<< . .

( 38)

. . >>

an individual is “just” in the legal sense has the same character logically
“as a statement by which we subsume a concrete phenomenon under an
abstract concept.” It is therefore “an objective judgment of value” to be
clearly distinguished from subjective judgments.11
Justice in any moral sense does not enter into Kelsen™s preference for a
Rechtstaat. He uses the term to describe a legal order, but for a very
different purpose than Kant. Insofar as Kant™s Rechtstaat is an associ-
ation defined by nothing but the rules governing it, Kelsen would appear
to echo Kant when he says, “If it is asked why an individual together with
other individuals does belong to a certain state, no other criterion can be
found but that he and the others are subject to a certain, relatively
centralized, coercive order. All attempts to find another bond that holds
together and unites in one unit individuals differing in language, race,
religion, world concept, and separated by conflicts of interests, is doomed
to failure.”12 Kelsen is not, however, asserting that the members of a
Rechtsstaat are autonomous persons, but he is asserting that there is no
reality to any notion of national character or spirit, nor to any ideology,
metaphysics, or mysticism about the state. He is concerned not with
establishing the incompatibility of a “common objective” with “freedom,”
but with eliminating any ground for passing judgment on the existing

Ibid., 14.
Hans Kelsen, “The Pure Theory of Law and Analytical Jurisprudence,” Harvard Law
Review 47 (1941): 55.
Kelsen, General Theory of Law, 14.
Ibid., 14.
Kelsen, Pure Theory of Law, 287.
Edward Levi, Hans Kelsen, H. L. A. Hart 207
legal order. His aim is to identify an object of legal cognition that is
untainted by “values” and therefore is a proper subject for scientific
investigation. Ultimately, like Kant, he wished to remove any impediment
to a world state.
The Pure Theory of law accordingly discovers what the law “is” and
not what it “ought” to be. Kelsen regards this distinction as a departure
from Austin™s command theory; what he and Austin share he confines to
a disposition to proceed solely by analysis of positive law. But there can
be no doubt that Kelsen significantly differs from Austin when he substi-
tutes the concept of a “norm” for that of a “command” to explain the
character of law.13
Law cannot be a command, Kelsen argues, because a command is the
expression of the will of an individual directed to the conduct of another
individual. There is a command only so long as both the will and its
expression are present. But legal rules remain valid even if the individuals
who made them have ceased to be. In fact, some legal obligations that
exist probably do not represent the “real will of anyone.” When the
members of a legislature die, the statutes enacted by them remain in force,
and although a statute is directly contrary to the will of those legislators
who voted against it, the statute is regarded as an enactment of the whole
legislature. Moreover, a large proportion of the members of a legislature
who vote for a bill either do not know its content or know it very
At most, then, to call a law a command is, Kelsen says, only a figurative
expression. It is based on the analogy between enacting or prescribing and
commanding. But the former two occur “without any psychic act of will.”
Therefore, law is better described as a “depsychologized” command; more
precisely it is a “norm.” Whereas a command prescribes a particular act, a
norm prescribes a manner of conducting oneself. To think of laws as
commands is to introduce a “superfluous and dangerous fiction of the
˜will™ of the legislator or the state.”15
The concept of a norm makes it possible, Kelsen believes, to keep clear
the distinction between “is” and “ought,” as Austin failed to do. A norm
states what “ought” to be done, but to describe the law as a system of
norms is to say not what it ought to be, but only what it is. Thanks to the
concept of a “norm,” it is possible to describe the law without either
leaving out its prescriptive character or entering into the realm of
“ought.” His new concept also enables Kelsen to reject Austin™s charac-
terization of a law as “enforceable” without denying that coercion is

Kelsen, “Analytical Jurisprudence,” 54“55.
14 15
Ibid., 55“56. Ibid., 56“57.
208 The significance of rules

central to the law. A law does not, Kelsen argues, force men to obey its
commands. Nor is it at all certain that the lawful behavior of individuals
is brought about by fear of the threatened sanction. Rather, a law is a
“norm which provides a specific measure of coercion as sanction.” This
sanction follows upon a “delict,” that is, illegal conduct.16 Or in other
words, the law says that certain consequences “ought” to follow upon
certain sorts of behavior.
A legal system consists of a hierarchy of norms. The hierarchy has the
character of a series of more and more confining frames. Interpretation of
legal norms, or movement from a more general to a less general norm,
consists in fitting a smaller frame into a larger one. When the judge
applies the law to an individual case, he fits the particular case into an
appropriate frame.
Kelsen opposes this conception of the law as a frame to what he takes
to be the traditional view that law can render certain decisions by an “act
of cognition.” At the same time, he dismisses analogical reasoning as a
worthless description of interpretation because analogies can lead to
diverse results, and “no criterion exists to decide when the one and when
the other should be applied.”17 As he cannot, therefore, credit analogical
reasoning with rationality, Kelsen assumes that the traditional insistence
on the rationality of law must mean, as it does for Kant, that there is only
one logically correct interpretation of the law. This conclusion, he argues,
is a fiction that, though useful for supporting legal security, is not a
legitimate part of a scientific description of positive law.
The act of interpretation is not, Kelsen emphasizes, an act of cognition
because there is always more than one possible object that will fit into a
given frame. The designation of one of several possibilities must then be,
according to the Kantian dichotomy between reason and will, an act of
will. And it is an act of will that “creates” law. If the interpretation is
being done by a “law-applying organ” it is an authentic creation of law.18
Thus, by a very different route Kelsen unwittingly arrives at the same
conclusion as the Realists. By describing every movement to a more
particular norm as an act that “creates” law, he denies the traditional
distinction between legislation and adjudication.
But this view of law contradicts Kelsen™s conception of validity, which
he treats as a logical criterion. The validity of a norm is always derived
from a higher norm, and in this context Kelsen describes the relation
between higher and lower norms as that between a major premise and the
conclusion derived from it.19 This means, Kelsen emphasizes, that the

16 17
Ibid., 58. Kelsen, Pure Theory of Law, 352“53.
18 19
Ibid., 353“55. Ibid., 194.
Edward Levi, Hans Kelsen, H. L. A. Hart 209
whole legal order can be understood as a logical system. At the same time,
however, Kelsen felt obliged to explain, as Kant had not, the uncertainty
of the law. To solve this problem he introduced the notion of law as a
“frame.” The result is a self-contradictory account of a legal system that
unwittingly describes it both as a chain of deductive reasoning and as an
act of will.
But to the problem of explaining the ultimate justification of the norms
that constitute a legal system, Kelsen proposed a more consistent and
ingenious solution. A valid law should be obeyed, according to Kelsen,
because it has been issued by someone who is authorized by a higher norm
to make such a law. As this leads to an infinite regress of norms that could
never give validity to the system as a whole, what is needed, Kelsen
concluded, is a norm at the base of the legal system that is not enacted.
This ultimate norm cannot be “directly evident,” in the sense that it
emanates from reason, “because the function of reason is knowing and
not willing, whereas the creation of norms is an act of will.”20 As the
validity of the ultimate norm can neither be derived from a higher norm
nor its reason questioned, it must then be a “hypothetical” norm or, in
other words, the presupposition of a “law-making authority whose
norms are, by and large, observed, so that social life broadly conforms
to the legal order based on the hypothetical norm.” Kelsen calls this
presupposition the “basic norm.”21
Kelsen™s “basic norm” has the same epistemological status as the
categories of Kant™s transcendental philosophy. Just as the categories
are not data of experience but conditions of experience, so the basic norm
is a condition of juridical science. The hypothetical basic norm answers
the question: “how is positive law possible as an object of cognition, as an
object of juridical science; and, consequently, how is a juridical science
possible?”22 A legal system can then be identified as a hierarchy of norms
whose validity can be traced to the same basic norm, which hierarchy also
constitutes their unity.
But Kelsen is not content with a merely formal solution. He adds an
empirical condition for the validity of the legal system as a whole. The
norms of a valid system remain valid, Kelsen tells us, “only on the
condition that the total order is efficacious; they cease to be valid, not
only when they are annulled in a constitutional way, but also when the
total order ceases to be efficacious.”23 This condition saves the efficacy of
the legal system from being questioned because a particular law or laws
have become ineffective.

20 21
Ibid., 196. Kelsen, General Theory of Law, 437.
22 23
Ibid., 437. Ibid., 119.
210 The significance of rules

But it also introduces a significant uncertainty. Kelsen appears to
suppose that whether or not a legal order is “efficacious” is self-evident.
Yet the difficulty of deciding after World War II whether the legal order
in Nazi Germany could be said to have broken down, or whether it was
sufficiently in effect to justify acts accepted as valid by it, suggests that
Kelsen™s ultimate criterion of “efficacy” may yield judgments not wholly
uncontaminated by “subjectivity.” Moreover, to say that the validity of a
legal system as a whole rests on its “efficacy” equates right with might.
All these difficulties arise out of Kelsen™s anxiety to disengage the
concept of a norm from any suggestion that it has to do with acts of
deliberation or choice. Although he distinguishes the world described
by natural science from that described by a science of norms, he refuses
to attribute free will to men. Free will, he believes, denies that a man™s
will is “causally determinable.” The law need make no such assumptions.
Indeed, it cannot make such assumptions because a “normative, behav-
ior-regulating order” would not be possible if human behavior were not
determined by causes. For a norm “commanding a certain behavior” can
only signify “the cause of a norm-conforming behavior.”24
The law needs to assume only that a man may be held responsible for
what he does. This means, according to Kelsen, that a man™s will may be
considered the end point of a relationship of “imputation,” which is as
central to the normative sciences as causality is to the natural sciences.
Thus, a criminal act may be “imputed” to a man when he has violated a
norm. A punishment is then imputed to his crime; “by imputation we
understand every connection of a human behavior with the condition
under which it is commanded or prohibited in a norm.”25 In this fashion,
by substituting cause and imputation for choice and deliberation, Kelsen
believes himself to have given a pure, scientific account of the law without
confusing the science of norms with natural science, or allowing any
illegitimate references to human purposes or understanding.
Nevertheless, Kelsen recognizes that he must explain an apparently
serious difference between imputation and causation. Whereas in a causal
relationship described by a law of nature the effect necessarily follows the
cause, in a relationship by imputation the connection may or may not
come about because the judge may or may not find the defendant guilty.
Kelsen solves this difficulty by saying that when a man lies or speaks the
truth, he is causally determined. But he is not determined by a law of
nature “according to which one must always speak the truth or always
lie.” He is determined rather by “another law of nature,” for example, by

24 25
Kelsen, Pure Theory of Law, 94. Ibid., 92.
Edward Levi, Hans Kelsen, H. L. A. Hart 211
“one according to which man chooses that behavior from which he
expects the greatest advantage.”26 Causality, then, is not incompatible
with uncertainty about the relation between two events, and imputation is
not essentially different from causation.
The shortcoming in Kelsen™s solution is that this second “law of
nature” is a rhetorical flourish covering a self-contradiction. For if a
man “chooses,” he is not determined by a law of nature; he deliberates
and decides. If he is “determined” by a law of nature, he is not under-
standing and considering alternatives, and so he cannot be choosing.
Moreover, if every choice is determined by a consideration of what will
produce the greatest advantage, there can be no objective judgments and
no way of distinguishing legal decisions from the pursuit of self-interest.
Kelsen™s Pure Theory of Law restored the equation of law with imper-
sonal rules that the Realists repudiated. But on the other hand, Kelsen
reenforced the Realist view of law by excluding from a true understanding
of law any moral judgments both about the legal system as a whole and
about particular laws. By coupling his definition of law as a set of
“norms” with a prohibition of “value” judgments, he suggested that
seeing the law as a set of rules requires denying that considerations of
justice have any relevance to the law.

H. L. A. Hart
H. L. A. Hart set himself the same task as Kelsen, but phrased it in more
matter-of-fact terms: “to give an account of what it is for a legal system to
exist.”27 And his style reflects the fact that his affinity is with Bentham
rather than Kant.
Although Hart accepts Kelsen™s criticism of Austin, he offers different
corrections. The command theory of law, he says, affords no way of
distinguishing the gunman™s order from the policeman™s warrant. What
matters for Hart is that although the gunman may get his way, his victim
feels no obligation to obey him, but only a need to do so in order to avoid
being shot. For the same reason Hart rejects the Realists™ view that what
defines a legal system is that it makes possible predictions of when one is
likely to suffer. In the Realist view, he argues, any obligation to obey the
law would disappear whenever one could confidently expect to avoid
punishment.28 In Hart™s view, the existence of a legal system implies that

Ibid., 95.
H. L. A. Hart, The Concept of Law (Oxford: Clarendon, 1961), 110.
Ibid., 82.
212 The significance of rules

the citizen has an obligation to obey regardless of whether he will or will
not in fact be punished for disobeying. This obligation cannot be ex-
plained by what citizens do, but only by how they understand their
relation to the laws. A description from an “external point of view”
necessarily ignores this “internal aspect” of rules because it can only
indicate regularities of conduct. It cannot show why a man who commits
a crime may not only predict a “hostile reaction,” but may also recognize
a “reason for hostility.”29
Hart criticizes both Austin and Kelsen for having too restricted a view
of what law does. The former, in his Command Theory of law, reduces
law in effect to a policeman halting a motorist at an intersection; Kelsen
reduces law to telling officials what to do if somebody commits a crime.
Hart regards Kelsen™s view as an improvement on Austin™s, but holds that
Kelsen, too, has fallen far short of an adequate analysis because he fails to
recognize two other kinds of laws.
The first kind gives private persons the security of knowing that the
contracts, wills, and other private arrangements they make in accordance
with the law can be enforced in the courts. Such laws are enabling acts
granting new powers to private persons rather than making any behavior
criminal or obligatory. They set conditions for engaging in certain activ-
ities, but they command nothing. By treating laws of this kind as parts of
an order to officials to apply sanctions, Kelsen obscures the fact, so Hart
argues, that such rules are “an additional element introduced by the law
into social life over and above that of coercive control.” They are designed
not for the “bad man,” but for any “man who wishes to arrange his
Besides, Hart distinguishes rules of a totally different kind because they
are rules for making or recognizing rules. Primitive societies make do with
only “primary rules,” which govern the activity of the members directly.
But a legal system appears only when primary rules are supplemented by
“secondary rules,” whose function it is to eliminate the uncertainty that
arises as soon as primary rules become at all complicated. Hart identifies
three kinds of secondary rules. A “rule of recognition” is fundamental
because it specifies marks by which rules may conclusively be known to
have the force of primary rules. By connecting what would otherwise be
discrete unconnected rules into a unified set, the rule of recognition
introduces the idea of a system. When the rule of recognition is supple-
mented by “rules of change,” it becomes possible to alter old primary
rules or introduce new ones in a deliberate fashion. Lastly, there are “rules

29 30
Ibid., 88. Ibid., 39“40.
Edward Levi, Hans Kelsen, H. L. A. Hart 213
of adjudication”31 that empower certain persons to determine authorita-
tively whether a primary rule has been adequately observed. They indicate
who is to decide such questions and according to what procedures. In
other words, they define concepts such as judge, court, jurisdiction, and
All these distinctions enlarge the notion of law from that of a ground
for private litigation or prosecution to a manner of controlling, building,
and planning “life out of court.”32 The old notion of law as coercive
orders or rules makes it impossible to see this larger function of law. For
the coercive aspect of law, in Hart™s view, is only incidental to its general
task of introducing or modifying “general standards of behavior to be
followed by the society generally.”33
Identifying secondary rules is also vital because they help explain the
idea of validity, which is as central to the concept of law for Hart as it is
for Kelsen. Whereas Kelsen resorts to a hypothetical “basic norm,” which
is itself not a legal phenomenon but a “presupposition” of legal phenom-
ena, Hart points to definite rules or procedures on which the validity of
laws in a given legal system is based. Moreover, this distinction enables
Hart to keep the validity of a legal system independent of its efficacy. For
Kelsen, the obligation to observe the laws of a legal system disappears
once the legal system ceases to be effective; Hart argues that if laws have
been made in accordance with established secondary rules, they remain
valid, even when the legal system is no longer efficacious. By emphasizing
the internal aspect of law, he is saying that what matters is not only how
laws are made, but how laws are understood to be made. This claim
suggests that “beliefs” about law are as important as observable phenom-
ena such as orders and acts and carries intimations of a concept of
Hart also departs from Kelsen by admitting considerations of justice
to the concept of law and emphasizing the distinction between what the
law “is” and what it “ought” to be. The Scandinavian Realists are
wrong, according to Hart, when they declare words such as ˜just™ and
˜unjust™ to be devoid of meaning. The fact that “concepts like justice”
depend on “implicit varying and challengeable criteria,” Hart says, “does
not render them meaningless when applied to law.” They are like other
variable standards such as those indicated by “long,” “short,” “genuine,”
“false,” and “useful.”34 The idea of fairness, that is, of treating like cases
alike and different cases differently, Hart considers essential to law.

31 32
Ibid., 94. Ibid., 39.
Ibid., 43.
H. L. A. Hart, “Scandinavian Realism,” Cambridge Law Journal 17 (November 1959): 235.
214 The significance of rules

Besides, Hart holds that something can be salvaged from theories of
natural law. But they have to be disentangled from the “teleological point
of view,” which is “latent in our identification of certain things as human
needs which it is good to satisfy and of certain things done to or suffered
by human beings as harm or injury.”35 A minimum natural law can be
constructed without resort to metaphysical notions simply by considering
“contingent” facts “which could be otherwise” but in general are not.36
Such a fact is the desire of men to survive. We cannot demonstrate the
necessity of this desire but can agree that it persists.
There are a number of such “natural facts” that have a “rational
connexion” with certain legal and moral rules. This connection, Hart is
careful to point out, is “not mediated by reasons; for they do not relate the
existence of certain rules to the conscious aims or purpose of those whose
rules they are.”37 But as men are vulnerable creatures, they must exercise
mutual forbearance and compromise if their lives are not to be nasty,
brutish, and short. As their altruism is limited, and the things that they
need in order to survive are relatively scarce, it follows that some form of
property arrangements must exist. Then, too, the division of labor brings
with it a need for dynamic “rules enabling men to transfer, exchange, or
sell their products” and recognize promises. And as human understanding
and strength of will are limited, there must also be “sanctions” to “guar-
antee that those who would voluntarily obey shall not be sacrificed to
those who would not.”38 These conclusions are all untainted by meta-
physics or reasons or purpose; they are “simple truisms,” Hart maintains,
that “disclose the core of good sense in the doctrine of Natural Law.”
They can be admitted to a positivist theory because they modestly
recognize the rational connections of certain facts, contingent though
persistent, with certain laws.39
Hart will have none of the Realists™ skepticism about rules of law. His
answer to their charge that rules of law must be meaningless because they
do not produce perfectly predictable decisions rests on a distinction
between the “core” and “penumbra” of a law. By the “core” Hart means
those particulars that are clearly known to belong to the class of acts
or things designated by a rule of law “ what he calls “the standard case;”
in the “penumbra” he includes those particulars that were not in the
minds of the legislators but are offered later as new candidates for
inclusion. The Realists make the mistake, according to Hart, of neglecting

35 36
Hart, Concept of Law, 186. Ibid., 188.
37 38
Ibid., 189. Ibid., 192“93.
Ibid., 194.
Edward Levi, Hans Kelsen, H. L. A. Hart 215
the “core” in every rule of law and supposing that a rule consists only of
the “penumbra,” although he agrees that what the Realists say about a
law applies to the penumbra.40

<< . .

( 38)

. . >>

Copyright Design by: Sunlight webdesign