Session 3 Hart, selections from The Concept of Law

Hart, like Austin, wants to retain the central tenet of legal positivism: the separation of the
concept of “law as it is” from the concept of “law as it ought to be”:

(1) The existence and content of laws is a matter of social fact.

(2) There is nothing in the nature of law as a social institution that guarantees its moral
worth.

But Hart also sees it as an essential feature of legal systems that they are normative: they
give rise to obligations of a sort, and generate (or are perceived to generate?) reasons.

This Hart saw as the essential seed of truth in Austin’s theory: laws generate requirements
that are in some sense non-­­optional.

But Hart is trying to occupy a somewhat slippery position: he wants to account for the
(necessary) normativity of law while denying that laws (necessarily) have moral content.
Can one account for the normativity of law without admitting that morality is “internal to
law,” in the sense the content of laws depends on the content of our moral requirements?

To see how Hart intends to answer this question, it’s fruitful to begin by looking at what he
sees as the short-­­comings of Austin’s Command Theory (CT) of Law:

(1) The CT can’t explain why/how laws bind their makers, too; and why not all
commands issued by a legislator are laws.
(2) The CT is ill-­­suited to explain slaws that are power--conferring, rather than duty--
creating: laws governing the making of wills, contracts, marriages, and more laws!
(3) The CT has difficulty recognizing as law any laws that aren’t brought into being by
explicit prescription (e.g., legal precedent).
(4) The CT has difficulty explaining the continuity of legal obligations across changes in
sovereignty.
(5) The CT mistakenly conflates being obliged with being obligated: commands backed
by sanctions may oblige us to act, but being obliged to at is neither a necessary nor a
sufficient condition for being obligated to act. And I’m not released from a legal
obligation just because I know I won’t be caught/punished. (The notion of an
obligation is normative. So the CT theory fails to adequately capture the sense in
which laws are normative after all.)

We can imagine responses to these worries on behalf of Austin; they too are informative:

(1) Austin could claim that law-­­makers, in their capacity as sovereigns, issue commands
that bind them in their capacity as citizens.
-­­ Hart would be sympathetic to this approach. But he asks: what allows us to
differentiate these roles? In virtue of what does something a legislator does

1
 count as making law? How are we to recognize when the legislator is acting as
sovereign and when she is acting as citizen?
(2) Austin might try to stretch his command model to accommodate power-­­conferring
laws.
-­­ One possibility is seeing laws governing, e.g., the making of a marriage contract
as a command “Get married this way!”, enforced by the sanction, “If you don’t,
you won’t count as married!”
o But if I choose not to get married, it seems a stretch to interpret me as
having violated a legal obligation and so having to suffer the appropriate
punishment: if I choose not to get married, not being married is no
punishment!
-­­ Another possibility is interpreting the laws as commands, backed by sanctions,
to public officials to treat people who have taken certain describable steps (filled
out certain forms, etc.) in certain ways.
o But Hart says this account has “no better claim to our assent than the
theory that all the rules of a game are ‘really’ directions to the umpire and
the scorer [score-­­keeper];” this account distorts the psychology of
ordinary citizens who take the relevant rules to be applicable to them.
(3) Austin might suggest that laws that result from legal precedent, and other laws not
explicitly issued, are tacit orders of the sovereign, because she doesn’t explicitly
repeal them.
-­­ But this seems overly simplistic. And such laws are law even if it’s clear that the
“sovereign” doesn’t approve of them (think of the Supreme Court’s recent
campaign finance decision). What gives these laws their validity?
(4) Again, Austin might suggest that any law that is a hangover from a previous
sovereign is binding in a new regime because, by not repealing it, the new sovereign
has tacitly reissued it.
-­­ But this might misdescribe our intuitive sense of the source of validity of the law,
and also can’t explain why we feel immediately bound by the laws issued by the
new authority, despite lacking the habit of obedience.
o We may have a habit of obedience to the role, which now has a new
occupant. But here Hart can return to his thoughts about question (1):
what establishes the role as something distinct from its occupant?
(5) Finally, we wanted to explain the bindingness of laws even when we know we won’t
get caught/punished for breaking them. Austin has suggested that a law is a
command backed by the possibility/probability of a sanction: that statements about
legal obligations are really predictions about what will follow if we perform certain
actions (how judges will rule, how authorities will act, etc.). This suggestion forms
the backbone of a theory of jurisprudence (misleadingly) called Legal Realism, and it
has some prominent adherents:

“Take the fundamental question, What constitutes the law? You will find some text
writers telling you that it is something different from what is decided by the courts
of Massachusetts or England, that it is a system of reason, that it is a deduction from
principles of ethics or admitted axioms or what not, which may or may not coincide
with the decisions. But if we take the view of our friend, the bad man, we shall find

2
 that he does not care two straws for the axioms or deductions, but he does want to
know what Massachusetts or English courts are likely to do in fact. I am much of his
mind. The prophecies of what the courts will do in fact and nothing more
pretentious are what I mean by law". (Oliver Wendell Holmes, Collected Papers, New
York, 1920, p.173).

-­­ But this approach again distorts the attitude of citizens bound by the laws, and of
judges and officials interpreting and enforcing them: how can judges’ rulings be
seen as predictions of how they themselves will act? And citizens see laws not
just as reasons to expect that punishment will follow certain behavior, but as
reasons for punishment to follows certain behavior.

All of these worries shape Hart’s own version of legal positivism. Our concept of law
should:
-­­ allow us to represent the difference between legislators as legislators and
legislators as citizens;
-­­ allow us to represent the difference between power--conferring and duty--
establishing rules;
-­­ allow us to explain the validity of laws not issued by a sovereign
-­­ reflect the distinction between being obliged and being obligated

For Hart, the central concept of law is not that of a command, but that of a rule:

The essential characteristic of a rule is that people who see themselves as bound by it see it
as providing them with a reason to act (to obey, or to apply sanctions…). (Question: can
this properly distinguish rules from commands?)

This is what Hart calls the internal aspect of rules:

It represents the difference between rule--following and merely habitual behavior. (The
latter can be fully described from the extreme external perspective.)

Points of View:

-­­ The extreme external PoV: can learn to see the red light as a sign that people will
stop (that’s how legal realists see jurisprudence).
-­­ The moderate external PoV: can learn to see the red light as a signal to people to
stop.
-­­ The internal PoV: sees the red light as a reason to for people to stop.

The essential flaw of both the CT and the predictive model (Legal Realism) is that they do
not recognize the internal point of view.

Mere habits of obedience: observable regularities.

Social rules: normative vocabulary used to draw attention to a standard of behavior and
criticize/punish deviations from it.

3
Examples: Rising inflection when asking a question (norm) v. rising inflection in
valleyspeak (habit).

Two Kinds (Levels) of Rules:

Primary Rules: forbid or require acts; generate duties or obligations.

Secondary Rules: rules about primary rules -­­ specify the ways in which the primary rules
may be conclusively ascertained, introduced, eliminated, varied, and the fact of their
violation conclusively determined; generate powers.

We can imagine a society governed only by primary rules: a society in which a certain set
of restrictions and requirements have been organically adopted as binding on the society,
by its members, as a matter of social fact. But that society will face uncertainty about what
the rules are; it will be static – unable to respond quickly to changes in circumstances by
changing the rules; and it will be inefficient – no good method to settle disputes about the
proper interpretation of rules, or to enforce compliance with them.

Secondary rules, which, Hart thinks, bring a society from a pre--legal state to a state
governed by a legal system, solve these problems:

Three kinds of secondary rules:

-­­ the Rule of Recognition “specifies some feature or features possession of which by a
suggested rule is taken as conclusive affirmative indication that it is a rule of the group
to be supported by the social pressure it exerts.” (p. 76)
o The RR solves the problem of uncertainty
o May recognize different ways of making laws, and grant some more authority than
others in cases of conflict (e.g, written constitution, enactment by legislature,
judicial precedent)
o The RR identifies a particular source of law as “supreme”, in the sense that rules
identified by reference to it count as valid rules of law even if they conflict with rules
identified by reference to other criteria. But this is not to say the power of that
source of law is unlimited (as Austin imagined sovereign power to be).
o The RR is a social rule – it exists simply in virtue of being generally accepted.

-­­ Rules of Change empower individuals to introduce new primary rules or eliminate old
ones.
o Solve the problem of stasis
o Explain the difference between the legislator as sovereign and the legislator as
citizen.

-­­ Rules of Adjudication establish proper procedures for settling disputes about the
meaning of laws.
o Help solve the problem of inefficiency

4
 o Go hand in hand with a rule of recognition identifying adjudicators’ judgments as
“sources” of law (since their judgments are authoritative as to what the law is)

Hart says for a legal system to exist, we must (i) have primary rules that are valid according
to the RR and are generally obeyed (though not necessarily seen by ordinary citizens as
legitimate – ordinary citizens needn’t take the internal PoV); and (ii) have secondary rules
to which the officials designated do take the internal PoV.

Validity v. Efficacy:

To say that a rule is a valid rule of law is just to recognize it as passing all the tests for
pedigree provided by the rule of recognition. To say a rule is efficacious is to say that it is
generally obeyed. But a rule can be a valid law even if it is not generally obeyed, so long as
it is identified as valid by the RR. (Some RRs may build an efficacy requirement into their
criteria for validity, such as those that include a “rule of obsolescence”.)

However, if there is a general disregard for all or most of the rules of a system, they may not
turn out to be valid rules of law. That’s because such general disregard for the rules would
seem to indicate that no RR is generally accepted (at least not anymore).

Hart says the question of validity doesn’t properly arise with respect to the RR. Why not?

• It is the ultimate source of validity – it establishes the criteria of validity of other

rules; but there are no rules providing the criteria for the assessment of it’s own
legal validity.
• We can ask “is this the RR that is used by people in this society?” and “is it a good
thing that they recognize this RR?”. But these are not questions of validity; these
questions are asked from the external perspective.
• People who take the internal perspective to a legal system presuppose an
affirmative answer to these questions. To claim a particular rule is or is not a valid
rule of law is to take the internal perspective to the system.
• The existence of the RR can be established from the external perspective (it is an
empirical question), but the validity of a rule of law can only be established from the
internal perspective.
• He says we should not say of the RR that it is “assumed to be valid” but rather that it
is “presupposed to exist.”
• Comparison to the meter stick: “[The RR] can neither be valid nor invalid but is
simply accepted as appropriate for use in this way. To express this simple fact by
saying darkly that its validity is “assumed by cannot be demonstrated,” is like saying
that we assume, but can never demonstrate, that the standard metre bar in Paris
which is the ultimate test of the correctness of all measurements in metres is itself
correct.” (p. 83)
• Comparison to the “pre-­­legal” simple system of primary rules (Hart says the
existence of the RR is like the existence of the primary rules in such a system…)
• Think also about the rules of a game…

5
MIT OpenCourseWare
http://ocw.mit.edu

24.235J / 17.021J Philosophy of Law

Spring 2012

For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.