Law and Morality

Cambridge Law Journal, 71(1), March 2012, pp.
59–85
doi:10.1017/S0008197312000207
ARTICLES
MORALITY, LAW AND CONFLICTING REASONS

FOR ACTION
PETER CANE*
ABSTRACT. In The Concept of Law, H.L.A. Hart suggested that four

formal features of morality distinguish it from law: importance, immunity
from deliberate change, the nature of moral offences and the form of
moral pressure. On closer examination, none of these supposed features
clearly distinguishes morality from law, at least in the broad sense of
‘morality’ that Hart adopted. However, a fifth feature of morality
mentioned by Hart – namely the role that morality plays in practical
reasoning as a source of ultimate standards for assessing human
conduct – does illuminate the relationship between law as conceptualised
by Hart and morality variously understood. Because morality has this
feature, law is always subject to moral assessment, and moral reasons
trump legal reasons. It does not follow, however, that law is irrelevant to
moral reasoning.
KEYWORDS: Jurisprudence; morality and law; Hart; legal reasoning;

reasons for action
This article is an expanded version of the 26th Maccabaean Lecture in

Jurisprudence delivered at the British Academy on 15 November 2011,
which marked the fiftieth anniversary of the publication of what Brian
Simpson has called “the most successful work of analytical jurispru-
dence ever to appear in the common law world”1 – HLA Hart’s
The Concept of Law (“CL”).2 The anniversary provided a good
* Australian National University College of Law. I have benefited greatly from invaluable
conversations with Tony Connolly. Niki Lacey and Nigel Simmonds made insightful comments
on earlier versions. I am grateful to Jeff Goldsworthy and other participants in the Melbourne Law
School Legal Theory Workshop, and participants in the Social and Political Theory Seminar at the
ANU, for helpful feedback. Address for correspondence: Professor Peter Cane, ANU College of
Law, The Australian National University, Canberra ACT 0200, Australia. Email: peter.cane@
anu.edu.au
1
A.W.B. Simpson, Reflections on The Concept of Law (Oxford 2011), 1.
2
H.L.A. Hart, The Concept of Law (Oxford 1961). References are to the 2nd edition, edited by
P. Bulloch and J. Raz and published in 1994. The 2nd edition reprints the 1st edition with the
addition of a posthumous Postscript based on material written but by Hart but not published in his
lifetime. The book has been translated into 17 languages, and more than 50,000 copies of the
2nd edition have been sold since 1998.
59
60 The Cambridge Law Journal [2012]
excuse – if an excuse was needed – to revisit what Hart, at the very be-
ginning of the book, identifies as one of three “persistent ques-
tions … concerning human society” (CL 1–6) – namely, the relationship
between law and morality. Hart described that relationship as “one of
the principal recurrent issues in the long and complicated history of
speculation about the nature of law” (CL 6). In various guises it has been
a recurring theme of Maccabaean Lectures since 1959 when Patrick
Devlin delivered the second in the series, entitled “The Enforcement of
Morality”, unwittingly launching one of the century’s most famous –
and notorious – jurisprudential debates, commonly known as “the
Hart/Devlin debate”. (I will say a little more about that exchange later.)
Law and morality are both concerned with practical reasoning –
that is, with reasoning about what to do, what goals to aim for and
what sort of person to be. In this sense, both law and morality are
about right and wrong, good and bad, virtue and vice. These contrasts
are “normative”: they express value judgments. Sometimes the terms
“moral” and “morality” are used in contrast to “immoral” and
“immorality” to distinguish normatively between right and wrong,
good and bad, virtue and vice. In a similar way, what is “legal” may
be contrasted with what is “illegal”, “legality” with “illegality”. On the
other hand, the terms “morality” and “law” may also be used to dis-
tinguish between different aspects of social life and different domains
of practical reasoning. Thus morality may be contrasted with tradition
or etiquette or custom and, of course, with law. We may, that is, use the
words descriptively, contrasting the moral not with the immoral but
with the non-moral.
Concerning the relationship between law and morality in their
normative senses, Hart’s view was that what is required or allowed by
law may be prohibited by morality – in other words, that what is legal
may be immoral. Conversely, what law prohibits, morality may require
or permit – in other words, what is illegal may be morally acceptable or
even required. This view is sometimes referred to as the “separation
thesis”.3 A corollary (or, perhaps, a different formulation) of the
3
The separation thesis must be distinguished from the thesis that there is no necessary conceptual
connection between law and morality. Hart did not subscribe to the “no necessary connection”
thesis. Indeed, it is inconsistent with the separation thesis, which entails that state law is necessarily
subject to moral evaluation. Hart also believed that there is a “naturally” necessary connection
between law and morality, which is found in the minimum content of natural law (CL ch. 9). To
mark the difference between the separation thesis and the “no necessary connection” thesis, some
scholars now associate Hart with the “separability” thesis to the effect that law and morality are
“separable” rather than separate. Concerning Hart’s motivation in adopting the separation thesis
see CL 207–212; see also P. Soper, “Choosing a Legal Theory on Moral Grounds” (1986) 4
Philosophy and Social Policy 32; F. Schauer, “Fuller on the Ontological Status of Law” in
W.J. Witteveen and W. van der Burg (eds.), Rediscovering Fuller: Essays on Implicit Law and
Institutional Design (Amsterdam 1999); L. Murphy, “The Political Question of the Concept of
Law” in J. Coleman (ed.), Hart’s Postscript: Essays on the Postscript to the Concept of Law
(Oxford 2001).
C.L.J. Morality, Law and Conflicting Reasons for Action 61
separation thesis is that what the law is and what it ought morally to be
are two different questions. Of course, the separation thesis is consist-
ent with the view that the law may be, and very often is, in agreement
with morality; but, say separationists, it ain’t necessarily so. The sep-
aration thesis can be distinguished from various other views about the
normative relationship between law and morality. According to a thesis
sometimes referred to as “hard natural law”, an immoral (or an “un-
just”) law is not a valid law at all. By contrast, according a position
that may be called “soft natural law”,4 the moral status of law does not
determine its existence or validity but it does affect its conformity with
ideals of rightness, goodness and virtue that are intrinsic to the way we
understand law5 or that are presupposed by law’s function.6
Underpinning the separation thesis are two assumptions: first, that
a reasonably clear descriptive distinction can be drawn between the two
normative phenomena; and secondly, that law is properly subject to
assessment in moral terms.7 These assumptions raise several questions.
How is morality to be distinguished from law? Why is law subject to
moral assessment? If law is subject to moral assessment, how, if at all, is
law relevant to moral reasoning? In this article I address each of these
questions using, as a point of departure, Hart’s account in The Concept
of Law of what morality is and how it differs from law. This aspect of
Hart’s work has received significantly less attention than his account of
law and its normative relationship to morality.8 My argument will be
that of the various characteristics of morality identified by Hart, the
one that most robustly distinguishes it from law is the function Hart
attributed to it of providing ultimate standards for assessing human
4
The leading modern exponent of this approach is John Finnis: J. Finnis, Natural Law and Natural
Rights, 2nd edn. (Oxford 2011).
5
E.g. N. Simmonds, Law as a Moral Idea (Oxford 2007); L.L. Fuller, The Morality of Law (New
Haven 1964).
6
E.g. R. Dworkin, Law’s Empire (London 1986), 93: “the most abstract and fundamental point of
legal practice is to guide and constrain the power of government. Law insists that force not be used
or withheld … except as licensed or required by individual rights and responsibilities flowing from
past political decisions about when collective force is justified.”
7
According to Martin Stone, it is the first, rather than the second, of these assumptions that
distinguishes the separation thesis from soft natural law: “Legal Positivism as an Idea about
Morality” (2011) 61 University of Toronto L.J. 313. He argues that (to the extent that law and
morality share common concerns) soft natural lawyers think of law as a “part of” (ibid., p. 341) or,
perhaps, an extension or realisation or concretisation of morality; whereas separationists conceive
of it as an instrument or application of morality. See also R Dworkin, Justice for Hedgehogs
(Cambridge, Mass. 2011), ch.19; concerning Kant’s similar view see A. Ripstein, Force and
Freedom: Kant’s Legal and Political Philosophy (Cambridge, Mass. 2009), 223–5, 255–6. In Stone’s
account, for soft natural lawyers, morality is incomplete without law. By contrast, for
separationists, morality is self-sufficient or, as he puts it, “self-standing” (ibid., p. 319). Stone
particularly associates the separationist view of morality with the development of utilitarianism as
a comprehensive moral theory and decision-procedure. The idea that morality can be fully
understood without taking account of law pervades modern moral philosophy.
8
The topic falls between two stools: most philosophers who discuss the nature of morality are
uninterested in law while legal theorists are typically more concerned with the normative
relationship between law and morality than with the descriptive relationship.
conduct. Because of this feature, in cases of conflict, moral reasons

trump legal reasons. I will also argue that the reason why morality
trumps law depends on the view taken of the nature of morality.
Finally, I will argue that it does not follow from the view that moral
reasons trump legal reasons that law is irrelevant to moral reasoning.
I. TYPES OF MORALITY
Hart conceived of moral norms as a category of non-legal norms. His
account of legal norms focuses on what he called “legal systems of
modern states” (CL 3) and “municipal legal systems” (CL 17) and is
informed by his familiarity with the English legal system as it was in the
middle of the twentieth century and, to a lesser extent, the US legal
system. In Hart’s view, the characteristic features of municipal law and
the “key” to understanding its nature are, first, a “rule of recognition”,
which determines what the law is; and secondly, officials authorised to
make and change the law, and to resolve disputes about what the law
is and how it applies in individual cases.9 He might have added that
municipal legal systems are also characterised by officials authorised
to implement the law, monitor compliance with the law, and detect and
impose sanctions for non-compliance. Municipal law, in Hart’s con-
ception, is characterised by particular types of social institutions.10 To
a greater or lesser extent, Hart argued, morality and other non-legal
normative regimes lack such an institutional framework. Understood
in this way, morality is a more-or-less non-institutionalised analogue
of law.
Hart pictured a “pre-legal” society in which more-or-less non-
institutionalised normative regimes played an important part in en-
abling people to live together successfully and well. He associated
the development of legal institutions with growth of social groups be-
yond a certain size. In larger, less cohesive societies, more-or-less non-
institutionalised normative regimes will continue to thrive amongst
small, cohesive groups within society but at the level of the society as a
whole, such practices will not be enough to enable the members of the
society to live together successfully and well: hence the development of
the sorts of institutions characteristic of municipal law.
9
Hart actually refers to “secondary” rules that empower the officials rather than to officials. But the
significance of secondary rules is that they create and empower the institutions that administer the
regime of primary rules.
10
Of course, there are many types of social institutions in addition to those identified by Hart as
characteristic of law. In this paper, the terms “institutional” and “institutionalised” refer only to
characteristically legal institutions. Ironically, Hart showed very little interest in actual legal
institutions and how they operate in practice: N. Lacey, “Analytical Jurisprudence versus
Descriptive Sociology Revisited” (2006) 84 Texas Law Review 945; “Out of the ‘Witches’
Cauldron’? Reinterpreting the Context and Reassessing the Significance of the Hart-Fuller
Debate” in P. Cane (ed.), The Hart-Fuller Debate in the Twenty-First Century (Oxford 2010).
In a pre-legal society, Hart says, there may be no clear distinction

between morality and “custom” (CL 86–7). Hart distinguished cus-
tomary normative practices from regular patterns of behaviour.11
A regularity of conduct becomes a custom, according to Hart, when
people start treating the behaviour in question as obligatory by
demanding compliance and criticising non-compliance. The facts that
particular conduct is demanded, that such demands are considered
legitimate, and also that non-compliance with such demands is con-
sidered a legitimate basis for criticism, together provide a reason to
comply independent of the content of the demand. Customary norms
are norms that arise from such social practices. By saying that in a pre-
legal society there may be no clear distinction between morality
and custom, Hart implies that morality and custom are both “con-
ventional” practices in the sense that they are based on shared patterns
of behaviour consisting of demands for compliance, widespread com-
pliance, and criticism of non-compliance, possibly accompanied by
other sanctions.
This sort of conventional morality Hart called “social” or “posi-
tive”. Like law, he said, it “may include much that is peculiar to the real
or fancied needs of a particular society” (CL 172) and may vary
significantly from one society to another. In his discussion of the
similarities and differences between morality and law in The Concept of
Law Hart adopted what he called a “broad” view of morality (CL 181)
that included such conventional social morality but also what he dub-
bed “personal” morality and “critical” morality. Personal morality he
understood as consisting of “ideals”12 of rightness, goodness and virtue
that an individual might neither share with anyone else nor use as a
basis for criticism of others. By contrast, he thought of critical morality
as standards that are “used in the moral criticism of society itself”
(CL 183) and which meet “conditions of rationality and generality”
(ibid). Hart resisted adopting a narrow view of morality that would
identify it solely with critical morality because he thought that all three
11
For a suggestive discussion of the relationship between habit, custom and law see J.B. Murphy,
“Habit and Convention at the Foundation of Custom” in A. Perreau-Saussine and J.B. Murphy,
The Nature of Customary Law: Legal, Historical and Philosophical Perspectives (Cambridge 2007).
For the criticism that Hart’s account of custom is purely structural see A.W.B. Simpson,
Reflections on The Concept of Law, p. 175.
12
Hart contrasted ideals with duties (CL 182), aspiration and supererogation with obligation. It is not
clear why he thought it important to include personal ideals in his account of morality since they
have no clear analogy in his account of law, which is social, not personal, and rooted in obligation,
not aspiration. Nor is it clear why he limited idiosyncratic, personal morality to ideals. An
individual might, for instance, consider vegetarianism an obligation for themselves but not for
others. Perhaps more importantly for present purposes, it is not clear whether or to what extent
personal morality shares with positive and critical morality the four characteristics listed later in
this paragraph – more especially, the last two. Social ideals have an important place in some
concepts of law and certain accounts of the normative relationship between law and morality. See,
e.g., the work of Simmonds and Fuller cited note 5 above. See also C.A.J. Coady, Messy Morality:
The Challenge of Politics (Oxford 2008), 67–70.
types of morality shared four characteristics that distinguished them

from law, which he referred to as “importance”, “immunity from de-
liberate change”, “the voluntary character of moral offences” and “the
form of moral pressure”. I will return to these suggested features of
morality later.
Although The Concept of Law attracted a large body of commen-
tary and criticism during his lifetime, Hart did not systematically ad-
dress or answer the commentators and critics. However, after his death,
Penelope Bulloch and Joseph Raz edited notes that Hart had made
over the years, which were published in 1994 as a Postscript to the
original text. In the Postscript, Hart radically revised the account of
morality that I have outlined above. In response to criticisms of his
ideas by Ronald Dworkin,13 Hart accepted Dworkin’s claim that there
is an important distinction between conventional social practices
and concurrent social practices. Custom, he said, is a conventional
social practice in the sense that the very fact that particular conduct
is generally thought obligatory provides a non-conclusive, content-
independent reason to engage in that conduct. By contrast, social
morality – he now thought – is a concurrent social practice. It consists,
he says, of an overlapping consensus of “conviction” amongst indi-
vidual members of the society “who have and generally act on the same
but independent reasons for behaving in certain specific ways”
(CL 255–6).14 In this new view, whereas custom is at least partly a
matter of what people generally think is the right thing to do, morality
is a matter of what the individual judges to be the right thing to do.
In this sense, although there can be customary law, there can be no
customary morality.
This change of mind by Hart is radical because it effectively col-
lapses the distinction he had previously drawn between positive and
critical morality. Understood as a concurrent practice, the morality
of societies and social groups is simply an aggregate of the moral
convictions of the individual members of the society or group. Whereas
Hart’s original thought seems to have been that conventional
social practice is the paradigm of morality (in one place he calls it
“primordial” (CL 183)),15 his revised view was that morality, properly
13
R. Dworkin, Taking Rights Seriously (London 1977), 53–58. See also G.J. Warnock, The Object of
Morality (London 1971), ch. 5.
14
See also R.B. Brandt, A Theory of the Right and the Good (Oxford 1979), 172. It does not follow, of
course, that social practices and institutions do not play a part – indeed, a very important part – in
the formation of individuals’ convictions and even their “personal ideals” (see e.g. P.F. Strawson,
“Social Morality and the Individual Ideal” in G. Wallace and A.D.M. Walker, The Definition of
Morality (London 1970)).
15
See also N. Cooper, “Two Concepts of Morality” in The Definition of Morality, esp. pp. 84–88.
According to James Wallace, the philosopher John Dewey considered customary morality to be
prior to critical morality: J.D. Wallace, Moral Relevance and Moral Conflict (Ithaca 1988), 73–6.
understood, is a matter of individual judgment and conviction and that

social morality is derivative of morality in that central sense.16
At around the same time as Hart was formulating his original
account of social morality, he made strategic use of the distinction
between positive and critical morality in his debate with Patrick Devlin
about the decriminalisation of private, consensual, male-homosexual
activity.17 In 1957 the Wolfenden Committee controversially re-
commended decriminalisation on the basis that private sexual behav-
iour is “none of the law’s business” unless it has a direct, adverse
impact on identifiable individuals other than the consenting partici-
pants. Devlin originally agreed with this recommendation, having
argued for reform in his submission to the Committee. However, in
the course of preparing his 1959 British Academy Maccabaean
Lecture, Devlin came around to the view that there was no reason, in
principle, to say that the law should never “track” moral judgments18
that certain behaviour is wrong regardless of its tendency to harm third
parties.19 This put Hart in a difficult position partly because he pre-
ferred to avoid controversy concerning the contents of morality
(CL 168)20 – what morality is about and whether or not particular
norms are part of morality.21 Some people think that private consensual
sexual behaviour is not a moral issue at all22 – in other words,
that private sexual behaviour is not only none of the law’s business
but none of morality’s either. But Hart was prepared to treat sexual
preference as a moral issue23 and was even apparently willing, for
the sake of the argument, to treat private male homosexual activity
as immoral. His position was essentially that of the Wolfenden
16
In Hart’s revised view, the “social rule” or “practice” account of norms fails to provide “a sound
explanation of morality, either individual or social” (CL 256). The extent to which “morality”
consists of “rules” is an ongoing topic of debate amongst philosophers that Hart does not address.
17
See P. Devlin, The Enforcement of Morals (Oxford 1965) (the Maccabaean Lecture is reproduced in
Chapter 1); H.L.A. Hart, Law, Liberty and Morality (Oxford 1963).
18
As Bob Goodin neatly puts it: R.E. Goodin, “An Epistemic Case for Legal Moralism” (2010) 30
Oxford Journal of Legal Studies 615.
19
According to Hart, Devlin later reverted to the view that “the law [on homosexuality] ought to be
altered”: D. Sugarman, “Hart Interviewed: HLA Hart in Conversation with David Sugarman”
(2005) 32 Journal of Law and Society 267, 285.
20
Following Hart’s lead, in most jurisprudential discussions of the concept of law, morality is
identified by its “form” as opposed to its “substance”. The substance of morality plays a larger role
in theoretical discussions of particular areas of the law, such as criminal law and contract law.
21
Hart said much about what he called “the minimum content of natural law” (CL ch 9). He argued
that as a matter of “natural necessity”, both morality and law address certain obstacles to
successful social life. The claim that morality is partly a function of human physical and mental
characteristics and the natural environment in which we live is, no doubt, relatively
uncontroversial; but how much of morality can be explained in this way certainly is not. For an
expansive view see J. Finnis, Natural Law and Natural Rights, 2nd edn (Oxford 2011). This is a key
question in contemporary evolutionary and cognitive science. See e.g. R.A. Hinde, Why Good is
Good: The Sources of Morality (London 2002); J. Mikhail, “Plucking the Mask of Mystery from Its
Face: Jurisprudence and HLA Hart” (2007) 95 Georgetown Law Journal 733.
22
E.g. T.M. Scanlon, What We Owe to Each Other (Cambridge, Mass. 1998), 171–2.
23
“[I]t would be absurd to deny the title of morality to emphatic social vetoes” such as sexual taboos:
CL 175. See also CL 182.
Committee – namely that the law should not concern itself with
“harmless” immorality – immorality “as such”, as both Hart and the
Committee called it.
One of the claims that Hart made in support of this position was
that when law-makers track morality, it is critical morality, not positive
morality that they should follow; and that certain (unidentified)
opponents of reform24 recommended the legal enforcement of positive
morality whereas those who supported reform had critical morality on
their side.25 This argument has two major weaknesses. First, even as-
suming that positive morality is conventional, it ignores the fact that
at the time, British public opinion, about homosexuality and the law’s
proper role in relation to it, was deeply divided. It puts in the mouth
of Hart’s opponents advice to the legislature that was, regardless of
its acceptability, impossible to follow. Secondly, however, if social
morality is understood to be a concurrent practice, the only possible
question for the law-maker who wishes to track morality in circum-
stances where opinion in society is significantly divided is which of two
(or more) critical, moral positions to track. Devlin offered highly con-
troversial content-independent advice to the legislature about how
to make this choice, whereas Hart’s strategy was to argue for the sub-
stantive superiority of his own critical judgment, namely that the law
should not proscribe harmless immorality. As a result, the protagonists
were, in a significant respect, arguing at cross-purposes. For Hart, the
fact that public opinion was divided was irrelevant to the task of the
legislature whereas for Devlin, it was of central importance.
Ronald Dworkin also joined in the debate against Devlin but, of
course, did not deploy the distinction between positive and critical
morality. In his view, what was “shocking and wrong” about Devlin’s
position was “not the idea that the community’s morality counts,
but his idea of what counts as community morality”.26 Dworkin dis-
tinguished between moral convictions in an “anthropological sense”,
referring to “whatever attitudes [a] group displays about the propriety
of human conduct, qualities or goals”,27 and moral positions in a
“discriminatory sense” which contrasts them with “prejudices, ratio-
nalizations, matters of personal aversion or taste, arbitrary standards,
mere emotional reaction and parroting”.28 In Dworkin’s opinion, law
should only track morality in the discriminatory sense. This advice
might help in cases of unreasoned moral disagreement but not where
disagreement is accepted on all sides to be reasonable in the sense that
24
But not Devlin: Law, Liberty and Morality (note 17 above), pp. 19–20.
25
Ibid., pp. 22–4.
26
Taking Rights Seriously (note 13 above), p. 255.
27
Ibid., p. 248.
28
Ibid., pp. 248, 250.
although each side thinks the other mistaken, the disagreement “has
survived shared deliberation conducted in good faith over an extended
period of time”.29
Hart’s change of mind about the nature of morality had other ra-
mifications, too. One of the reasons why he originally adopted a broad
understanding of morality was in order to avoid controversy about its
content and “status” – what morality is about and what sort of thing it
is. His aim was to identify features that were common to diverse ac-
counts of morality and that served to distinguish morality, broadly
understood, from law (CL 168–9). In this respect, Hart adopted a dif-
ferent strategy in relation to morality than in relation to law. As
we have seen, in constructing a concept of law he focused on municipal
law. So, for example, instead of developing a broader concept that
accommodated international law as well as municipal law, in the last
chapter of The Concept of Law he discussed at length whether the in-
ternational phenomenon is properly called “law” understood in terms
of the concept he had developed in the rest of the book. As a result of
this restrictive approach, Hart’s concept of law has become increas-
ingly controversial, attracting criticism from sociologically minded le-
gal pluralists. They point to a rich diversity of practised understandings
of what law is, to the fact that there may be more than one legal system
within a single social space, to the fact that municipal law is not the
only type of highly institutionalised normative system, and to the de-
velopment in recent decades of various more-or-less institutionalised
trans-national and supra-national normative regimes that further de-
stabilise the paradigm of municipal law. Hart has also been criticised
for failing to consider the possibility that different legal traditions or
families of legal systems might be based on significantly different con-
cepts of law.30
Avoidance of distracting controversy about the nature of morality
is precisely what Hart sought by adopting a broad and eclectic ap-
proach to that concept and precisely what he inevitably invites by
opting for a narrower concept of morality as individual conviction,
which is only one of a diverse array of understandings of the phenom-
enon. For the purposes of illuminating the relationship of morality to
law, what reason is there for ignoring, for instance, accounts of mor-
ality in terms of convention, or accounts that ground morality in
the authority of God or a sacred text or religious officials, or “realist”
accounts that liken moral truth to scientific truth? Putting the point
somewhat differently, perhaps we should entertain the thought that
29
C. McMahon, Reasonable Disagreement: A Theory of Political Morality (Cambridge 2009), 9.
30
E.g. Simpson, Reflections, note 1 above, pp. 157–60.
there are as many relationships between law and morality as there are
understandings of the two phenomena.
A possible advantage of narrower accounts is that, although
sociologically less satisfying, they may be analytically more powerful.
For instance, in his path-breaking book, A General Jurisprudence of
Law and Society,31 Brian Tamanaha sets out to develop a concept of
law designed to accommodate a large body of empirical data about law
and legal systems. He comes up with the idea that law is whatever
people call law. What this broad concept of law gains in sociological
fidelity it loses in analytical power. Hart’s sociologically narrower con-
cept is analytically powerful because it deploys a set of relatively clear
criteria that can be used to identify specific similarities and differences
between various normative regimes.32 It is very easy to spot funda-
mental differences between municipal law on the one hand, and social
morality understood narrowly in terms of individual conviction on the
other. It may be more difficult to identify clear differences between
municipal law and morality broadly understood, and even more diffi-
cult to identify clear differences between morality broadly understood
and law broadly understood.33 Indeed, legal pluralists have tended
to shift their attention away from attempting to categorise social
normative regimes, instead devoting their energies to understanding
how various normative regimes interact. Nevertheless, a significant
aspect of Hart’s project in The Concept of Law was to identify simila-
rities and differences between morality broadly understood and
municipal law.
II. THE STATUS OF MORAL JUDGMENTS

Just as Hart sought to avoid controversy by adopting a broader rather
than a narrower account of morality, he also looked for points of dis-
tinction between law and morality that side-stepped disagreements
about what he called the “status” of moral judgments. By “status” he
refers to the issue of whether morality is “part of the fabric of the
Universe not made by man, but awaiting discovery by the human in-
tellect” (CL 168) or whether, by contrast, it consists of “expressions of
changing human attitudes, choices, demands or feelings” (ibid.). The
former view is sometimes referred to as “realism” and the latter as
31
B.Z. Tamanaha, A General Jurisprudence of Law and Society (Oxford 2001).
32
For a suggestive comment by Hart along similar lines see Sugarman, “Hart Interviewed” (n 19
above), 291. For a related point see M. Krygier, “The Concept of Law and Social Theory” (1982) 2
Oxford Journal of Legal Studies 155, 159–61. By adopting a concept of morality as autonomous
judgment, Neil MacCormick is able to establish a “conceptual” distinction between law and
morality: N. MacCormick, Institutions of Law: An Essay in Legal Theory (Oxford 2007), ch. 14.
33
For an extended argument along these lines see J. Shklar, Legalism: Law, Morals and Political
Trials (Cambridge, Mass. 1986), 39–64.
“constructivism”.34 At first sight, Hart’s later revised view of morality

as personal conviction (unlike his earlier eclectic approach) might seem
to commit him to some version of constructivism; but in the Postscript
to The Concept of Law he reiterates the opinion that “legal theory
should avoid commitment to controversial philosophical views of the
general status of moral judgments” (CL 253–4).
Hart made this declaration of philosophical agnosticism not in the
context of his revised view of the nature of morality but in his dis-
cussion of what is called “soft” or “inclusive” positivism (CL 250–4).
This somewhat arcane and technical topic deserves our attention be-
cause of the light it sheds on Hart’s understanding of morality and its
normative relationship to law. Hart’s general approach to law is known
as “legal positivism”. Post-Hartian positivists divide themselves into
two camps: “soft” (or “inclusive”) positivists and “hard” (or “exclus-
ive”) positivists. Hart, by his own admission, was a soft positivist. Soft
positivists and hard positivists disagree about whether the validity of a
law can ever depend on its conformity with “moral principles or sub-
stantive values” (CL 250).35 Soft positivists say that it can, provided the
rule of recognition so provides. According to soft positivists, the effect
of a constitutional provision that empowered a court to strike down
parliamentary legislation on the ground that it infringed principles
of freedom of speech or equality before the law or due process (for
instance) would be to make the very existence of the legislation as valid
law depend on its conformity with certain moral principles.36 By con-
trast, hard positivists would interpret such provisions as empowering
judges to invalidate the legislation and in that way change the existing
law to bring it into conformity with morality.37
34
There are many versions and varieties of both moral realism and moral constructivism. My only
concern here is with Hart’s unwillingness to commit himself to any version of either. Christopher
McMahon develops a view he calls “moral nominalism”, “intermediate” between realism and
constructivism, to explain the important phenomenon of reasonable moral disagreement, to which
Hart paid little or no attention and which, McMahon says, neither realism nor constructivism can
adequately explain: McMahon, Reasonable Disagreement (note 29 above). Central to nominalism is
the idea that moral ideas evolves and can change.
35
Because this phrase appears in the Postscript to The Concept of Law, I assume that by “moral
principles” Hart means personal moral convictions as opposed to conventional morality. Hard
positivists need not and do not reject Hart’s view that conventional norms can be part of the law if
the rule of recognition so provides: CL 44–8.
36
I assume for the sake of the argument that freedom of speech, equality, due process and so on are
“moral” issues. However, it is arguable that framing the debate between hard and soft positivists in
terms of the relevance of “morality” to the validity of law raises the stakes unnecessarily. Instead,
the difference between the two positions might be said to concern whether the content of a law (as
opposed to its source) can affect its validity. This reading of the debate is consistent with the
argument made in the next paragraph below. In this light, a provision of the type we are
considering would authorise judges, in assessing its validity, to take account of the content as well
as the source of a law without raising any issue about the “status” of the relevant “substantive
values”, for conformity with which the law is to be tested.
37
Hart may have been attracted to soft positivism because it seems descriptively superior to hard
positivism. As a matter of legal doctrine, when legislation is held unconstitutional it is invalidated
with retrospective effect, not deprived of effect prospectively. On the other hand, hard positivism is
On the surface, at least, soft positivism may seem inconsistent with

the separation thesis I described earlier, a plausible interpretation of
which is that the validity of a law does not depend on its conformity
with substantive moral principles. However, since Hart unwaveringly
espoused the separation thesis, we should look for an interpretation of
soft positivism consistent with that thesis. The view that morality is a
matter of personal, conviction may provide such an interpretation. The
idea would be that legal provisions imposing substantive limitations on
what can count as law authorise legal officials to give effect to what
Hart calls their “best moral judgment” (CL 254) about what such lim-
itations prohibit, require or permit in particular cases. According to
this interpretation, such provisions do not make the existence and
validity of law depend on what moral principles prohibit, require or
permit but on what particular legal officials in particular cases believe
particular moral principles prohibit, require or permit. Because the
personal convictions of judges are open to moral assessment, this
approach preserves the distinction between asking what the law is and
what the law ought to be even in cases where what the law is depends
on substantive moral principles. This interpretation also has the
advantage, from Hart’s perspective, of being neutral as between realist
and constructivist views about the status of moral judgments.38 This is
because even if realism provides the correct or best account of morality,
there is no widely accepted method for discovering moral truth. It fol-
lows that the best that even a realist judge can do is to give effect to his
or her own personal moral convictions about what the moral truth is.39
Of course, regardless of the theory, in practice many judges think
that they should not give effect to their own personal moral convictions
in deciding what the law is or ought to be.40 They worry that the
legitimacy of the judicial role depends on finding some external source
likely to be normatively attractive to those who distrust judicial power or who think that citizens
should be able to identify what the law is without recourse to substantive moral reasoning.
38
Matthew Kramer is a soft positivist who commits himself to the view that there are objective moral
truths and that it is such truths that the sort of provision we have been discussing incorporates into
the law: M.H. Kramer, Where Law and Morality Meet (Oxford 2004), ch 2. However, he also
accepts, of course, that judges are morally fallible. His view is that a judge who makes a mistake
about what morality requires changes the pre-existing law while a judge who correctly identifies
what morality requires applies the pre-existing law. Kramer argues that the interpretation of soft
positivism I suggest here “can very easily be extended into an audacious form of rule-scepticism” in
relation to the application as opposed to the ascertainment of law (ibid., 119). This conclusion
seems to me to confuse moral reasoning with textual interpretation or, at least, to rest on a highly
formal, rule-based understanding of morality.
39
J. Waldron, “The Irrelevance of Moral Objectivity” in Law and Disagreement (Oxford 1999).
MacCormick apparently makes a similar point when he argues that for realists, autonomy is
morally rather than metaphysically relevant to their moral lives (MacCormick, note 32 above,
p. 250). Unlike Hart, MacCormick appears to have committed himself to some version of
constructivism.
40
See e.g. Lord Justice Balcombe, “Judicial Decisions and Social Attitudes” (1994) 84 Proceedings of
the British Academy 209, 229; Lord Bingham of Cornhill, “The Judges: Active or Passive” (2006)
139 Proceedings of the British Academy 55, 70. For an extended scholarly argument to this effect see
M.A. Eisenberg, The Nature of the Common Law (Cambridge, Mass. 1988).
of norms in cases where the extant legal materials provide no determi-

nate answer to the issue they have to decide. However, at least in cases
where the relevant issue is the subject of genuine, reasonable and in-
tractable moral disagreement, the most appropriate course of action
for the judge is to follow his or her own best moral judgment. Typically,
in principle at least, the moral views of an individual judge will not
determine what the law is, and the legitimacy of the input of the indi-
vidual judge to the law-making process must be assessed in terms of the
broad institutional environment in which cases are decided including,
for instance, majority voting in multi-member appeal courts and the
possibility of reversal by a legislature.41
III. FOUR FEATURES OF MORALITY

So far we have seen that Hart’s descriptive account of morality
was eclectic. It included various types of morality (conventional, con-
current, personal and so on) and was neutral as between realist and
constructivist views of the status of moral judgments. It is against this
descriptive background that Hart identified four main features of
morality that, in his opinion, differentiated it from municipal law: “im-
portance”, “immunity from deliberate change”, “the voluntary nature
of moral offences” and “the form of moral pressure”. On closer scru-
tiny, none of these features provides a straightforward contrast be-
tween morality broadly understood and municipal law. I will look at
each in turn.
A. Importance
“In contrast to morals”, Hart says, “some, though not all, rules of law,
occupy a relatively low place in the scale of serious importance”
(CL 174). On the other hand, moral rules, he asserts, are necessarily
important (CL 175). We can easily agree with Hart that law and mor-
ality both deal with issues of great human and social importance.
However, as Hart himself admits, the statement that morality deals
with serious matters is “both truistic and vague” (CL 173). We would
need to know much more about the content of morality (a topic that
Hart largely avoids – except in his discussion of what he calls “the
minimum content of natural law”) before we could assess in what ways
and to what extent law and morality differ along the dimension of
importance.42
41
P. Cane, “Taking Disagreement Seriously: Courts, Legislatures and the Reform of Tort Law”
(2005) 25 Oxford Journal of Legal Studies 393.
42
C.D. Johnson, “Moral and Legal Obligation” (1975) 72 Journal of Philosophy 315, 329–31 argues,
with particular reference to promising, that in both law and morality importance is related to the
level of abstraction and, conversely, specificity, at which their respective “requirements” are
stated – the more abstract the more important, the more specific the less important.
We would also need to know much more about the idea of

importance. Here, Hart is more forthcoming, and he fills out the
concept in three different ways (CL 173–4). First, he says, “[m]oral
standards are maintained against the drive of strong passions which
they restrict and at the cost of sacrificing considerable personal inter-
est”. Secondly, “serious forms of social pressure [are] exerted not only
to obtain conformity … but to secure that moral standards are taught
or communicated as a matter of course to all in society”. Thirdly, “if
moral standards were not generally accepted, far-reaching and dis-
tasteful changes in the life of individuals would occur”.
1. Curbing the Passions and Sacrificing Personal Interest

As a psychological generalisation, the claim that moral standards re-
strict strong passions seems false even in relation to fundamental moral
standards, such as the prohibition on murder. Under normal circum-
stances, very few people have to struggle mightily against the desire to
kill or injure others, for instance, or even to steal or defraud. In fact, it
is a reasonable hypothesis that the strength of the desire to breach
a moral standard is typically inversely proportional to its perceived
importance, whether in the abstract or in particular circumstances. In
this regard, it may be significant that the Concept of Law was written
around the time Hart was also debating with Patrick Devlin about
the decriminalisation of male homosexual activity. “[L]aws enforcing
sexual morality”, Hart said in Law, Liberty and Morality,43 “create
misery of a quite special degree. For the difficulties involved in the
repression of sexual impulses … are quite different from those involved
in abstention from ‘ordinary’ crime”. Indeed, Hart goes on to say that
“the impulse to steal or to wound or even to kill is not, except in a
minority of mentally abnormal cases, a recurrent or insistent part of
daily life”.44
Hart’s related claim that morality conflicts with personal interest
reflects an association, common in jurisprudential literature, between
morality and regard for others, and a corresponding contrast between
morality and “prudence” or “self-regard”. By contrast, there is a long
philosophical tradition that considers self-interest, one’s personal pro-
jects, aspirations and life-plans, as being integral to morality.45 One
of the standard objections to utilitarianism as a moral theory is that
it is too demanding and allows too little place for self-interest.46 The
43
Note 17 above, p. 22.
44
Ibid.
45
“Throughout the history of philosophy, by far the most popular candidate for the position of the
moral point of view has been self interest”: K. Baier, The Moral Point of View: A Rational Basis of
Ethics (Ithaca, NY 1958), p. 187. See also, e.g., S. Scheffler, Human Morality (New York 1992),
ch. 2; B. Leiter, “Nietzsche and the Morality Critics” (1997) 107 Ethics 250, esp. pp. 259–60, 272.
46
See, e.g., M. Timmons, Moral Theory: An Introduction (Lanham, MD 2002), 134–5.
degree to which morality is other-regarding is a matter of complex and

continuing debate and disagreement amongst moral philosophers.
Of course, we should not mistake philosophical moral theory for
morality; but disagreement amongst philosophers about the moral
status of self-interest should at least cause us to pause before accepting
Hart’s position.
At all events, even if we agree with Hart about the relationships
between morality, the passions and self-interest, it is surely the case that
law, too, may conflict with what we want and what we perceive to be in
our own interest. Even assuming that it makes sense to assess the
relative importance law and morality in the aggregate as opposed to
comparing particular legal rules and moral standards, there is no
obvious reason to think that the contrast between morality and law in
this regard is any more than a matter of degree; and it would be no easy
task to determine the relative positions of law and morality along
the dimensions of curbing the passions and frustrating the pursuit of
self-interest.
2. Serious Social Pressure

Hart’s second criterion of importance – that morality is characterised
by “serious” forms of social pressure – is extremely vague.47 Moral
sanctions vary in severity, and the least severe (a mild expression
of disapproval, perhaps) might not deserve to be called “serious”.
Unless one subscribes to the implausible view that all moral rules,
principles, standards, reasons and judgments are of equal importance
and force, not all immoralities would or should be met by equally
serious forms of social pressure. Moreover, by comparison with some
legal sanctions – for instance, the death penalty, long-term imprison-
ment, and large monetary fines and compensation awards – even very
serious moral sanctions (such as scape-goating and social ostracism)
may be experienced as relatively mild. Amongst the resources that
municipal law can put at the disposal of morality are stronger forms of
coercion than a non-institutionalised normative regime can effectively
provide without risking social disorder. An important consideration
to be taken into account when deciding whether or not to “legalise” a
particular non-legal norm is whether any of the various forms of
coercion that the law has at its disposal would appropriately be made
available to reinforce the norm in question. This may help to explain
why, for instance, many legal systems impose no “general duty to
rescue” even though many people believe that there is such a moral
47
Hart treats this criterion inconsistently, in one place offering it as a point of similarity between law
and morality, not difference: CL 172, where he summarises five “striking similarities between moral
and legal rules of obligation”.
duty. It is one thing to disapprove of and criticise failure to rescue but

another to impose legal liability and to award monetary compensation
for failing to rescue.
3. The Consequences of Immorality

Hart’s third criterion of importance – relating to the serious, negative
consequences for individuals of widespread rejection of moral
standards – is ambiguous. If it refers to any and every moral standard,
the claim seems as implausible as it would if it were made in relation to
any and every legal rule. Widespread failure to comply with a moral or
legal requirement of trivial social importance may have only trivial
social consequences. On the other hand, if the criterion refers to mor-
ality in the aggregate, it seems equally true of law. A general failure of
voluntary conformity to law would destabilize society and potentially
have a negative impact on its members just as much as would general
rejection of prevailing moral standards. Perhaps the clearest difference
between municipal law and morality so far as the consequences of
widespread non-conformity are concerned is that an individual legal
rule may be entirely a dead letter, neither complied with nor enforced
by anyone, but still exist as a law of the system; whereas the death of
compliance and enforcement would spell the end of existence for moral
standards. But even this conclusion might be questionable in relation to
understandings of morality that find a significant place for authority.
B. Immunity from deliberate change

1. Moral Legislation
Hart formulates this supposed feature of morality in various signifi-
cantly different ways: that morality cannot be deliberately made or
changed,48 that it is “something ‘there’ to be recognised”, and that the
idea of a moral legislator with authority to make and change morality
is “repugnant to the whole idea of morality” (CL 175–8). In consider-
ing these claims, it is important to remember that at the time he made
them, Hart took a broad view of morality and professed neutrality
as between moral realism and moral constructivism. So the question
is not whether there is any particular understanding of morality of
which these claims are true, but whether they are true of a broad range
of understandings of morality, whether realist or constructivist,
48
For the suggestion that the legal principle that promises should be kept is just as immune to
deliberate change as its moral analogue see Johnson, “Moral and Legal Obligation” (note 42
above), pp. 322–4. In Johnson’s view, the important distinction is not between law (changeable)
and morality (immune to change) but between fundamental principles (immune to change) and
specific rules and obligations (changeable).
whether conventional, critical or personal, and whether based on

authority or individual conviction.
In that light, consider first the claim that the concept of a legislator is
repugnant to the very idea of morality. The force of this claim depends
significantly on how the concepts of a legislator and legislation are
filled out. If the paradigm legislator resembles the UK Parliament, for
instance, or the US Congress, and if paradigm legislation resembles a
statute, the distinction between morality, however broadly understood,
and law might seem quite clear. There are no formal moral parliaments
or congresses and no moral statutes. On the other hand, if one’s image
of a legislator and of legislation is looser and less formal, the ideas of a
moral legislator and moral legislation may not be repugnant to an
understanding of morality that, for instance, finds a significant place
for authority.49 From a completely different perspective, in the words of
Gilbert Harman, “Kant argued that we must think of the principles
of morality as principles that each of us legislates for himself and for
others”.50
Anyway, even the absence of moral parliaments and moral statutes
may not be quite as decisive as it first appears. One of the important
social functions of legislation is to reinforce and supplement morality
in various ways; and to the extent that such legislation is morally
acceptable, it can generate new moral reasons for action as well as legal
reasons. Nor is legislation the only mechanism by which law is made.
For instance, during the first 800 years of the English legal system,
courts were the main source of law and parliament played only a
subsidiary role. In England, courts pre-dated parliament, and it is not
clear that we would want to deny the title of “legal” even to a system
in which all law was made as a by-product of adjudicating disputes.
When they make law, judges commonly justify their decisions by
reasoning in ways that are structurally similar to formalised models
of moral reasoning and judgment such as John Rawls’s reflective
equilibrium method.51 The common law treats every principle as revis-
able at the point of application and provides a working model of how
change can occur in a normative regime, like morality, that puts high
value on stability and in which change is typically gradual and
incremental.
49
Indeed, G.E.M. Anscombe famously argued that the concept of morality only makes sense within a
religious framework: Ethics, Religion and Politics: The Collected Philosophical Papers of G.E.M.
Anscombe, Volume Three (Oxford 1981, ch 4 (“Modern Moral Philosophy”). For a discussion of
the role that authority might play in secular morality see S. Fleischacker, The Ethics of Culture
(Ithaca, NY 1994), esp. chs. 3 and 4.
50
G. Harman, The Nature of Morality: An Introduction to Ethics (New York 1977), 112.
51
Of course, judicial reasoning is substantively constrained by institutional factors in a way that the
moral reasoning of individuals is not.
2. Morality “there” to be recognised

Contrary to Hart’s professed agnosticism about the status of moral
judgments, on its face the claim that morality is something “there” to
be recognised apparently assumes the truth of some version of moral
realism. However, in context, it seems that this is not what Hart in-
tended because he says that morality shares this feature with traditions,
which, in his view, are conventional social practices. Traditions, Hart
says, acquire and lose their status as such “by growing, being practised,
ceasing to be practised, and decaying … by … slow, involuntary
processes” (CL 176). The idea that traditions and other conventional
practices are involuntary is bizarre and quite at odds with Hart’s ac-
count of how social rules come into existence and are sustained, which
involves not only voluntary but also deliberate conduct.52 In fact, it
seems that Hart meant only to restate the claim that traditions, like
morality, are not the product of legislation in a quite formal sense. As
he goes on to say, immunity from legislative change “must not be
mistaken for immunity from other forms of change” (CL 176).
Of course, Hart’s revised view was that social morality is not anal-
ogous to tradition because, unlike tradition, it is not conventional.
Although Hart did not revisit his discussion of the amenability of
morality to deliberate change in the light of his revised view, the ideas
of personal morality and individual ideals that he included in his orig-
inal catalogue of types of morality are analogous to his later concept of
individual conviction. About personal morality he says that “though
conversions are possible, the notion that such ideals could be adopted,
changed, or eliminated by deliberate choice is chimerical” (CL 184).
This statement might seem to take us well beyond the idea that mor-
ality cannot be formally legislated in the direction of some version of
realism. Once again, however, given Hart’s desire to avoid commitment
to any particular view about the status of moral judgments, we should
try to interpret his position consistently with that stance. A possibility
is to say that when people adopt moral positions for the first time or
change their existing moral views they treat those new positions and
views as always having been the right ones independently of their
having espoused them. This might be thought to provide evidence for
the objectivity of morality. However, of course, it does not follow from
the fact that we treat our moral views as timeless and pre-existent that
they actually are. We might explain such an attitude in functional terms
as providing a psychological basis for treating morality as “the ultimate
standard by which human actions … are evaluated”, as Hart puts it
(CL 230).
52
To similar effect: Warnock, The Object of Morality (note 13 above), pp. 49–50.
At all events, Hart gives us no good reason to accept that immunity

from deliberate change is a necessary feature of morality unless, per-
haps, we understand “deliberate change” in a very narrow way.
C. The Nature of Moral Offences

In law, Hart says, liability regardless of fault may be open to criticism,
but in morality it “comes as close to being a contradiction in terms as
anything in this sphere” (CL 173). Moral obligations, he adds, are
“within the capacity of any normal adult” (CL 171). “In morals” he
continues, “ ‘I could not help it’ is always an excuse, and moral obli-
gation would be altogether different from what it is if the moral
“ought” did not in this sense imply “can” ” (CL 179). If these state-
ments are understood as referring to morality as commonly practised,
the claim that morality, unlike law, does not recognise strict liability
seems quite obviously false. Hart’s view to the contrary may be the
result of failing to distinguish clearly enough between responsibility
and liability. It is one thing to say that a person is not “morally re-
sponsible” for the adverse consequences of their conduct unless that
conduct was blameworthy, but quite another to say that a person can
have no moral obligation to make some sort of amends for the adverse
consequences of their conduct unless that conduct was blameworthy.
In neither legal nor non-legal normative regimes is liability necessarily
responsibility-based. The conduct required to discharge a moral obli-
gation to make amends may be much less onerous than that required
by the law in similar circumstances – perhaps an apology would be all
that is needed or an expression of sympathy or regret. And, of course,
people may reasonably disagree about the content, and even the exist-
ence, of such an obligation outside the law. However, there is no reason
to think that moral strict liability is “a contradiction in terms” or even
close to it.
Hart may have been led to this view by his tendency to associate
morality with criminal law – a tendency shown by his description of
moral wrongs as “offences”.53 People who would be prepared to accept
that strict legal liability to compensate victims of harm is morally
acceptable or required might feel much less comfortable about
criminalizing, and punishing people for, faultless conduct. A judgment
that moral punishment for blameless behaviour would be unacceptable
does not lead to the conclusion that a moral obligation to make amends
for the consequences of such conduct would be equally unacceptable.
53
More generally, this tendency may explain why philosophers deny the possibility of strict moral
liability: J. Watkins, “Responsibility in Context” (2006) 26 Oxford Journal of Legal Studies 593,
606.
The claim that moral obligations are necessarily within the capacity
of normal adults also seems problematic, at least on the assumption
that there can be “conflicts between irreducibly moral considera-
tions”.54 Suppose that a Good Samaritan can provide vital assistance
at a crash scene only by breaking a long-standing promise to meet a
disabled relative off a long-haul flight. The Good Samaritan cannot
comply with both the obligation to assist and the obligation to keep the
promise. In this sense, “ought” does not imply can.55 Leaving aside
normative conflicts, it is at least plausible to think that in the way most
people understand morality, it does not impose obligations that are
beyond the capacity of normal people to perform or prohibitions that
are beyond the capacity of normal people to observe – albeit, for some
people and in some circumstances, only by dint of considerable effort.
However, the same, in general, could also be said about law. What is
more, just as particular individuals may lack the capacity for moral
responsibility, so people may lack the capacity for legal responsibility.
In morality, “normal capacity” is commonly treated – as in law – as a
threshold above which differences in capacity amongst “normal” peo-
ple are more or less irrelevant to responsibility, if not to liability. It is
widely accepted that in morality as much as in law, the relevant re-
sponsibility question is not whether a person who fails to behave as law
or morality requires on a particular occasion and in particular cir-
cumstances could have conformed to the relevant legal or moral norm
on that occasion and in those circumstances – even assuming that this
question makes sense. Rather, the question is whether, on such occa-
sions and in such circumstances, the person is generally able to behave
as law or morality requires.56 Requiring people to do the impossible is
not a characteristic feature either of legal or moral ideas of responsi-
bility.57
D. The Form of Moral Pressure

Securing compliance with morality, Hart claims, involves “appeals to
the respect for the rules, as things important in themselves, which
is presumed to be shared by those addressed” and never involves
54
Wallace, Moral Relevance and Moral Conflict (note 15 above), p. 12. Wallace argues that solving
moral conflicts requires active interpretation and adaptation of moral rules and principles in a way
that changes existing rules and principles and generates new ones. Morality, he asserts, is “an
intellectual and cultural achievement” (ibid, 53) and a matter of “learned dispositions” (ibid.,
p. 55).
55
To similar effect: C.E. Larmore, Patterns of Moral Complexity (Cambridge 1987), 149–50.
56
T. Honoré, “Can and Can’t” (1964) 73 Mind 463; reprinted in T. Honoré, Responsibility and Fault
(Oxford 1999). I am leaving aside the issue of causal determinism. Whatever the truth may be at the
micro-level of physical processes, at the meso-level of practical reasoning and action there is no
good reason to ignore the lived experience of freedom and control.
57
Fuller’s famous list of pathologies of rules includes “requiring conduct beyond the powers of the
affected party”: L.L. Fuller, The Morality of Law, revised edn. (New Haven 1969), 39.
“only threats of physical punishment or unpleasant consequences”

(CL 180–1; see also 86, 228). Hart apparently intends this as a con-
ceptual claim, but I think it is better understood as an empirical
hypothesis about what Hart calls the “typical forms” of legal and moral
pressure respectively: “physical” in the case of law and “non-physical”
in the case of morality. Empirically, Hart’s claim strikes me as being far
from obviously true either in relation to morality or law. We might
hypothesise with equal plausibility that those who seek the compliance
of others with either morality or law typically use whatever mix of hard
treatment and appeals to legitimacy they think most likely to achieve
the desired result. Moreover, if, with Hart (CL 38–9), one believes that
law’s prime function is to guide the law-abiding rather than to control
law-breakers, and that general conformity with moral and legal norms
is a precondition of social life and human flourishing, it seems better to
think of physical sanctions as pathological rather than as characteristic
of law.
We may conclude this discussion of Hart’s four points of dis-
tinction between morality and law by saying that it suggests that
the two phenomena – or, at least, particular legal rules and moral
standards – can usefully be contrasted along various dimensions, but
not that they occupy categorically distinct social and conceptual
spaces.
IV. MORALITY AND CONFLICTING REASONS FOR ACTION

Although Hart took a quite narrow approach to law focused on mu-
nicipal legal systems, he approached morality more broadly; and this
is one reason why he was not successful in identifying crisp distinctions
between law and morality. Success would probably have been
even more elusive if he had adopted a broader approach to law.
Nevertheless, there is one characteristic of morality identified by Hart,
almost in passing and without much elaboration, which does feature in
most accounts of morality of which I am aware,58 and which also ex-
plains why law is subject to moral assessment. In Hart’s words, it is
“the part played by morality in the lives of individuals” (CL 176)
“as the ultimate standard by which human actions … are evaluated”
58
At any rate, accounts of the morality of obligation as opposed to the morality of aspiration: see
n 12 above. One exception that I know of is Bernard Gert’s account of “common morality”, which
he defines as “the moral system that most thoughtful people use, usually implicitly, when they make
moral decisions and judgments”: B. Gert, Common Morality: Deciding What to Do (Oxford 2004),
v. In Gert’s scheme, the ultimate normative standard is “rationality” (ibid., pp. 91–5); and (he says)
while it is never irrational to act morally, it may be rational to act immorally. Gert’s view is
explained by the fact that “common morality” is a substantive, not a formal concept. According to
Gert, the basic principle of common morality is to “do no harm”. This understanding of morality
also leads him to draw a distinction between religious principles about right and wrong, good and
bad, virtue and vice on the one hand, and moral principles on the other (ibid., p. 4).
(CL 230).59 This feature differs from the others discussed by Hart in two
significant ways. First, it refers to the function of morality as opposed
to what we might call its “form” or “structure”. Secondly, it refers to
individuals” reasoning about what do to and what sort of person to be,
about their own conduct and that of others. By contrast, in Hart’s main
discussion of the distinctive features of morality, both morality and law
are treated as things in the world – they are, if you like, reified or per-
sonified. Lawyers and legal theorists often make statements about “the
law” – that it has aims and functions, for instance, and that it makes
“claims” (of being morally justified, for instance)60 – rather than talk
about the aims and purposes that people pursue through law and the
claims they make about it. Similarly, it is not only moral realists who
reify morality and think of it as a subject rather than a product of
human reasoning. The shift, from reifying morality and law to thinking
about them in terms of individuals’ practical reasoning, opens up
fruitful ways of understanding their inter-relationship.
The function of providing individuals with ultimate standards for
assessing human conduct is not only common to diverse under-
standings of morality; it also distinguishes morality from municipal law
and from other social normative regimes whether or not they are
understood to be legal.61 In practical deliberation about the right, the
good and the virtuous, when a moral reason for action conflicts with a
reason derived from another normative regime,62 the moral reason
trumps the non-moral reason. More particularly, when a moral reason
for action conflicts with a reason for action derived from the law, the
moral reason trumps the legal reason because one of the functions of
morality in practical deliberation is to provide ultimate standards for
the assessment of law.63
59
Cf Scheffler, Human Morality (note 45 above), p. 12: “a shared reference point for the formulation
and adjudication of challenges to existing institutions and practices more generally”.
60
Or even that it is male or female: G. Criscuoli, “Is the Common Law Masculine or Feminine?”
(1981) 1 Oxford Journal of Legal Studies 305.
61
For this reason, amongst others, Hart resisted classification of international “law” as
(international) morality: CL 227–32.
62
In the philosophical literature there is much discussion of a different question of whether and the
extent to which moral reasons for action override or, on the contrary, accommodate reasons of self-
interest (and, perhaps, some other types of “non-moral” reasons). See note 45 above. There is a
similar debate about the relationship between moral reasons and reasons of public interest:
P. Cane, Responsibility in Law and Morality (Oxford 2002), 276–7; see also Hart’s discussion of the
relationship between “justice” and “social/public/common good” (CL 166–7). The point I am
making in the text assumes that the question of what reasons are moral reasons has already been
answered. Under a theory that allows reasons of self-interest to trump moral reasons, legal reasons
would presumably trump reasons of self-interest. One view of moral reasons for action is that they
are “all-things-considered” reasons. So understood, morality would trump law as a matter of
definition: L. Alexander and F. Schauer, “Law’s Limited Domain Confronts Morality’s Universal
Empire” (2007) 48 William and Mary Law Review 1579.
63
This way of presenting the issue will not appeal to those who think that all genuine reasons for
action are moral. They might prefer a formulation in terms of a conflict of the moral reasons
provided by the law and moral reasons apart from the law. See note 6 above.
This is not to say, of course, that law does not claim to trump
morality. Understood as a reified or personified entity, law may be said
to claim supremacy over morality just as morality can be said to claim
supremacy over law. Acting in accordance with a moral reason
and contrary to a legal reason may attract legal sanctions just as the
imposition of sanctions in response to such conduct may be judged
morally unjustified. However, because morality provides ultimate
standards for assessing human conduct in a way that law does not,
moral reasons trump legal reasons in the deliberations not only of
subjects of the law but also of legal officials who administer the law.
Even while submitting to law’s claim of supremacy and its superior
force, individuals may assess its claim and its application of force to be
morally unjustified; and even while enforcing law’s claim to supremacy
and applying its force, legal officials may assess that claim and that
force to be morally unjustified – at the retail level, anyway, if not
wholesale.
Although this feature of providing ultimate standards is common to
very many, if not all, accounts of morality, the diversity of those ac-
counts is reflected in the variety of explanations they suggest of the role
of morality as the source of ultimate standards of conduct. For in-
stance, if morality is understood as a matter of authority, moral rea-
sons will trump non-moral reasons in normative conflicts because they
are thought to derive from an authority higher than any other norm-
making authority. According to realism, moral reasons trump legal
reasons because, unlike law and other social normative regimes, mor-
ality is not a human creation but in some sense part of the natural order
and so, timelessly, unchangeably and objectively “true”. Understood
conventionally, morality may provide civil society and, more particu-
larly, social groups with a normative bulwark against the exercise of
power through law,64 while understood as a matter of individual con-
viction, it can function to protect individual freedom of thought and
conscience against the demands of society and the state. The general
point is that there are many, diverse accounts of morality which share
the feature of trumping law and many, diverse explanations of why
morality has this feature.
What are the implications of the ultimacy of moral reasons? In
particular, what does it tell us about the relevance of law to moral
reasoning? The argument I want to make here is that the belief that the
law is always subject to moral evaluation and that moral reasons trump
legal reasons does not lead to the conclusion that the law is irrelevant to
64
Remember that this statement refers to the function of morality in individuals’ practical reasoning.
In a moral theory that distinguishes between conventional, positive morality on the one hand and
individual, critical morality on the other, critical morality provides ultimate standards for assessing
positive morality as much as other conventional normative systems.
moral reasoning and deliberation. In particular, I suggest that a person

who accepts either or both of the following two propositions about
morality may have good cause to take account of the law in their moral
reasoning. The first proposition is that the requirements of morality in
particular situations and circumstances may be indeterminate because
moral rules and principles tend to be more-or-less abstract. The second
proposition is that people can genuinely, reasonably and intractably
disagree about what is morally right, good and virtuous.
Consider, first, the indeterminacy of morality. By dint of its in-
stitutional, norm-generating, resources – legislatures, courts and so
on – law can supplement morality by giving it a degree of determinacy
that it would otherwise lack and that may be essential for successful
social life.65 For instance, law can supplement a moral injunction to
respect other people’s property by establishing a detailed regime of
property rights; it can supplement a moral injunction to contribute
financially to the support of social life by establishing a detailed tax
regime; it can supplement a moral injunction to do no harm by estab-
lishing rules of the road; and it can supplement a moral injunction to
promote social good by conferring specific powers and imposing spe-
cific obligations on public officials.
Next, consider disagreement. Because human beings are social ani-
mals, successful social life is a fundamental human value the mainten-
ance and promotion of which can provide good reasons for acting in
certain ways and refraining from acting in certain other ways. In si-
tuations where moral disagreement about the right thing to do creates a
risk of dysfunctional social instability and conflict, the institutional,
norm-generating resources that law possesses and morality lacks can be
used to manage such disagreement by giving people reasons to act or to
refrain from acting in particular ways, for the sake of maintaining and
promoting successful social life, even if they consider that in requiring
or forbidding such action, the law conflicts with some standard of right
conduct to which they subscribe independently of the law and of the
disagreement which the law addresses. In some situations, a reason to
act generated by the law for the sake of maintaining and promoting
successful social life may outweigh a reason for action recognised in-
dependently of the law and of the fact of disagreement. In this way, law
can promote social good by enabling people to live together success-
fully and well despite significant and serious differences of moral opi-
nion and outlook.
A person who accepts that the legitimate functions of legal institu-
tions include supplementing morality and managing potentially dis-
ruptive moral disagreement has good cause to take account of the law
65
T. Honoré, “The Dependence of Morality on Law” (1993) 13 Oxford Journal of Legal Studies 1.
in their moral reasoning because the law possesses institutional

resources that enable it to generate reasons for action that could not be
generated easily or at all without those resources.66 Of course, on the
assumption that it is the function of morality, not law, to provide
ultimate standards for judging human conduct, the reasons for action
that law supplies are always subject to assessment against the ultimate
standards that individuals recognise independently of the law. In
particular, even assuming acceptance of the propositions about
indeterminacy and disagreement, law can provide reasons for action
only to those who accept the moral legitimacy of legal institutions and
legal processes for making law, enforcing compliance, adjudicating
disputes and so on. However, a person who accepts that morality is not
self-sufficient but is dependent on law in the ways I have outlined, and
who accepts the legitimacy of that dependence, has good reason to take
account of law in their reasoning about ultimate standards of rightness,
goodness and virtue.
Acceptance of the propositions about indeterminacy and disagree-
ment would also give philosophers good cause to take account of law in
analysis of and normative theorising about morality. My point here is
not that the literature of the law may provide a richer source of raw
material for philosophical moral reasoning than thought experiments
can offer – although I believe this to be true; and it is an argument that
philosophers are, in my experience, quite willing to accept when it is put
to them. The point, rather, is that if the propositions about in-
determinacy and disagreement are accepted, it is a philosophical mis-
take to think that morality can be properly and fully understood
without reference to law. Some philosophers (and ordinary practical
reasoners) may be inclined to reject this conclusion on the ground that
there are many legal systems; that law is a local and culturally relative
phenomenon, tied to particular communities; and that legal systems
can and do render morality determinate and resolve moral disagree-
ments in various different ways. For some people, moral standards are,
necessarily rather than contingently, not only ultimate but also uni-
versally valid. However, while universality provides a reason, ad-
ditional to that provided by ultimacy, for treating the law of any
particular legal system only as a source of provisional standards, it
provides no good reason to ignore law, in the sense of a set of in-
stitutionalised social practices, as a potential source of non-provisional
66
Even legal norms that are morally “arbitrary” in the sense that the norm chosen (e.g. drive on the
left) is no more or less acceptable than some alternative (e.g. drive on the right) so that the choice
between them is of “no intrinsic importance” (CL 229–30) may provide moral reasons for action.
Hart believed that morality “logically” cannot be arbitrary in this sense (CL 229). However, if
morality is understood not only to require and prohibit conduct but also to permit conduct that it
neither requires nor prohibits, it is logically possible that in any particular situation several courses
of conduct may be of equal moral value so that none is morally preferable to any alternative.
standards given law’s capacity, by reason of its institutional resources,

to make a distinctive contribution to practical reasoning.67
Another way of making this point is to adopt, for the sake of
argument, the common view that a moral reason for action tells us
what we should do, all things considered. A person who accepts (1) the
indeterminacy of morality and the fact of moral disagreement; (2) that
law has institutional resources that enable it to make a distinctive and
valuable contribution to practical reasoning by rendering morality
more determinate and managing moral disagreement; and (3) that
these facts are relevant to practical reasoning, also has cause to accept
that a legal reason for action may, in principle at least, be an all-things-
considered reason.68
On the other hand, in order to have good cause to take account
of law in moral reasoning it is not necessary to believe that “the
establishment of legal order as distinct from discretionary or statically
customary order is … a moral ideal if not a compelling demand of
justice”;69 or that law provides “the only possible set of conditions
within which one can inhabit a human community and yet enjoy a
degree of freedom as independence from the power of others”.70 It
is necessary only to accept that law has institutional resources that
enable it to make a distinctive contribution to answering practical
questions about what one ought to do or what sort of person one
should aim to be.
V. CONCLUSION
In this paper, using Hart’s discussion in The Concept of Law as a
starting point, I have addressed three questions about the relationship
between law and morality: How is morality distinguishable from law?
Why – as the separation thesis assumes – should moral reasons for ac-
tion trump legal reasons? And, given that moral reasons trump legal
reasons, what is the relevance of law to moral reasoning? Concerning
the first question, we have seen that Hart originally took a broad and
eclectic approach to morality but later re-conceptualised it more nar-
rowly in terms of individual judgment and personal conviction. It was
in the context of the earlier approach that Hart identified four sup-
posedly distinctive features of morality. I have argued that none of
these features clearly distinguishes morality from law, and I have
67
I hope that this account of the relationship between law and moral reasoning, both “everyday” and
“philosophical”, is an improvement on the account I gave in Responsibility in Law and Morality,
which was fairly criticised by Jeremy Watkins, “Responsibility in Context” (2006) 26 Oxford
Journal of Legal Studies 593.
68
I am not sure whether Alexander and Schauer (note 62 above) go quite as far as this.
69
Finnis, op. cit. note 4 above, p. 14.
70
Simmonds, op. cit. note 5 above.
suggested that Hart’s lack of success in this regard is the analytical

price of his sociologically catholic approach to the concept of morality.
His later, narrower, understanding of morality is much more easily
distinguishable from law.
Hart also identified a fifth feature of morality, namely that of pro-
viding individuals with ultimate standards for the assessment of human
conduct. This feature differs significantly from the other four in that it
relates to the function of morality as opposed to its form or structure,
and it shifts the focus of attention from morality as a reified social
phenomenon to its place in practical reasoning. I have argued that
because this feature is shared by many, diverse understandings of the
form and structure of morality and the status of moral judgments it
provides a robust point of distinction between morality and law. It also
explains, in answer to the second question, why moral reasons trump
legal reasons, although the details of the explanation vary amongst
different understandings of the form and structure of morality and the
status of moral judgments. Addressing the third question, I have
argued that it does not follow from the fact that moral reasons trump
legal reasons that law is irrelevant to moral reasoning. Law deserves to
be taken seriously in reasoning about what is ultimately right, good and
virtuous by people who accept that morality is dependent on law in
certain ways.
An attractive aspect of this analysis is that it links the separation
thesis, which encapsulates Hart’s view about the normative relation-
ship between law and morality, with a descriptive account of their re-
lationship in terms of their respective places in practical reasoning. The
separation thesis rests on the assumption that moral reasons trump
legal reasons, and the function of morality as a source of ultimate
standards for assessing human conduct explains why this is so.
Copyright of Cambridge Law Journal is the property of Cambridge University Press and its content may not be
copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.

Law and Morality

Uploaded by

Copyright:

Available Formats

Law and Morality

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Law and Morality

Uploaded by

Copyright:

Available Formats

Cambridge Law Journal, 71(1), March 2012, pp.

MORALITY, LAW AND CONFLICTING REASONS

ABSTRACT. In The Concept of Law, H.L.A. Hart suggested that four

KEYWORDS: Jurisprudence; morality and law; Hart; legal reasoning;

This article is an expanded version of the 26th Maccabaean Lecture in

conduct. Because of this feature, in cases of conflict, moral reasons

In a pre-legal society, Hart says, there may be no clear distinction

types of morality shared four characteristics that distinguished them

understood, is a matter of individual judgment and conviction and that

II. THE STATUS OF MORAL JUDGMENTS

“constructivism”.34 At first sight, Hart’s later revised view of morality

On the surface, at least, soft positivism may seem inconsistent with

of norms in cases where the extant legal materials provide no determi-

III. FOUR FEATURES OF MORALITY

We would also need to know much more about the idea of

1. Curbing the Passions and Sacrificing Personal Interest

degree to which morality is other-regarding is a matter of complex and

2. Serious Social Pressure

duty. It is one thing to disapprove of and criticise failure to rescue but

3. The Consequences of Immorality

B. Immunity from deliberate change

whether conventional, critical or personal, and whether based on

2. Morality “there” to be recognised

At all events, Hart gives us no good reason to accept that immunity

C. The Nature of Moral Offences

D. The Form of Moral Pressure

“only threats of physical punishment or unpleasant consequences”

IV. MORALITY AND CONFLICTING REASONS FOR ACTION

moral reasoning and deliberation. In particular, I suggest that a person

in their moral reasoning because the law possesses institutional

standards given law’s capacity, by reason of its institutional resources,

suggested that Hart’s lack of success in this regard is the analytical

You might also like