm
adopted by legal science. So judges have a reason, independent of any
political commitment, to accept the norms enacted by revolutionaries.

(viii) The positivist doctrine maintains that there is a legal duty of

judges to accept the norms of a révolution. As good judges, as legal
science judges, they ought to continue with their valuable task under the

new law. However, this does not exclude the possible existence of moral

considérations of a différent kind that would justify judges to resign their

posts once a révolution succeeds.
I shall not discuss all the steps of this reasoning. I only want to make
a general observation about its structure and to point out some of the

mistakes that I think Harris commits. The general observation is that the

reasoning at stake has the curious feature of starting from what appear to

be evaluatively neutral premises, among which there are the rejection of

ail association between validity and justification, and the identification of

validity with consistency, in spite of which the reasoning leads to a norm

ative conclusion about what judges ought to do in the face of a révolu

tion (a conclusion which coincides with the ideology that Ross calis
"pseudopositivism", distinguishing it from the theoretical positivism that
he and others defend18). The pitfalls that I want to point out are the
following:

18
See "El concepto de validez y otros ensayos", op. cit. Buenos Aires 1969, p. 25.

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
370 Carlos Santiago Nino

(a) Harris limits the relevance of the Kelsenmn basic norm to the fonc
tion of enabling jurists to see the law as a consistent corpus. Besides the

fact that this disregards large parts of Kelsen's élaboration, in which he

seems to assign quite a différent significance to the basic norm, it is
difficult to see how the basic norm may assist jurists to describe the law
as a coherent system. In order to elimínate contradictions from a legal

System, ail what we need is a criterion of membership which establishes

that two norms with incompatible solutions cannot be included as part

of the same legal system. This cannot conceivably be the fonction of a

norm which prescribes that what other norms stipulate is obligatory. It is

true that Kelsen once maintained19 that legal Systems are necessarily
consistent and that he seemed to attribute the fonction of preserving this

consistency to the basic norm. The adoption of this stränge thesis could
receive différent explanations; it may well be a conséquence of identify

ing validity with justification and seeing the basic norm as the expression
of that justification; there is, indeed, a case for saying that one cannot

justify, or suppose the justification of, contradictory norms20. But this

exclusion of contradictions is not, at any rate, a fonction of a norm with

the content of the Äe/seman basic norm (even when it could be a requisite

for formulating a rational basic norm), less is it the only fonction of

that basic norm.

(b) Neither the context nor the dérivation from another rule is
enough, in Kelsen's theory, for distinguishing a norm from a command.

Whatever Harris means by "context", I am confident that Kelsen would

reject its relevance for identifying legal norms (my confidence is based on
the simple fact that nowhere in Kelsen's works any significance is given
to circumstances such as the purposes with which a prescription is given,
the factual way in which power is acquired, etc.). In relation to the déri
'
vation from another mie (the sense (a) of "validity" in Harris classifica
tion) as a criterion which would allow to distinguish norms from com

mands, its ineffectivity is apparent. A command of a gangster may well

"derive" from another command, viz. one of the chief of the gang

prescribing his subordinates to émit the former command. Kelsen ob

J. Finnis has called my attention to the fact that Kelsen had abandoned his
position about the necessary consistency of legal Systems, in the article "Law
and Logic", published in Philosophy and Christianity, Amsterdam 1965.
See this point in J. Raz, "Kelsen's Theory of the Basic Norm", American Jour
nal of Jurisprudence 19(1974),p.106.

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
 Some Confusions around Kelsen 's Concept of Validity 371

viously requires dérivation from a valid norm, and "valid" here cannot

merely describe the fact that this other rule derives from a further one,
ñor can it mean membership to a consistent System (since the command

of the gangster may well satisfy this condition). Curiously enough, Harris
suggests, in a footnote, a possible distinction between "subjective" and
"objective" normative meanings (which in Kelsen corresponds, respec

tively, to mere commands and valid norms) which points in the direction
that he refuses to follow up in the text: "Kelsen's distinction between

and normative - he - is not

'subjective' 'objective' meaning says clear,
but the following interprétation is suggested: when the complete artic

ulation of a normative expression would require the use of a verb subject

other than the subject of the normative verb, its meaning is 'subjective';
if it does not require such an additional verb subject, its meaning is
- contrast: 'I (you, he) say(s) you (he, all men) ought to
'objective'
do such and such; and I (you, he, all men) ought to do such and suh'"21.

Clearly this grammatical distinction is not accidentai, but reflects a

distinction in the meaning or force of the respective Statements. The

second kind of Statement is formulated when a claim about the justif

iability of the prescription is implied. According to Kelsen, it is the

formulation of this kind of Statement, in giving account of norms,
which distinguishes them from mere commands. A prescription is a valid
norm when, in describing it, we assume its justifiability. If any of the
senses of "validity" enumerated by Harris is connected with Kelsen's

criterion for distinguishing norms from mere commands, it is the sense

(d) whose relevance is hastily rejected by him.
(c) In spite of Harris' rejection of the identification between validity
and justifiability or obligatoriness, it is obvious that that identification
is surrepticiously in his reasoning. The task of presenting the
accepted
legal System as a consistent corpus is undoubtedly a useful activity of

legal science and there could be a case for saying that this task should be

carried out even in relation to highly immoral but efficacious legal

Systems that people are morally obliged to disobey (it can be alleged,
that an immoral and contradictory legal system is worse than
perhaps,
an immoral but consistent one). But from this nothing follows about the
of the norms of the system; jurists could coherently présent
acceptance
the legal in force as a consistent whole and, at the same time, urge
system
courts not to apply it. Besides, the relationship between legal science and

In op. cit., footnote 38, p. 113.

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
372 Carlos Santiago Niño

judicial activity is more complex than what Harris suggests. If courts

were to refuse the application of norms enacted by revolutionaries and

adopted, instead, other norms, legal science could still continue with "its
useful function", presenting those other norms as a consistent whole.
If legal science were as descriptive and evaluatively neutral as Kelsen
thinks it must be, it should limit itself to describe, and perhaps to impose
consistency upon, the norms that courts happen to accept (the effica
cious norms), without this activity bearing any conséquence upon the
matter of what norms courts should accept. The origin oí Harris' confu

sion is manifested in his remark that positivistic doctrine maintains the

legal duty of judges to accept the norms of an efficacious legal System.
If this is a positivistic posture at all, it corresponds to the ideology that,
as I have already mentioned, Ross calis "pseudopositivism", viz. the

ideology that efficacious Systems have some inhérent justifiability; it

does not reflect the theoretical conception to which Kelsen explicitly
subscribes (even when it is arguable that pseudo-positivist assumptions
have surrepticiously pervaded his theoiy22),and it certainly does not
coincide with the posture of authors like Ross and Hart. One critical
point in which Harris départs from Kelsen in this respect is in his thesis
that not only jurists but also judges should présupposé the basic norm.
Even when Kelsen is not very clear about this, I doubt very much that
he would be enthusiastic about this extension of the class of people who
should accept the basic norm, for one thing is to reason theoretically
under the assumption that an efficacious system is justified, and another

thing is to act under that assumption; in the latter case there is no

significative différence between merely assuming that the legal system is

justified and fully accepting that justifiability, and to say that courts
must présupposé the basic norm towards efficacious Systems amounts to

saying that they should act in the same way as they would act if they
believed that the system is justified. The mistake that Harris commits in
his characterization of positivism is exactly the same that R. Dworkin
commits when he says23 : "I have no answer to the argument that the
term 'law' can be used in such a way as to make the positivist's thesis
true by stipulation. It can be used in such a way that the speaker re
cognizes as 'legal' standards only those standards that judges and lawyers

"
As Ross maintains, see op. cit., p. 43.
23
"Social Rules and Legal Theory", Y aie Law Journal 81, 1972.

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
 Some Confusions around Kelsen 's Concept of Validity 373

cite which are in fact identified by some commonly recognized features.

No doubt 'law' can be used in that way, and perhaps some lawyers do so.
But I was concerned with what I took to be an argument about the con

cept of law now in general employment, which is, I take it, theconcept of
the standards that provide for the rights and duties that the government
has the duty to recognize and enforce, at least in principie, throughthe
familiar institutions of mies and police." The concept of law which
Dworkin does not object is,by and large, just the concept which positivists
adopt; the other one (which obviously présupposés a direct normative
sense, and not merely an "inverted commas" use, of the word "duty"
in the phrase . . the government has the duty to recognize and en

force. ..") is that which is put forward by natural law théories, including
the pseudo-positivistic ideology that Dworkin and Harris confuse with
theoretical positivism. In sum, despite his rejection of any association
between judgements of validity and the hypothetical assumption that the
system is justified, Harris ends up by accepting a close connection be
tween judgements of validity and a füll avowal of the justification of the
System, which, in so far as validity is considered to be definitory of
"law" and "legal norm", is inconsistent with the posture of Kelsen and

other positivists.

IV

For Kelsen to say that a legal system or a legal norm is valid (that is to
say, that they are just a legal system or a legal norm) is to say that they
are justified or obligatory. Legal science does not need to commit itself
to the moral posture that the legal system it describes is actually valid

or justified, but it must at least work under the assumption of that

validity or justification in order to présent the law as a normative phenom

enon and not merely as a sequence of facts. As Rai says now24, for

Kelsen, as for natural law theorists, the only conceivable type of norma

tivity is "justified normativity". One cannot speak of legal norms, legal

rights, legal duties, etc. without accepting or at least presupposing their
justification. The important différence between Kelsen and natural
lawyers is that, while the latter maintain that legal science must adopt
a moral commitment in order to identify the law, Kelsen thinks that

"Kelsen's Theory of the Basic Norm", op. cit., p. 99.

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
374 Carlos Santiago Nino

that commitment can be avoided, and should indeed be avoided by legal

science, through the expédient of taking the justification of what is
described as law as a working hypothesis and not as a moral conclusion.
Kelsen's view that one cannot identify norms unless one accepts or

présupposés their justifiability might be contrasted with the fanciful view

that one could not describe beliefs unless one accepts or présupposés
that they are true. In the case of judgements about beliefs the same

grammatical distinction that Harris notes in relation to "oughf'-judge

ments, can be found, and it may ground a parallel distinction between

"subjective" and "objective" judgements: when one is certain about the

truth of some proposition, one can say, for instance, "the snow is

white", instead of the uncommitted phrase "I (you, he) believe(s) that
the snow is white". Of course, the point of bringing forth this parallel is
that, in the same way that we do not need to accept or présupposé the
truth of a belief in order to describe it, we do not need to accept or
présupposé the justification of a norm in order to describe it.
But the matter is more complex than the foregoing conclusion
assumes. In the case of beliefs we need to distinguish between the mental
State or attitude of some person which we describe as "believing in

something" and the proposition which is believed. If we want to give

account of the fact that a certain person has the mental state or attitude
towards some proposition which is characteristic of believing in that pro
position, of course we do not need to accept or présupposé the truth of
that proposition. But if we are interested in giving account of the content
of the belief, that is the proposition itself, irrespective of the fact that
somebody in particular believes in it, one is tempted to think that we
cannot do that without stating the proposition, and that we cannot state
the proposition without playing some language-game in which the truth
of the proposition is either asserted or assumed, etc.25. This question is
relevant to the Kelsenian theory of norms sin ce he seems to conceive of

I am not sure about whether the idea of "describing a proposition" has any
sense at all when it is distinguished from the détermination of the truth-value of
the proposition and the display of its logical conséquences. If my doubts were
confirmed, this would be a décisive objection to the view assumed by some
juiists that the central task of legal science is to describe norms and that this
task can be distinguished from the empirical task of determining which norms
are efficacious, the evaluative task of assessing which norms are valid, and the
logical task of displaying the conséquence of norms.

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
 Some Confusions around Kelsen 's Concept of Validity 375

norms as something parallel to descriptive propositions, as the "ideal

content" of some attitudes or acts, which cannot be described without

re-stating it through normative judgements. This would explain why he

insists in his apparently stränge thesis that juristical "rules of law" are
ought judgements despite the fact that they are purely descriptive of
norms; it is precisely because they are descriptive of norms (that is de
scriptive of certain propositional contents), that Kelsen thinks that rules
of law must re-state those norms, though within a language-game
différent from that within which they were enacted or are accepted by
people who adhere to them. Therefore, when Raz in one of his recent

works26 argues against Kelsen that the "justified" conception of nor

mativity is not the only one available and that there is a "social" con
ception of normativity, according to which norms are a kind of social
practice, Kelsen would certainly answer that norms are not social prac
tices (in the same way that descriptive propositions are not the psycholo
gical states or attitudes which we call "beliefs") but are what may or may
not be (depending on the efficacy of the norm) the "ideal content" of
some social practices.
Raz seems to have lately27 oriented himself in this direction when he
asserts that the normative language used by jurists in describing law
cannot be analyzed as descriptive of certain practices, beliefs or attitudes,
since that language, without apparent change of meaning or force, is used

as well to describe hypothetical norms which are not in force, and that,
consequently, the use of that language must imply the adoption of some
normative point of view not very différent than that described by Kelsen
as typical of legal science. However, Raz does not think that norms can

be identified with deontic propositions, since deontic propositions are

not reasons for actions, and, for Raz, this feature is of the essence of

norms. This position poses very complex and interesting problems which
I hope to discuss in another work.

In the meanwhile, perhaps it is worth pondering the following possible

to Kelsen: If norms are deontic propositions, it is absurd to charac
reply
terize the role of legal science as just the description of norms. What legal
science describes is the fact that some norms are accepted in certain
social practices (that which makes some norms to be part of the law in

26
In op. cit., note 24, p. 111.
27 and in "Legal
In Practical Reason and Norms, cited, p. 170 et seq. Validity"
unpublished.

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
376 Carlos Santiago Niño

force, depending on the character of the practices), it develops, besides,

the logical conséquences of those norms, and perhaps (if we relax the
purely descriptive role that Kelsen assigns to it) it should, further,
reformulate the system so as to elimínate its contradictions and to fill its
gaps. None of these tasks require the assumption of the validity or

justifiability of the norms in question. The firstof them, in particular, is

not a "description" of deontic propostions which would, supposedly,
require to re-state them within some normative language-game. Perhaps
the origin of many puzzles around this matter is that we have become

accustomed to use the phrase "description of norms" as a shorthand for

a variety of intellectual activities about norms some of which do not

quite consist in describing them.

CARLOS SANTIAGO NINO

Some Confusions around Kelsen's Concept of Validity

Summary

The concept of validity employed in Kelsen's theory is a central issue of

discussion in contemporary Jurisprudence. This article puts forward the
thesis that that concept has been misinterpreted in so far as its normative

nature, which it shares with the notion of validity used by natural law

philosophers, has not been sufficiently grasped. Once this feature of the
Kelseni&n concept of validity is taken into account, many problems in
the interprétation of the Pure Theory vanish and some of its shortcomings
are brought to notice. Some wrongaccountsofÄefen's notion of validity
are discussed in the article. Finally, some hypothèses are suggested in
order to explain why Kelsen includes in his theory a notion of validity
which is apparently in consistent with its basic postulations.

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions
 So me Confissions around Kelsen 's Concept of Validity 377

CARLOS SANTIAGO NINO

Quelques confusions sur le concept de validité de Kelsen

Résumé

Le concept de validité dans la théorie de Kelsen a une place centrale dans

les discussions de la philosophie du droit contemporain. Dans cet article
on propose la thèse que ce concept a été mal interpreté pour ne pas avoir
advertie sa nature normative, qui le rapporte avec la notion de validité

employée par des philosophes du Droit Naturel. Une fois qu'on voit ce

caractère du concept kelsenien de validité, beaucoup des problèmes de la

Theorie Pure s'évanouissent et quelques unes des ces limitations sont
mises en valeur. On analyse, en particulier, quelques interprétations ratées
de la notion de validité qu' emploie Kelsen. Finalement, on suggère
quelques hypothèses qui ont pour but d'expliquer pourquoi Kelsen incor

pore à sa théorie un concept de validité apparemment inconsistent avec

ses postulâtes basiques.

CARLOS SANTIAGO NINO

Einige Schwierigkeiten bei der Deutung des Kelsenschen Geltungsbegriffs

Zusammenfassung

Kelsens Begriff der Geltung der Rechtsnormen ist ein zentrales Diskus

sionsthema in der heutigen Rechtsphilosophie. In diesem Aufsatz wird

die These vertreten, daß dieser Begriff oft irrtümlich gedeutet wir , in
dem man seine normative Natur verkennt: es handelt sich um einen

normativen Begriff, der auch von der Naturrechtsphilosophie gebraucht

wird. Diese Erkenntnis erlaubt es, mehrere Probleme der Interpretation
der Reinen Rechtslehre zu klären. Dadurch kommen auch einige ihrer

Fehler zum klar zum Vorschein. Daraufhin werden einige Fehldeutungen

des Geltungsbegriffs bei Kelsen besprochen, und zum Schluß wird der
Versuch unternommen zu erklären, warumMe« diesen normativen Be

griffgebraucht, der mit den Grundprinzipien seiner Theorie unverträglich

zu sein scheint.

This content downloaded from 137.132.123.69 on Mon, 30 Nov 2015 19:42:44 UTC
All use subject to JSTOR Terms and Conditions