CHAPTER 2

Law as Command
2.1 Chapter overview
Law is a way of guiding human conduct, but it is not the only way. Qur lives are also governed
by other standards, such as those of morality, religion and etiquette. What, then, is distinctive
or special about law? What differentiates law as a type of social institution from other social
institutions? This question is at the heart of many of the jurisprudential debates that will be
discussed in this book.

This chapter will explore the 'command theory' of law as articulated by the nineteenth-century
British jurist John Austin. This is the theory that law is distinguished from other social
standards, such as morality and religion, by the exercise of force: Although Austin was not the
first theorist to defend the command theory Jeremy Bentham and, before him, Thomas Hobbes
had spoken of law in similar terms—Austin is generally acknowledged as having provided its
most influential exposition.

This chapter will also examine the criticisms of the command theory put forward by H L A Hart,
who is a key figure in contemporary legal philosophy. Hart argues that Austin's account of law
is incapable of explaining some central features of legal systems. Notwithstanding their
differences, both Austin and Hart are positivists. The chapter will conclude with a discussion of
some key positivist claims.

2.2 Austin's command theory of law

In proposing the command theory of law, Bentham and Austin were reacting in part to natural
law ideas, which at this point can loosely be defined as maintaining that law is necessarily
linked to justice. William Blackstone, who wrote an influential eighteenth-century treatise on the
common law, was a particular target of their criticism. Blackstone tended to present the
common law as the embodiment of reason and Christian moral principles, whereas Bentham
and Austin were legal reformers who wished to demystify law by providing an accurate
description of law with all its imperfections. This, they thought, would provide the best basis for
criticizing and improving the law.

Whereas Blackstone had blurred the line between dispassionate analysis and moral
evaluation, Bentham made a clear distinction between what he called `expository
jurisprudence' and 'censorial jurisprudence'. Expository jurisprudence is concerned with
analyzing the law as it actually exists, without regard to its justice or injustice. Censorial
jurisprudence is a critical and moral enterprise designed to improve or reform the law. Austin
made a similar distinction between general jurisprudence and the art of legislation.
Conceptualizing law as the exercise of power was an important plank in this overall project.
Obviously, if law is the exercise of power, law is not guaranteed to be the embodiment of
reason, for power can be exercised for both good and bad ends.

In his set of lectures entitled The Province of jurisprudence Determined, published in 1832,
Austin begins by distinguishing 'laws properly so called' (laws intended to guide behaviour)
from 'laws by analogy' (such as the laws of fashion or honour) and 'laws by metaphor' (these
being the laws of science).2 He then turns in more detail to the category of laws properly so
called. All laws properly so called are commands, claims Austin, a command being an order
backed up by a `sanction' (a threat of harm) in the event of non-compliance with the
command.3

Some commands, Austin tells us, are general. These are commands that prescribe types of
conduct, directed to classes of persons. Other commands are directed to individual people.'
But all commands issue from a superior, that is, a person or group of persons possessing the
power to inflict harm. Austin writes: `[T]he term superiority signifies might: the power of
affecting others with evil or pain, and of forcing them, through fear of that evil, to fashion their
conduct to one's wishes.' Austin then goes on to draw a further distinction between commands
that issue from God and commands that issue from humans. Of those that issue from humans,
some, he says, are laid down by the sovereign in a state, while others, such as the commands
of a father to his child, are not.

Who or what is the sovereign in the state? The sovereign, Austin explains, is that person (such
as an absolute monarch) or body of persons (such as a democratically elected parliament)
whose orders are habitually obeyed by the bulk of society and who does not habitually obey
any other person or body.'

Austin goes on to say that only those general commands that emanate from the sovereign in a
state are laws 'strictly so called' or 'positive laws', and it is these laws that constitute the
subject matter ofjurisprudence.8 This is Austin's simple answer to the question with which this
chapter began, namely: how can we distinguish legal standards from the many other kinds of
standards that regulate and govern human conduct? Legal standards, on Austin's view, are the
commands of a supreme political superior or sovereign, which are backed up by the threat of
sanctions. Here we have what Austin calls 'the key to the science qfjurisprudence'. 9 Hart
explains that, on Austin's view, 'in every human society where there is law, there is ultimately
to be found ... [a] simple relationship between subjects rendering habitual obedience and a
sovereign who renders habitual obedience to no one'. 10 Where this vertical structure is not
present, we cannot, according to Austin, speak of law.

Austin's views about the nature of law are not of only abstract interest. Consider the Australian
case of Mllirrpurn v Nabalco Pty Ltd (the Gove Land Rights case),11 in which Austin's account
of law was directly relevant to the legal issues and the court therefore had to confront the
question of its adequacy. The plaintiffs were certain Aboriginal clans who claimed that a mining
company had unlawfully interfered with rights they had under Aboriginal customary rules to
occupy and enjoy certain areas of land in the Gove Peninsula in the Northern Territory.

Blackburn found that the clans in question had a religious basis and a connection with the
land. They did not, however, have an internal organization, were not ruled over by a chieftain,
and were not in control of a definable territory. This raised the question whether in their world
there was anything recognizable as law, because, if not, their relationship to their land could
not amount to a proprietary right and their argument would have floundered at the first hurdle.
On Austin's view, the answer would be `No', since there was no identifiable sovereign authority
with a capacity to enforce the clans' claims to land. In fact, the defendants implicitly relied on
Austin's view, arguing that, in the absence of any machinery for enforcement, the clans'
customs could not be described as legal rules and therefore that no invasion of legal rights had
taken place. But is Austin correct that force applied by the state is the defining characteristic of
law? The Federal Court rejected his view, and as we will see, there are powerful arguments in
favour of the position it took. In fact, there are few contemporary legal theorists who accept the
command theory of law.

In order to understand why this is so, we need to take note of an important distinction: that
between normative statements and descriptive statements. Normative statements tell us what
we have a right to do, what we ought to do and what we ought not to do. They aim to guide our
conduct and to provide a standard for its appraisal. They are therefore 'practical' or
`prescriptive' in the sense that they prescribe how we should behave. Normative statements
can be contrasted with descriptive statements, which do not tell us what we ought to do but
instead describe the way the world is.

Austin's analysis of law treats propositions flaw, such as propositions about the legal validity of
a rule and propositions about legal rights and duties, as descriptive or predictive statements
about the empirically observable behavior of the sovereign and the sovereign's subjects. In
other words, Austin thinks that statements of what individuals are legally required to do can be
reduced to non-normative, factual statements about the giving of orders, habits of obedience,
and the likelihood of being punished. Austin therefore sees legal concepts as ordinary factual
concepts, not normative concepts. His account makes no reference, for instance, to the
lawmaker as the person or body with the authority or entitlement to make laws, or to
statements of law as statements that tell people what they ought to do, or to breaches of legal
duties as justifying sanctions; Instead, he analyses statements of what individuals are legally
required to do in non-normative, descriptive terms, namely, in terms of the factual likelihood of
being harmed and facts about commands and obedience.' 2
Austin took this approach because he wished to extend the methods of science to the study of
social phenomena such as law. This led him to analyze law from the outside, in terms of
empirically observable regularities in behaviour, rather than from the inside, by reference to the
role played by legal rules in the internal reasoning processes of those who are committed to
legal institutions and practices. H L A Hart is the theorist who is best known for exposing the
difficulties in Austin's behaviourist approach. As we will see, although both Hart and Austin are
legal positivists, their versions of positivism are very different. Let us begin with Hart's reasons
for objecting to Austin's approach.

2.3 Can threats place us under obligations?

On Austin's picture, as Hart points out, law is tv be found in the 'gunman situation writ large'. 13
A gunman who accosts you, saying 'your money or your life', achieves what he wants by
threatening to kill you. We say that you have been coerced or forced to hand over your money.
Law, for Austin, is coercive in the same way. The only points of difference are that the
gunman's orders are directed temporarily at a particular individual and the gunman is not the
supreme source of commands in the society, whereas for Austin, laws are general, are obeyed
over time, and issue from the sovereign. Nevertheless, both the gunman's command and the
dictates of the law can be seen as orders backed up by a threat of harm. For Austin, force or
coercion is the essence of law: the state is a gunman on a large scale.

Hart objects to this picture on a number of grounds. For one thing, he argues that threats are
not capable of giving rise to obligations. Hart draws a distinction between 'being obliged' and
'being under an obligation'. If the threat of being killed leads me to hand over my money to a
gunman, then we would say that am obliged to hand over the money, in the sense of having no
real choice. But, says Hart, we would not say that I was under an obligation to do so was duty-
bound to do so.' Nor would we say that the gunman has authority to tell me what to do. 15

The notion of authority will be considered in detail in 4.6, but for the moment it will be sufficient
to make the following slightly simplified points. Authority is a normative matter. Someone who
has authority over me has the normative capacity to bind me by issuing directives with which i
am under a duty to comply just because of the person's status as an authority, and no matter
what I may think of the merits of the directive.' The gunman exercises control over me, not
authority. His threats do not impose a duty on me and I am not bound to obey him. He merely
influences my actions by making it prudent for me to do as he orders, and if someone were to
disarm him, I would have no further incentive to give him my money. Furthermore, if for some
reason I were to resist the gunman, while this would no doubt lead to my suffering a harm, no
one would suggest that the harm was justified by my resistance. It makes no difference
whether it is the state or a lone villain who plays the role of gunman. Threats backed up by
sanctions may affect my actions, but they cannot create an obligation for me to comply or
make it justifiable to punish me for failure to comply.

The law, by contrast, does speak in the normative language of rights and obligations. Although
both law and threats are intended to make certain forms of conduct non-optional, and although
law may be backed up by threats of harm, law's primary intention-is that we should take its
directives as authoritative. Law claims, in other words, that it has authority to tell us what to do.
It tells us that we are bound to obey its directives. When, for instance, judge’s order someone
to pay money as compensation for wrongdoing, they claim, as bank robbers do not, that the
person is under an obligation to comply. We are supposed to obey the law not because we will
be punished if we fail to do so but just because it is the law.

But what are these obligations that law purports to impose on us and what sort of authority to
create obligations does the law claim to have? Hart argues that they are legal obligations, not
necessarily moral obligations, and that the law need not claim morally legitimate authority.
Joseph Raz, by contrast, thinks that law's claim to authority is a claim to morally legitimate
authority and that the law therefore claims to impose moral obligations on us. He thinks that
the law asserts that we are under a moral obligation to obey the law just because it is the law.
We will examine the dispute between Hart and Raz in 3.7. At this point, however, we should
take note of the fact that, although Raz thinks that the law necessarily claims moral authority,
he also thinks that its claim to such authority is frequently unwarranted. We will examine his
reasons for saying this in 4.6. For the moment we simply need to be aware of the difference
between saying that the Law necessarily claims moral authority and saying that it necessarily
has moral authority. Raz makes the former assertion about the law, not the latter.

Finally, to say that the law claims authority is not to say that everyone who obeys the law
accepts the claimed authority of the law. Some people may obey the law because it happens
to coincide with what they independently think is right. Mark Murphy explains this point as
follows:
[T]he person who acts under authority goes along with what the authority demands. [lone is
told to do something; and then deliberates about what to do, then one is not acting as
someone under authority, even if one ends up doing just what the authority told one to do. VI
tell my daughter to clean her room, and she thinks about how messy her room is, and why
now is a good time to do it, and then decides to go ahead and do it, she is not acting as
someone under my authority—she is just making up her mind that, on the merits, cleaning her
room is the thing to do.)18

Others may obey the law out of fear of sanctions. In this case, as Hart says, sanctions are
intended to provide a motive for conformity, for those who refuse to accept the authority of the
law. But sanctions merely function as a back-up in cases where the system has failed and the
law's directives are not accepted as giving rise to a duty to obey. 19
2.4 Can Austin account for certain familiar features of legal systems?
Even if commands backed up by threats were capable of imposing obligations, Austin's
account of law is too simple, according to Hart. It ignores the variety of laws and therefore
imposes on law a 'spurious uniformity.’ In particular, it is not the case that all laws order people
to do or not to do certain things on pain of sanction. Austin's picture comes closest to
describing criminal laws, which in some ways resemble commands backed up by a threat of
punishment. But even criminal laws cannot be seen as analogous to a gunman's threats,
because a gunman's orders do not apply to himself, whereas criminal laws have to be obeyed
even by those who make them.21 Thus it is not appropriate to conceptualize even criminal laws
as Austin's account would.

Furthermore, Hart argues that there are many laws that are even more difficult to force into
Austin's duty-imposing model. Consider laws that confer power, whether on private individuals
or officials. Such laws do not impose duties or demands that require us to behave in certain
ways, whether we wish to do so or not. Rather, they provide the means to realize our wishes,
and they enable officials to act in certain ways. Far from reducing our freedom, they enlarge it.
There are, for instance, laws that empower us to make a will, or to enter into a contract or a
marriage. And there are laws that confer power on courts to hear certain matters and on
legislatures to make laws. Failure to observe the requirements laid down by these laws leads
to legal invalidity, not a sanction. If, for instance, I fail to sign my will, I will not be punished. I
will merely have failed to make a valid will. Likewise, if there is a law conferring power on a
legislative body to pass laws in accordance with certain procedures, an attempt to pass a law
that does not comply with the procedures will not be punished. It will merely be ineffective. Nor,
according to Hart, can it be said that legal invalidity is a kind of penalty, for the purpose of
sanctions is to discourage certain conduct, whereas legal invalidity is not used for that
purpose. Hart concludes that the command theory of law is incapable of adequately
accounting for private and public power-conferring rules.'

Hart turns next to Austin's account of sovereignty in terms of habits of obedience and argues
that there are other typical features of legal systems that it cannot explain. One such feature is
that legal systems make provision for the uninterrupted continuity of lawmaking power when
one lawmaker succeeds another. When a new set. of legislators take their seats in parliament,
the first bills they pass are laws. When an absolute monarch's son succeeds him, the son's
first orders are laws. But, for Austin, we can talk of law only when there is a habit of general
obedience. It follows that when one lawmaker succeeds another there is, for Austin, no law
until the populace has acquired the habit of generally obeying the new lawmaker. There, will, in
other words, be an in-between period in which law cannot be made. Hart concludes that
Austin's account therefore cannot explain a central feature of legal systems. This is the fact
that there is a seamless, transfer of authority to the successor lawmaker: the first law made by
the new lawmaker is already law, despite the fact that the new lawmaker has not yet received
habitual obedience.'

Another familiar characteristic of legal systems for which Austin's concept of sovereignty
cannot account, according to Ham is the fact that laws that were made by an earlier lawmaker,
now dead, can still be valid law even though that lawmaker is, by definition, no longer
habitually obeyed. When one lawmaker replaces another there is no reason to fear that all the
old laws have disappeared. A law enacted 100 years ago can still be law today.' Yet Austin's
theory has obvious difficulty in explaining this phenomenon. How can the orders of previous
lawmakers still be law when those lawmakers are no longer habitually obeyed? The only way
in which Austin can attempt to account for this phenomenon is to say that if the new lawmaker
has not expressly repealed the ad statute, then this amounts to 'tacitly' commanding it. In
particular, by not interfering when the courts enforce the old statute, the new lawmaker has
given a tacit order that the old law should be obeyed. The trouble with this response though,
-
as Hart points out is that it implies that the old law is not law until it us applied by the courts
during the tenure of the new sovereign. But this is false. A statute does not become law only
after it is applied by the courts. It is already law before any cases arise to which it may be
applied.

Other features of Austin's concept of sovereignty to which Hart objects are the indivisibility of
the sovereign power (the fact that, for Austin, there can bp only one sovereign body or one
sovereign individual) and the sovereign's legally illimitable status (the fact that the sovereign is
outside the law). Insofar as the latter is concerned, since Austin's sovereign by definition does
not habitually obey any other person or body, there can be no legal limits on what the
sovereign can command and no laws that the sovereign cannot make. According to. Austin,
this must be the case in all societies in which there is law: it is a necessary condition of law.
Austin writes: “Supreme power limited by positive law, is a flat contradiction in terms ... Every
supreme government is legally despotic.”

As Hart points out, however, there are many legal systems in which we do not find a sovereign
who is above the law. Austin's definition of sovereignty may seem to fit British constitutional
arrangements in which there is no written constitution limiting the legislative powers of the
Westminster parliament and (leaving aside certain complications arising from British
membership of the European Union) the doctrine of unlimited parliamentary sovereignty has
long been regarded as a fundamental rule. But there are many other countries in which the
legal situation is different. Furthermore, there are many systems in which sovereign power is
divided among different organs.27

In both the United States and Australia, for instance, there is a written constitution, which not
only divides power between the federal parliament and the Legislatures of states, but also
prevents legislative interference with certain individual rights and entrenches the separation of
powers. To give just one example, s 116 of the Australian Constitution prohibits the federal
parliament from interfering with the free exercise of religion. This means that if the federal
parliament attempted to pass a law with such a purpose, the High Court would strike it down
as invalid. In such a case, the parliament would have failed to make a valid law. But although
the power of the federal sovereign is subject to this and other legal limitations, it is obvious that
Australia has a legal system.

At this point, someone sympathetic to Austin's view might point out that in his later work Austin
did not identify the sovereign in a democracy with the legislature, but with the body of electors
—a body perhaps more plausibly thought of as necessarily free of legal limitation. Thus, if we
consider the Australian Constitution again, we find that it contains a provision, s 128, that
provides for alterations to the Constitution. Leaving aside certain special cases in which
proposed amendments must pass an additional hurdle, the normal rule under s 128 is that a
proposed amendment has to be approved by a majority of voting electors in a majority of the
states and by an overall majority of voting electors. This seems to suggest that the true
sovereign in Australia is the people—not, as we first assumed, their elected representatives in
parliament. And if it is the people who are sovereign—the people who issue the commands
that provide for the legislative powers of the federal and state parliaments and for the limits on
those powers—have we not found a body whose power is legally illimitable, in which case
Austin would be vindicated, at least in this respect?

Hart's reply is to argue that if it is the people who are sovereign, then it is impossible to
understand their sovereignty on Austin's original model of commands issued by a political
superior and backed up by sanctions. After all, who is doing the commanding and who is doing
the obeying? It seems that if the people are sovereign, and if Austin's account of sovereignty is
correct, then the people must be issuing commands to themselves, which they themselves
habitually obey. They must, in effect, be holding a gun to their own heads—an idea that is hard
to comprehend.28

In summary, Hart argues that Austin's theory of law as the will of a sovereign ruler is too crude.
An adequate theory of the nature of law must explain the central features of legal systems, but
there are many features of legal systems that Austin's theory lacks the resources to explain. It
cannot, for instance, account for the fact that there are many rules of law that do not have their
source in a lawmaker's will; that a lawmaker's powers are conferred by law and can be limited
by law; that legal powers may be used in ways that are not equivalent to prohibiting conduct;
and that, when powers are used in such ways, what has been done can be assessed for legal
validity or invalidity.

Hart does not deny that, in practice, some degree of coercion is needed to sustain a legal
system and that many people do, in fact, comply with legal demands out of fear of sanctions.
He insists, however, that coercion does not play a central role in a theoretical account of law.
As we will see in the next chapter, Hart thinks that law is not a matter of mere regularities in
behaviour, but a matter of rules. Hart argues that once we see that law consists in rules, we
will be able to explain the features of legal systems that Austin cannot explain.

2.5 Positivism: some preliminary points

Both Austin and Hart are legal positivists. The debate between legal positivists and those who
are, broadly speaking, in the natural law tradition, will be a recurring theme in the chapters that
follow. It is nevertheless necessary to introduce some relevant concepts at this stage, in order
to lay the foundation for subsequent discussion. It needs to be appreciated at the outset,
however, that there are many versions of positivism and this makes it difficult to single out
certain elements as the 'defining' features of positivism. In what follows, I will therefore speak
of 'central' doctrines or 'key ideas' associated with the positivist tradition without necessarily
suggesting that all positivists accept all of these ideas or give them equal prominence in their
theories. I will also highlight some of the differences between positivist theories.
One such key positivist idea is the idea, as Leslie Green puts it, that law is a 'social
construction'.29 On this view, law is ultimately a matter of 'social facts'. Social facts are facts
about the behaviour, beliefs and attitudes of people in their social interactions. 3° For instance,
the fact that the sovereign has issued a command on pain of punishment is a social fact.
Social facts are non-normative. They can be contrasted with normative facts, such as moral
facts. Moral facts are facts about what is morally right and wrong. 31 For instance, the fact that
slavery is wrong is a moral fact. It should be emphasized that moral facts are not facts about
what people believe is right and wrong but facts about what is really right and wrong. In some
societies, slavery may be thought morally acceptable. This does not, however, make slavery
acceptable. Facts about what social groups find acceptable or unacceptable are non-
normative, social facts. Facts about what is really acceptable or unacceptable are moral facts.
The view that moral judgments can be true and that they are true when they correctly report
facts is called 'moral realism' or 'cognitivism'.32 There is obviously much more that could be
said about this view. For our purposes, however, it will be sufficient to note that we do
generally think of our moral views as purporting to report facts and that we make an intuitive
distinction between non-normative facts (among which are social facts) and normative facts
(among which are moral facts). It is beyond the scope of this book to counter the skeptical
doubts of those who think that the idea of objectivity in morality is mistaken. 33
For Austin, law owes its existence to the social facts of commands, habits of obedience and
punishment. Laws are the orders of those who have power. What the law requires will
therefore be an empirical and contingent matter, depending on what has been ordered by
those in power. It is perhaps worth mentioning that Austin's view explains the origins of the
label 'legal positivism'. Another way of saying that rules owe their status as law to the
sovereign's deliberate acts of will is to say that law is put in place or 'posited' by the sovereign.
And a theory which stresses the posited nature or 'positivity' of law is naturally called a
'positivist' theory.
Although Hart agrees with Austin that law owes its existence to social facts, he has a different
view, as we will see in 3.3, about the nature of the social facts to which law owes its existence.
For Hart, law is a social practice. Social practices, as Matthew Noah Smith explains, are things
that people do together, in the sense that there is some sort of systematic unity to their
activity.34 For Hart, law owes its existence to official agreement on the criteria for identifying
valid laws and to the fact that members of the population generally obey the rules that are valid
according to the criteria of legal validity. Hart puts this by saying that legal officials follow a
'social rule' in identifying and applying the law, and he calls this rule the 'rule of recognition'. A
particular norm will therefore be a law if it conforms to certain criteria found in a special kind of
rule accepted by legal officials.

The 'social rule' account is widely accepted among contemporary positivists, although there is
disagreement as to what exactly a social rule is. One well-known answer is that a social rule is
a convention, although not all positivists agree with this answer. Here the word 'convention' is
being used in a technical sense, which will be explained in 3.3. But whatever the exact nature
of social rules, the main point for present purposes is that Hart takes the view that the law of a
jurisdiction is the set of norms that satisfy the criteria of legal validity accepted by officials in
that jurisdiction. This means that what the law requires is for Hart, as it is for Austin, ultimately
an empirical and contingent matter, depending on what criteria of validity officials happen to
agree upon.

Natural law theorists oppose the view that the existence of legal rights and duties is simply a
matter of social fact or social practice. The natural law view is that the existence of law or
'genuine' law is in some way tied to morality. Although most natural lawyers do not deny that
law has an institutional, human dimension, they believe that law is not a 'mere man-made
artifact'.35 It is something less dependent on arbitrary or optional human choices and social
practices. Law (or 'true' law) derives its authority not from human choices, or not solely from
human choices, but from being reasonable: it 'comes to [us] from something outside and
bigger than [ourselves]'.36

Another key doctrine associated with the tradition of legal positivism relates to the relationship
between law and morality. It is obvious that there are many contingent connections between
law and morality, or connections that hold as a matter of fact. Law is, for instance, frequently
morally defensible: consider laws prohibiting murder and assault. Positivists, however, deny
certain (and even, in the case of some positivists, any) necessary or conceptual links between
law and morality—links, in other words, deriving from the nature of law. Positivists therefore
say that law and morality are conceptually 'separable'. This view is for obvious reasons called
the 'separability thesis'.37
Some positivists adhere to the separability thesis because they believe that law is objective
and that moral standards are subjective and relative. Most contemporary positivists, however,
believe in the objectivity of moral standards—they believe in what I called 'moral facts'—and
they therefore have different reasons for adhering to the separability thesis. These positivists
make a distinction between 'critical morality' and 'popular morality'. 'Popular morality' refers to
the moral beliefs of a particular group or society. it is, in Hart's words, the 'accepted morality of
[a] group ... even though this may rest on superstition or may withhold its benefits from slaves
or subject classes'.38 Popular morality therefore refers to whatever beliefs a particular group
happens to think are right, whether or not these beliefs are sound. 'Critical morality', by
contrast, refers to `the general moral principles used in the criticism of actual social institutions
including [popular] morality'.' Critical morality therefore refers to sound moral principles,
whether or not these are accepted by a particular group or society. These sound principles can
be used to evaluate popular morality.

Most positivists think that popular morality is necessarily connected with law, law being one
way in which the mores of the group are expressed.4° The more interesting question, and the
one to which positivists and natural law theorists give different answers, concerns the nature of
the connection between law and critical morality. From now on all references to morality should
therefore be understood as references to 'critical morality', unless otherwise stated.
Although all positivists hold that law and morality are conceptually separable, it would be
wrong to exaggerate the uniformity in the positivist tradition. There is, in fact, considerable
disagreement among positivists as to the exact extent of the separability, and by implication,
about the significance to be attached to the separability thesis in the overall positivist project.
In particular, although positivism is sometimes said to be committed to the thesis that there are
no necessary connections between law and morality, there are many contemporary positivists
who do not accept so expansive a claim.41

All positivists are united, however, in taking a narrower view. This is that there is nothing in the
nature of law that requires laws to pass moral tests to be valid, or that requires moral facts to
play a role in determining the content of the law. The tests in a particular jurisdiction for
determining whether a rule is 'law' and for identifying the content of legal rules may refer only
to social or non-moral facts, such as the rule having been declared by a court or enacted by a
legislature. It follows from this view that legal rules are not necessarily just. For positivists,
therefore, the law as it is' is not necessarily the law as it ought to be'.

As explained, all positivists take the view that law and morality are separable, in the sense that
there is nothing in the nature of law that requires laws to pass moral tests to be valid or that
requires moral facts to play a role in determining the content of the law. There are, however,
two versions of this view. At a minimum, it implies that law may be identifiable without recourse
to moral reasoning. Some positivists, however, assert a stronger claim, namely, that law not
only may but must be identifiable without recourse to moral reasoning. We will examine this
issue in detail in Chapter 4. We will see that some positivists, who are called 'inclusive
moral contexts. We will examine these claims as well as various criticisms of
positivists', take the view that it is possiblethem
(although, of course, not necessary) for laws to be
in Chapters 3 and 5.
identified using moral tests. Thus, in particular jurisdictions, the accepted tests for legal validity
So far, we have been dealing with positivist views about the nature of
might require judges to make reference to matters of moral merit in deciding whether a law is
law. Such views fall under what is called 'substantive' legal positivism. There
valid or determining its content. Exclusive positivists, by contrast, take the view that this is
is, however, also another kind of positivism, which is called 'methodological'
impossible. They say that law must be identifiable without recourse to moral standards. They
positivism. Whereas substantive legal positivism involves claims about the
say that, necessarily, there cannot be moral tests for law.
nature of law, such as the claim that law and morality are conceptually distinct,
Austin expressly affirmed the view that laws do not have
methodological legalto pass aismoral
positivism a claimtest,
aboutsaying: 'The
legal theory or jurisprudence.

existence of law is one thing; its merit or demerit is another. Whether it be or be not is one
It is the view that theorising about the law can be a morally neutral enterprise,
enquiry; whether it be or be not conformable to an assumed standard, is a different enquiry. A
which aims merely to describe what is distinctive about law as a special form of
law, which actually exists, is a law, thoughsocial
we happen to On
organisation. dislike it.' It
this view, weis easy
can to illuminating
give an see why and Austin
fully adequate
took this view. For Austin, in order to determine whether a rule is a valid law and what its
account of the nature of law without being committed to any moral views and
content is, we have to find out what commandswithout have been
attempting issuedaby
to provide thejustification
moral sovereign andThis is
of law.
whether people are likely to be punished for disobeying them. But the fact that a sovereign has
a methodological form of positivism because it is a claim about the method of
commanded something does not guarantee the moral soundness of what has been
jurisprudence or how it should be done. Methodological debates have been a
commanded. Hence law is not necessarily good law. 47
feature of much recent work in jurisprudence.

Furthermore, since ascertaining the law's requirementsMethodologicalinvolves an do

positivists empirical inquiry
not take the intoview
implausible thethat theorising
sovereign's will, rather than a moral inquiryabout
intothethe
law can be entirely value-free. They concede that legal theorising, like all
acceptability of the commands, identifying
theorising, is an evaluative activity, in the sense that there are certain desiderata
the law is, for Austin, a different kind of enterprise from investigating whether the law is just or
or norms of theory construction, such as simplicity, clarity and coherence, by
ought to be obeyed. The first is a matter ofreference
non-moral or social facts. The second is a matter of
to which theorists are guided and the adequacy of their theories
moral facts. Although both questions havejudged.objective answerspositivists
Methodological for Austin,
are atdifferent
pains to point out, however, that values
considerations are relevant to answering them.such as simplicity and coherence are not moral values. 48

Austin thought, in particular, that the justice of the law is tested by its conformity to utilitarian
Methodological positivists also concede that a descriptive account of law
principles of the general good. (For discussion of utilitarianism, see 10.3.) Bentham's
will inevitably be based on only some facts about law, namely, Those facts that the
distinction between expository and censorial jurisprudence, mentioned in 2.2, rests on the
theorist considers central or important to law. A description of law is, as Green
same distinction between identifying the law and aevaluating
explains, it, and
selection of facts abouthe too thought
it, which are taken tothat
be interesting and
utilitarian morality was a measure of the justice of law. 49
important. This means, as Raz notes, that 'any good theory of [law] is based on
evaluative considerations in that its success is in highlighting important social

We have seen that all positivists take the view thatandthe

structures law asand
processes, it is does
every not necessarily
judgment of importance is evaluative'. ° 5

Furthermore, such evaluative judgments of importance will need to relate to the

coincide with the law as it morally ought to be. Some positivists believe, however, that the
conceptual independence of law from morality goes further. These positivists deny additional
necessary connections between law and morality. Their view is that law has no intrinsic moral
47 S R Perry has been influential in highlighting the issues about the methodology ofjurisprudence. See S
value but is rather a `morally neutral instrument which
R Perry, is as serviceable
'Hart's Methodological Positivism', in Jfor evil(ed),
Coleman as Hart's
for Postscript:
good'.'Essays on the Postscript to
The Concept of Law, Oxford University Press, Oxford, 2001, pp 311-54.
They say that there is no role morality must play in law and that any moral value or moral merit
that law in general or individual laws may have is contingent rather than necessary, depending
on the content of the law or the use to which the law is put in a particular system. This is
sometimes expressed by saying that there are no necessary connections between law and
morality—a slightly hyperbolic formulation which lays itself open to various objections, such as
the objection that both law and morality consist of norms. 44 The thesis is not, however,
intended quite as literally as the objections assume. When it is claimed that there are 'no
necessary connections between law and morality', what is meant is that there are no
interesting necessary connections.45 The 'no necessary connections' thesis should therefore be
understood as a rough or shorthand way of rejecting all of the supposedly necessary
connections which have been proposed by other legal theorists. 46

In subsequent chapters we will examine a number of possible necessary connections between

law and morality, which have been denied by some (though, as noted, not all) positivists.
Positivists have argued, for instance, that legal systems do not secure anything of moral value
merely by virtue of complying with the principles of the rule of law; that legal officials such as
judges do not have to believe (or even profess to believe) that the rules that they apply and
enforce are morally legitimate or have moral authority; and that normative terms such as
'obligations' and 'rights' have different meanings in legal and moral contexts. We will examine
these claims as well as various criticisms of them in Chapters 3 and 5.

So far, we have been dealing with positivist views about the nature of law. Such views fall
under what is called 'substantive' legal positivism. There is, however, also another kind of
positivism, which is called 'methodological' positivism. Whereas substantive legal positivism
involves claims about the nature of law, such as the claim that law and morality are
conceptually distinct, methodological legal positivism is a claim about legal theory or
jurisprudence.

It is the view that theorizing about the law can be a morally neutral enterprise, which aims
merely to describe what is distinctive about law as a special form of social organization. On
this view, we can give an illuminating and fully adequate account of the nature of law without
being committed to any moral views and without attempting to provide a moral justification of
law. This is a methodological form of positivism because it is a claim about the method of
jurisprudence or how it should be done. Methodological debates have been a feature of much
recent work in jurisprudence.47

Methodological positivists do not take the implausible view that theorizing about the law can be
entirely value-free. They concede that legal theorizing, like all theorizing, is an evaluative
activity, in the sense that there are certain desiderata or norms of theory construction, such as
simplicity, clarity and coherence, by reference to which theorists are guided and the adequacy
of their theories judged. Methodological positivists are at pains to point out, however, that
values such as simplicity and coherence are not moral values.48

Methodological positivists also concede that a descriptive account of law will inevitably be
based on only some facts about law, namely, those facts that the theorist considers central or
important to law. A description of law is, as Green explains, a selection of facts about it, which
are taken to be interesting and important. 49 This means, as Raz notes, that 'any good theory of
[law] is based on evaluative considerations in that its success is in highlighting important social
structures and processes, and every judgment of importance is evaluative'. 5° Furthermore,
such evaluative judgments of importance will need to relate to the judgments of importance
made by those who are subject to the law, because, as Raz says, 'unlike concepts like "mass"
or "electron", "the law" is a concept used by people to understand themselves'. 51 Julie Dickson
explains Raz's point as follows: ([A]ny explanatorily adequate legal theory must, in evaluating
which of law's features are the most important and significant to explain, be sufficiently
sensitive to, or take adequate account of, what is regarded as important or significant, good or
bad about the law, by those whose beliefs, attitudes, behaviour, etc. are under consideration.' 52

Thus, no doubt Austin highlighted sanctions in his account of law because the fact that we are
punished for failing to obey the law is a fact we find significant, in contrast to other facts about
law, such as the fact that judges may wear robes, which is of less or no significance to us.
Other theorists, such as Raz, highlight the way in which law asserts the authority to tell us what
to do, regardless of whether we think the law's demands are justifiable or not. It is likely that
those who find this feature of law important do so because they are concerned about a
possible conflict between law and the value of autonomy, that is, the value of not subjecting
our will to the will of another person.53

Once again, however, methodological positivists maintain that we should not confuse
evaluative judgments of importance with moral judgments of merit. In their view, the
identification of the important features of law does not rest on moral judgments concerning the
moral attractiveness or justifiability of those features.' For instance, in saying that law claims
ultimate authority over how we should behave, Raz merely wishes to report on an important
way in which law as a social institution invariably operates and impinges on us, without in any
way suggesting that it is a good thing (or a bad thing) that the law should make such a claim. 55
Analyzing law is therefore, in the view of methodological positivists, like analyzing other
concepts, such as knowledge. No one supposes that an account of knowledge requires moral
argument or must seek to justify knowledge. Similarly, methodological positivists claim to
provide a morally neutral, conceptual account of law as a distinctive form of social
organization. Such an account does not tell us whether law's central features are good or bad.
Nor does it tell us why and when law ought to be obeyed. Legal theory is an essential
precursor to answering these latter questions, but legal theory cannot itself answer them.
These questions are moral and political questions, and it is therefore the disciplines of moral
and political philosophy that provide the answers to them. This, then, is the specific sense in
which methodological positivists aim to offer a 'descriptive' as opposed to a 'normative' account
of law.

John Finnis and Ronald Dworkin are vigorous opponents of this kind of approach to legal
theory, as we will see in Chapters 5 and 6. For them, law is a morally charged concept and
they believe that an account of the nature of law therefore has moral implications for how
citizens and officials ought to behave. Consider, as an analogy, the concept of 'democracy'.
Can we arrive at a view about what democracy is without considering the moral point, function
or purpose of democracy? Must anyone who wishes to participate in the conceptual debate
about the nature of democracy have a view about the values which democracy is meant to
serve? If you answer 'yes' to this question, you will believe that 'democracy' is a morally
charged concept.

This is what Dworkin and Finnis believe about the concept of 'law'. They believe that explaining
what law is involves explaining its value. They therefore ascribe a morally valuable purpose to
law, with a view to showing how it’s serving this purpose can give citizens and officials reasons
to obey it. In the case of Finnis, law's morally valuable purpose is the coordination of conduct
for the common good. In the case of Dworkin, the purpose °flaw is to limit and morally justify
the use of state force. Either way, their conception of jurisprudence is entirely different from
that of methodological positivists, insofar as they believe that a theoretical account of what law
is, must be governed by the need to show what is morally valuable about law, such that it is
capable of generating moral obligations to enforce and obey it. For Dworkin and Finnis, in
other words, analyzing the concept of law is not a descriptive project but a morally engaged
project, and on their view, articulating a theory of law necessarily involves making moral
arguments.

Both Finnis and Dworkin deploy their normative methodology in the service of defending anti-
positivist theories of law, but it should be noted that moral arguments can also be used to
support substantively positivist theories of law, and that there are, indeed, some positivists who
do use them in this way. Once again, we need to be alert to the variety of different views that
make up the positivist tradition. Just as there are substantive differences among positivists (for
example, the difference between inclusive and exclusive positivism mentioned above), there
are methodological differences. Thus some positivist theorists, as we will see in 4.2, argue that
there are good moral reasons to conceptualize law as different from morality because, for
instance, such a conceptualization makes it more likely that we will resist unjust laws. Dworkin
goes so far as to suggest that a theory that ties law to social facts for moral and political
reasons is the only plausible form of positivism (see 6.9).

There is also, as we will see in 4.8, a prescriptive form of legal positivism. This is a theory that
recommends a rule-based form of governance, which minimizes the need for citizens and
officials to rely on moral judgments in identifying and applying the law. These theorists argue
that laws should be framed as clearly and precisely as possible and that judges should
interpret and apply legal rules according to their clear meaning because this is a way of
securing certain benefits. The benefits include public guidance and predictability and ensuring
that disputed moral questions are resolved by majoritarian decision-making rather than by
judicial decision.
We have distinguished between substantive positivism and methodological positivism. It
remains to mention a third kind of positivism, which is called 'sociological positivism'. Roger
Shiner describes sociological positivism by saying that it [regards] the legal system and its
operations as an inanimate machine'. 56 Shiner gives the example of the work of Donald Black,
a contemporary sociologist of law who takes an empiricist or scientific approach to law. Black
writes: 'Law consists in observable acts, not in rules ... law is not what lawyers regard as
binding or obligatory precepts, but rather, for example, the observable dispositions of judges,
policemen, prosecutors, or administrative officials.' 57The behaviourist character of Austin's
account of law puts him in the sociological positivist camp. As we have seen, Austin thought
that social phenomena such as law should be approached in the way that we approach natural
phenomena, using the methods of science. It was for this reason that Austin attempted to
reduce normative legal statements to non-normative, factual statements about observable
regularities in human behaviour.

Austin's command theory is positivist in all of the three ways considered above: it is
substantively, sociologically and methodologically positivist. However, as indicated above, it
represents just one contribution to the positivist school of thought. Thus, as we will see, Hart is
not a sociological positivist. Although he is a methodological positivist, who does not seek to
justify law but rather to describe and explain it in a way that carries no moral commitment to its
institutions, he does not believe law lends itself to being understood in scientific or naturalistic
terms, that is, in terms of regularities in observable behaviour. Nor does Hart adhere to the
command theory of law. In both these ways, his version of positivism is more sophisticated
than Austin's. We turn to Hart now.

Questions
1. What are the defining features of Austin's positivism?

2. Can the rules of contract law be explained on a coercive model of law?

3. Can Austin explain the duty of a judge to apply the law?

4. Roger Cotterrell writes: 'Writers who have argued forcefully against [Austin's account
()flaw] have usually wanted to see the legal system as being governed by rules, even in
its highest regions of authority, rather than—as Austin's theory so starkly claims—
governed by people, mere human decision-makers with all their frailties and potential for
arbitrary or tyrannous exercise of human power ... Austin's is not a theory of the Rule of
Law: of government subject to law. It is a theory of the 'rule of men': of government
using law as an instrument of power. Such a view may be considered realistic or merely
cynical. But it is, in its broad outlines, essentially coherent' (R Cotterrell, The Politics
 ofjurisprudence: A Critical Introduction to Legal Philosophy, 2nd edn, LexisNexis UK,
London, 2003, p 70). Discuss.

5. Are the rules of customary societies—rules that do not derive their

obligatory force from the existence of a state—law?