II
The pattern of discourse about the concept of law
in modern legal philosophy emerges in the nineteenth
century with the work of the English jurist, John
Austin. Austin described law as a set of general com-
mands issuing from a sovereign. The sovereign he
defined as a determinate human superior who receives
habitual obedience from the bulk of a given society
and is not himself in the habit of obedience to any
superior. The command of the sovereign is charac-
terized by the sanction which is held out as a threat
in the event of noncompliance and such a command
backed by a sanction imposes a duty on the citizen.
Command, sanction, and duty are thus key terms in
the Austinian scheme.
Austin was bent on freeing the concept of positive
law from entanglements and confusion with notions
of justice and natural law. Not only did he select hard
and concrete key terms for his description of law but
he also insisted explicitly on the separation of law and
morals. “The existence of law is one thing; its merit
or demerit is another. Whether it be or be not is one
enquiry; whether it be or be not conformable to an
assumed standard, is a different enquiry” (1954, p. 184).
This severance of the realms of law and morality has
characterized a continuing school of legal philosophy
which is sometimes knows as analytical positivism,
signifying its preoccupation with the analysis of the
content and structure of law as found (positum) in a
given legal system. Austin's position on this issue is
reiterated in the work of the most distinguished con-
temporary analytical positivist, H. L. A. Hart. But
while the positivist thesis on the separation of law and
morals has held firm, there has been radical revision
since Austin's time of the terms used in elucidating
the nature of positive law.
Here the foremost architect of the modern positivist
position has been the Austrian legal philosopher, Hans
Kelsen, who has lived for many years in the United
States. Kelsen, in two celebrated works, Allgemeine
Staatslehre (General Theory of Law and State, 1925)
and Reine Rechtslehre (The Pure Theory of Law, 1934),
departed from Austin's attempt to describe law in
terms of a human commander laying down rules for
subjects and substituted as his key concept the notion
of laws as consisting of normative ought-propositions
which, in a legal system, are all linked together and
acquire unity through their common derivation from
a basic ought-proposition or set of propositions which
he called the Grundnorm. The most concrete and par-
ticular propositions of law in a legal system ultimately
derive their validity through a process of tracing back
to the basic norms of the system. So the proposition
that X ought to pay Y $100 may be valid because it
is contained in a contract duly made in conformity with
general rules of the legal system which prescribe how
binding agreements may be made. These general rules
in turn are valid because they are contained in a statute
or in decisions of the courts. The statute or decisions
of the courts are valid because they have been enacted
or decided in conformity with constitutional provisions
which prescribe the proper procedures for enacting
statutes and for appointing judges with definitions of
their jurisdiction and powers. If we ask why the provi-
sions of the constitution are valid we must, according
to Kelsen, simply accept as necessary for compre-
hending the existence of a legal system the proposition
that the provisions of the constitution ought to be
complied with.
It is apparent that when Kelsen uses the term “valid”
with reference to a particular rule of the system it has
no connotation of approbation or moral approval but
signifies only that the rule has been identified as be-
longing to the system by the criteria of recognition.
To speak of the basic norm as “valid,” however, intro-
duces an element of confusion, since this cannot be
a question of identification by further formal criteria
of recognition, but must refer either to an empirical
observation about actual acceptance in society or to
a moral precept that functioning coercive orders ought
to be obeyed. The failure to clarify the precise import
of his assertion that the basic norm has validity has
been a source of difficulty with Kelsen's theory of the
nature of law.
With respect to the relation between law and morals
Kelsen is squarely within the positivist tradition. In
What is Justice?... (1956, p. 4) he tells us that
questions of justice “cannot be answered by means of
rational cognition,” and takes up a thoroughly non-
cognitivist position in ethics, asserting that choices
about values and ends ultimately rest on intuitions. His
basic concept of the Grundnorm can encompass the
totalitarian society as easily as the democratic, vicious
and depraved laws as well as just and beneficent ones.
Kelsen's system is a powerful demonstration of the
unity and scheme of action of a legal system. His
pyramidal image of a set of norms linked ascendingly
to a basic norm reveals the essentially common features
of the legislative and judicial roles, for both judge and
legislator are creating new legal norms while at the
same time drawing upon and applying superior norms
which confer validity upon their actions. Just as Austin
insisted on the central place of sanctions in a legal
system, so does Kelsen find the distinctive element of
law in the element of coercion institutionally applied
through the normative structure. For Kelsen all legal
norms are directives to officials to apply force in cer-
tain prescribed circumstances though this may not be
superficially obvious. For example, a rule that directs
that a will should have two witnesses appears to say
nothing directly about the imposition of coercion. For
Kelsen, however, the aspect of the rule which gives
it a legal character is to be found in the proposition
that coercion will be applied to those who seek to act
in defiance of the terms of a valid will. This, in Kelsen's
scheme, is the primary rule, and the direction to pri-
vate citizens about how they should make a will is a
secondary or derivative rule. The terms “secondary”
or “derivative” here do not imply any sense of prece-
dence or superiority but are only a figurative way of
expressing the notion that the distinctive characteristic
of a legal rule is in its reference to the prescribed
circumstances for the application of institutional force.
The most powerful and subtle contemporary expo-
nent of analytical positivism is the English jurist, H.
L. A. Hart. In his book, The Concept of Law (1961),
Hart offers a devastating critique of Austin's attempt
to elucidate the nature of law in terms of a human
superior issuing commands, backed up by sanctions
which create duties. This elucidation, Hart argues, will
not serve to explain the nature of laws which confer
powers (such as the power to make a will) and which
cannot be seen as imposing duties, while the notion
of law being founded in the habit of obedience to a
sovereign commander does not explain the continuity
of a legal system which, by the operation of basic
constitutional procedures of succession, proceeds un-
interruptedly after the death of the head of state. Who,
after all, are those determinate human beings whose
commands the law could be said to be? The members
of the legislature know only a little of the law and
are themselves bound by the law. (Similar criticisms
of the Austinian position have been made by Scandi-
navian jurists, notably Karl Olivecrona.)
Hart suggests that the key to understanding the
nature of a legal system is to distinguish between what
he calls primary and secondary rules. Primary rules
are those which impose duties and secondary rules are
those which confer powers. It is the union of primary
and secondary rules which gives a legal system its
dynamic, highly structured, and rapidly creative char-
acter as compared with a body of customary rules.
Secondary rules are rules about rules. They provide
procedures for the creation, modification, and abroga-
tion of primary rules. At the base of a legal system
we find secondary rules which are fundamental rules
of recognition and which embody the constitutional
procedures for valid lawmaking in the system.
It is apparent that Hart's analysis owes a great deal
to the earlier work of Kelsen but it differs in some
significant aspects. For Hart the basic rules of recogni-
tion are not described in terms of validity which Kelsen
used in constructing his concept of the Grundnorm.
The existence of a basic rule of recognition is presented
rather as an empirical phenomenon evidenced by the
actual acceptance of the rules in a given society. The
notions of obligation and duty are also analyzed by
Hart in more subtle and complex terms than Kelsen's
reduction of all legal rules to a uniform pattern of
directives to officials about the application of coercion.
Hart elucidates the meaning of statements about duty
and obligation in the context of a legal system as
involving social practices of reference to certain stand-
ards. In the light of these standards we justify criticism
and condemnation of the behavior of others and the
application of sanctions to them, and we offer reasons
to explain and justify our own behavior. The mainte-
nance of a general system of coercion in society no
doubt psychologically sustains feelings of obligation,
but statements of obligation are not simply statements
of the probability that coercion will be applied. Our
ordinary speechways evidence this, for we do not cease
to speak of a person as being in breach of an obligatory
rule simply because he has effectively removed himself
from the jurisdiction and so from any threat of sanction.
Statements of obligation do entail a general acceptance
in society of the basic rule which is taken to validate
the primary rules which formulate particular duties,
but this is to be distinguished from an individual's
acceptance of any particular rule. So if I say that X
has broken his legal obligations by smoking opium, this
does imply my recognition that the rule against smok-
ing opium (primary rule) is properly derived from the
constitutional procedures for lawmaking in the juris-
diction (basic rule). But it does not logically entail the
prediction that X will probably be prosecuted and
punished, and it says nothing at all about what I or
X may feel about the sense and wisdom of the particu-
lar law in question.
