Enforcing Morality1
(Criminal Law and Philosophy 2013)
Fifty years ago H.L.A. Hart initiated the “Hart / Devlin debate” with the publication of
Law, Liberty and Morality.2 And he won that debate. Hart established that Devlin had failed to
make a sound case for the general claim that political authorities permissibly can use the criminal
law to enforce the positive or social morality of their societies. As Hart showed, Devlin was
wrong to think that changes in the actual social morality of a society must threaten its very
existence and he was wrong to think that states have a general right to preserve and protect the
social morality of their societies.
But while Hart won this debate, the debate was disfigured
from the start by its focus on positive morality, as opposed to critical or sound morality. Hart
claimed that the “modern form” of the debate over the legal enforcement of morals centered on
the “significance to be attached to the historical fact that certain conduct, no matter what, is
prohibited by a positive morality.”3 This form of the debate was politically important in 1963 in
Britain and America, and it remains politically important in these countries today and elsewhere;
but it is not the philosophically most interesting form the debate can take. An older form of the
debate appealed to natural law. It centered on the question of whether political authorities could
properly use the criminal law to enforce the natural law, including natural law prohibitions on
conduct that was not harmful or disrespectful to others.4 The language of natural law has
associations that can be distracting, but it conveys the key idea that morality can be sound or
true, that it does not refer merely to the norms that are accepted by a particular society at a
particular time, and that it can be used to criticize the accepted norms of any actual society. Hart
termed this idea “critical morality.”5 The Hart/Devlin debate, as Hart pointed out, is a debate
within critical morality, but it is was a debate about positive morality. The older natural law
debate was a debate both within and about critical morality. This paper engages with this older
2
form of the debate. It offers some reasons for thinking that there is a presumption in favor of the
view that it is a proper function of the criminal law to enforce critical morality, including that
part of critical morality that is not directly concerned with preventing harm or disrespect to
others.6 It then defends this presumption against some arguments recently pressed by Ronald
Dworkin.
I. Articulating the Presumption
The presumption that I want to discuss – hereafter simply the ‘Presumption’ – stands in
need of clarification, as it is easily misunderstood. It is a presumption in favor of legal moralism,
but legal moralism is itself an elusive position. The legal moralist, as I conceive him, is
concerned with good character and with living well, and he thinks that the law, including the
criminal law, can be used to further these concerns. But to get a reasonably clear view of the
Presumption, and hence of legal moralism as I construe it, we must attend to some distinctions. I
have already said that the morality I have in mind is critical, not merely social; but this is just the
first step in clarifying the Presumption.
Not all critics of the Presumption are skeptics about critical morality and many will
accept that it is a proper function of the criminal law to enforce it. The term “proper function” is
vague, but I mean something reasonably precise by it. If it is a proper function of the criminal
law to enforce x, then it is in principle morally permissible, and may be morally required, for the
criminal law to enforce x.7 In articulating the Presumption, I will refer to different domains of
critical morality, and the claim that it is a proper function of the criminal law to enforce a
designated domain will mean that it is at least morally permissible, and may be morally required,
for the criminal law to enforce measures that fall within that domain.
3
Friends and opponents of the Presumption can agree on the general claim that the
criminal law properly enforces morality, but they will part company over more specific claims
about which parts or domains of morality are, or can be, properly enforced by the criminal law.
Without pretending that the divisions are either easy to draw or exhaustive, I will distinguish four
domains of critical morality. These concern autonomy, well-being, character and excellence.
There is overlap between the domains. The same criminal law may protect both the autonomy
and the well-being of those who are subject to it, for example. Still, the domains, as I construe
them, are not coextensive. By attending to them we can come to an understanding of the content
of the Presumption.
Autonomy has been understood in different ways. Here I assume that it refers to a
sovereign right of self-government to make choices from a wide range of valuable options, a
right that grounds a range of claims against coercive and manipulative interference from others
to engage with these options. Well-being refers to a person’s good or interests. If some event
sets back a person’s interests, then his well-being declines. Character refers to moral
dispositions to act and feel in certain ways. The dispositions are moral in a very broad sense.
They encompass both other-regarding and self-regarding virtues and vices. Some people think
that character is a necessary constituent of well-being. As one’s character is corrupted, one’s
well-being declines, holding other things constant. Here I assume that there is no such tight
connection between the two concepts. One’s character can be elevated or debased, I assume,
without a change in one’s well-being. Finally, excellence, as I understand it here, refers to
objects that have impersonal intrinsic value. The paintings in the Louvre and the natural beauty
of the Grand Canyon are candidates for excellence in this sense. These objects have intrinsic
4
value, and their value may not be fully accounted for by their effect on the autonomy, well-being
and character of human beings.8
Since these four domains of critical morality are distinct, they can inform distinct theses
on the proper legal enforcement of critical morality. The Presumption can be articulated by
identifying the theses it commits us to. Start with autonomy.
E1: It is a proper function of the criminal law to enforce respect for the autonomy
of persons.
Critical morality, I am assuming, includes respect for autonomy and so those who accept the
Presumption should grant E1. Drawing inspiration from Kant’s legal and political philosophy,
some people hold that the criminal law should not be used to pursue any goal other than the goal
of respecting and protecting autonomy of persons.9 They will grant E1 and hold the line there.
But it is possible for people to be harmed or wronged in ways that do not infringe their
autonomy. Hart claimed that persons have an interest in not being subjected to profound offense
and that the criminal law can be properly used to protect this interest. The setting back of this
interest need not constitute an infringement of autonomy. Others have suggested that one can
injure a person without setting back his autonomy.10 Yet even if our autonomy is not infringed
when others subject us to profound offense or injure us in these ways, our well-being may be set
back by their doing so. With this possibility in mind, consider
E2: It is a proper function of the criminal law to protect persons from rightsviolations that lower their well-being.
5
Not all rights-violations infringe autonomy. If being subjected to pain or profound offense can
set back our interests, then this fact may explain why we have a right to be free from it.
Accordingly, many will want to accept both E1 and E2. Accepting both theses can present
conflicts. For example, if the criminal law could protect persons from rights-violations that
lower their well-being, but it could do so only by infringing the autonomy of others, would it be
legitimate for it to do so? I return to the question of conflicts between enforcement theses below.
For now, notice that neither E1 nor E2 addresses the issue of legal paternalism. The legal
paternalist allows legal officials to interfere with people’s choices and decisions for their own
good. In criticizing Devlin, Hart pointed out that one can reject Devlin’s position – his “legal
moralism” – while affirming legal paternalism. And Hart expressed considerable sympathy for
the paternalist position, citing “a general decline in the belief that individuals know their own
interests best, and to an increased awareness of a great range of factors which diminish the
significance to be attached to an apparently free choice or to consent.”11
Hart’s discussion of legal paternalism did not mention the important distinction between
“soft” and “hard” paternalism. Both types of paternalism allow legal officials to interfere with
the activity of persons for their own good. But soft paternalism, which some claim is not really
paternalism at all12, limits this interference strictly to cases in which people act with a defective
will13, whereas hard paternalism allows interference in cases in which people act with a fully
nondefective will. Much more could be said about this key distinction, but this brief
characterization should suffice for present purposes. It allows us to formulate two further
enforcement theses.
E3: It is a proper function of the criminal law to restrain persons from harming
themselves when they act with a defective will.
6
E4: It is a proper function of the criminal law to restrain persons from harming
themselves even when they act with a nondefective will.
From Hart’s remarks it is not clear whether he was expressing sympathy for both E3 and E4 or
only E3. But many contemporary writers on legal paternalism unambiguously affirm E3 and
reject E4.
The claim that people can harm themselves, and the claim that it may be permissible for
the criminal law to restrain them from doing so, do not on their own tell us what kind of actions
constitute self-harmings. Speaking very loosely, we can distinguish standard harms from moral
harms. The former include garden variety physical, psychological and financial setbacks of
interests. The later refers to setbacks to our interest in having a good character. Earlier I
suggested that we construe character broadly. It includes moral dispositions, but also prudential
dispositions. The idea of character in this broad sense is related to living well.14 Possibly, it is
always in our interests to live well. If so, then moral harms would constitute setbacks to our
well-being. But here I want to allow that well-being and living well are independent notions.
Moral harms can prevent us from living well, but they need not set back, and (in some contexts)
may even augment, our well-being. Since legal paternalism is animated by a concern for the
well-being of those who are subject to the interference, we should, accordingly, construe E3 and
E4 as referring exclusively to standard harms.
Many people will think that the criminal law should not be indifferent to moral harms,
however. They will think that the criminal law should encourage good dispositions and suppress
bad ones, or at least that this is, in principle, a proper function of the criminal law. These people
will accept a fifth enforcement thesis.
7
E5: It is a proper function of the criminal law to promote good character, and to
restrain or discourage people from engaging in activities that cause moral harm to
themselves or to others.
Having and sustaining a good character is part of living well. And the law, including the
criminal law, may have a role to play in enabling or assisting those who are subject to it in
achieving this good. A person with good character, it may be said further, would be disposed to
respect excellence. He or she would not be inclined to destroy intrinsically valuable objects,
unless there were some compelling reason for doing so. Yet the main reason why people should
respect excellence is not because doing so is part of having a good character, which is part of
living well; but rather because objects of excellence simply ought to be respected and preserved.
There are many reasons why objects of excellence ought to be respected and preserved.
One reason why one should not destroy valuable art is that it will deprive others of opportunities
to experience the art, for example. But, or so it may be thought, there are reasons to respect and
preserve objects of impersonal intrinsic value that do not derive from the interests of human
beings. The reasons to prevent people from destroying the rain forests may not be exhausted by
the effects this would have on the interests of present and future humans. Accepting the
existence of these reasons would commit one to rejecting the humanistic principle that the value
of all things derives ultimately from its contribution to human life.15 But, on the assumption that
reasons of this kind exist, they could be invoked to justify certain legal measures and we could
ask whether these measures could be a permissible part of the criminal law. This yields our final
enforcement thesis.
8
E6: It is a proper function of the criminal law to preserve objects of excellence
and to ensure that they are respected.
One could reject this enforcement thesis and still hold that the criminal law should be used to
protect objects of excellence insofar as damage to these objects would constitute a rightsviolation or harm to others. E6 identifies new grounds for the criminal law, grounds that are not
already included under the other enforcement theses we have distinguished.
We can now, finally, articulate the content of the Presumption. The first three
enforcement theses, E1 – E3, are not typically denied by those on Hart’s side of the Hart / Devlin
debate. Foes of the Presumption can allow that the criminal law can and should be used to
protect autonomy, enforce rights and prevent self-harms that result from a defective will. E4 is a
borderline case. Many who are inclined to side with Hart over Devlin are also inclined to reject
hard legal paternalism. Still, since E4 is concerned with preventing only standard harms and not
moral harms, it is not essential to the Presumption. Acceptance of either E5 or E6, by contrast,
puts one squarely on Devlin’s side of the debate. Some friends of the Presumption will accept
one of these two theses and reject the other; but the vindication of either would be sufficient to
justify the Presumption.
An adequate consideration of E6 would require an investigation into the plausibility of the
humanist principle. This would be a large undertaking, and it would take us far afield from the
issues that historically have animated the debate over the legal enforcement of morals.
Accordingly, in what follows, I will focus exclusively on E5. If E5 can be shown to be
defensible, then the Presumption will be justified.
A final clarification is in order. I have claimed that if one accepts the Presumption, then
one falls on Devlin’s side of the Hart/Devlin debate. Could one, however, accept the
9
Presumption and reject what Hart termed “legal moralism”? The answer is not straightforward,
since “legal moralism” is sometimes used by Hart and others to refer to the view that it is a
proper function of the criminal law to punish or suppress acts that are inherently immoral. The
immorality of such acts, it is said, needs no further explanation. In this sense of the term, one
could reject legal moralism and accept the Presumption. But this sense of “legal moralism” is
not, I think, the best one. Sophisticated proponents of legal moralism, including Devlin, offer
arguments for why certain alleged immoralities should be restrained. These arguments present
grounds that purport to justify legal restraint, grounds which do not advert to respect for
autonomy or the promotion of well-being. The Presumption articulates these grounds, or at least
some of them.
II. Self-Respect and Character
With the Presumption now in clear view, we can consider what might be said in support
of it. Many arguments might be advanced. My strategy here is to outline one argument that
supports E5. The argument appeals to the value of self-respect and to the state’s role in securing
this value for its members. Before presenting the argument, I need to revisit an assumption that
stands behind E5. This is the claim that living well and well-being are independent notions.
Even if these two notions are conceptually distinct, it might be true that setbacks to a person’s
character – what I have referred to as “moral harms” – always set back his or her well-being.
This would be so if it were in the interests of each person to have a good character. In this paper,
I have presumed that this substantive claim is false. But someone who was sympathetic to it
could construe the argument presented below as an argument for E4 and so not a proper defense
of the Presumption.
10
Since I will not exclude this view, I want to allow that the argument presented in this
section, even if sound, could fail to provide support for E5 as I have construed it. This
possibility, however, should be cold comfort to the critic of the Presumption. For I suggested
that E4 was a borderline case for friends and foes of the Presumption and I suggested further that
since it encompassed only standard harms its acceptance was not dispositive for the Presumption.
But, on the possibility that we are presently considering, E4 is read to include moral harms as
well as standard harms. On this broadened construal, acceptance of E4 plausibly makes one a
friend rather than a foe of Devlinite legal moralism. Having noted this complication, I set it
aside.
The argument I now present rests on three main claims. It will be helpful to lay them out
before commenting on them.
(1) If it is a proper function of the criminal law to protect and promote the well-being of
those who are subject to it, then it is also a proper function of the criminal law to
assist those who are subject to it to lead well lived lives.
(2) To live well a person must have a sense of self-respect (and merit it).
(3) To have a sense of self-respect, and to merit it, a person must be committed to
pursuing a sound conception of the good and must care about his character.
The conditional form of the first claim expresses a limit to the argument that I will present. As I
have noted, some hold that the only proper function of the criminal law is to enforce respect for
the autonomy of persons. This view will not be discussed here.16 We have seen, however, that
opponents of the Presumption need not deny that it is a proper function of the criminal law to
protect and promote the well-being of those subject to it, even when doing so is not necessary to
11
enforce respect for their autonomy.17 The first claim addresses critics of the Presumption, such
as Hart, that grant this much. It holds that if a concern for the well-being of its subjects is a
proper rationale for the criminal law, then a concern to help subjects lead well lived lives is also
a proper rationale for the criminal law.
The first claim rests on the appealing thought that a concern for people extends beyond a
concern for their well-being to include a concern for how well they live. That thought is
appealing, since, on a plausible conception of well-being, a person’s well-being is advanced if he
succeeds in his rational aims. His rational aims are those aims that he actually has and that he
does not have good reason to revise or abandon.18 (An actual aim to count blades of grass in an
open field would not qualify as a rational aim, since there is nothing to be said for pursuing it.)
If this is right, then a concern for a person’s well-being naturally extends to some concern that he
succeed in his rational aims. However, in the typical case, a person’s rational aims are not
centered on himself, but on goals and projects that he judges to be valuable or worthwhile.
Indeed, to succeed in his rational aims, a person may know that he must sacrifice his own wellbeing to some extent.19 In such cases, our concern for him pulls us beyond a concern for his
well-being to a concern that he succeed in his efforts to live well.
The basic point I am pressing here can be expressed in a different way. A person, insofar
as he cares about his life, aims to live well. Success in this aim may advance his well-being, but
he may not, and typically will not, pursue it in order to advance his well-being; and it may be
true, and he may know that it is true, that success in this aim will not advance his well-being
overall. To the extent that others, including legal officials to whom he is subject, have concern
for him, then they ought not be concerned only with how he fares, but also with how well he
12
succeeds in his aim of living well. Personal concern for him, in short, encompasses not only his
well-being, but also his efforts to live well.
These claims can be resisted; and a full defense of them is not possible here. I believe,
however, that many opponents of the Presumption would not be inclined to reject them. They
will accept that, in principle, the criminal law permissibly could be used to help people lead well
lived lives, but they will deny that it can do so effectively or without harm to other values. They
also may deny that living well is as tightly bound up with good character and a sound conception
of the good as the third claim in the argument above asserts.
So, granting the first claim, let us consider now what can be said in defense of the other
two claims. The second claim ties living well to self-respect. This claim is often asserted and
widely believed. As such, it provides a fairly uncontroversial entry point into the discussion of
what must be true for a person to live well. Following the implications of accepting this claim,
we will see, clears a path to legal moralism. To have self-respect a person must not deny his
own worth or be indifferent to whether he lives well or lives poorly. Consider Rawls’ well
known account of the idea. Rawls held that a secure sense of self-respect is an attitude that is
vital to living a good life. “Without it nothing may seem worth doing, or if some things have
value for us, we lack the will to strive for them.”20 This attitude, as Rawls correctly observed, is
judgment-dependant. It rests on the judgment or conviction that one’s plan of life or conception
of the good is “worth carrying out.”21 Rawls put the focus squarely on those who lack this
attitude, doubting their own value or the worth of their plans. And he pointed out, once again
correctly, that the social conditions in which people live, including the basic institutions of their
society, have a profound effect on whether people form and sustain the requisite attitude. But
Rawls said much less about the implications of the judgment-dependent character of this attitude,
13
and it is by reflecting on these implications that we can move beyond his discussion and come to
a more accurate view of self-respect and its connection to living well.
A person can fail to have self-respect if she fails to have the requisite attitude. But she
may have the attitude, but fail to have warrant for it. Enmeshed in worthless and degrading
pursuits, she may believe, falsely and unjustifiably, that her conception of the good is worth
carrying out and she may, as a result of this mistaken belief, acquire a lively sense of her own
worth and a self-confidence to pursue her ends. Reflection on this kind of case suggests that it is
warranted, not deluded, self-respect that is the good that contributes to the success of a life. Still,
even the person with a deluded sense of self-respect cares that her life go well rather than poorly.
Her failure is a failure to take proper responsibility for that attitude. If having the attitude of selfrespect is partly constitutive of living well, then a person with that attitude must also be
concerned that her sense of self-respect is warranted.
To appreciate the point better, consider the case of a person I will call Charlie. In
believing that his conception of the good is worth carrying out, Charlie will have the attitude that
Rawls identified. But, in virtue of being a person, Charlie has the capacity to critically evaluate
his conception of the good, and if he comes to see that it is misguided, then he will cease to
believe that it is worth carrying out. At this point, Charlie will lose the attitude in question, since
that attitude was premised on beliefs that he now rejects.
You might object that the attitude of self-respect is valuable to Charlie, and that it would
remain valuable, even if it were premised on mistaken and unwarranted beliefs. Suppose Charlie
says: “I am committed to a certain conception of the good, but even if I am wrong about its
value, it would still be a good thing for me to believe that it was worth carrying out.” In saying
this, Charlie would reveal that he did not have the attitude that Rawls identified. He does not
14
think that it is important that he live well. Rather, he thinks that it is important for him to believe
that he is living well. In being indifferent to whether or not he, in fact, lives well, Charlie would
thereby show that he did not really respect himself.
Is this judgment too harsh? Charlie might think that it is best both to believe that one’s
conception of the good is worth carrying out and for it, in actuality, to be worth carrying out.
But he also might think that if one’s conception of the good were not worth carrying out, then it
would remain good for one to believe that it was worth carrying out. Charlie might concede that
the belief that one’s conception of the good was worth carrying should be responsibly held. But
so long as this belief was not unjustified, then a person would have self-respect and self-respect
would contribute to his living well, even if the belief were false.
If this were Charlie’s claim, then it is not implausible. I have been exploring the thought
that living well consists, in part, in having the right attitude towards one’s life and the pursuits
that contribute to its value. In caring that he live well, and in having responsible beliefs about
the value of his pursuits, Charlie would evince the right attitude towards his life. But now we
need to ask about the value of warranted, but mistaken, self-respect. Anyone that values selfrespect must think that it is more valuable when it is premised on true rather than false beliefs.
The attitude of self-respect, I have been emphasizing, is judgment-dependant, and this feature of
the attitude implies that those who have it must care about actually living well as opposed to
merely believing, however responsibly, that they are living well. Accordingly, to merit selfrespect, one must accept that in order to live well one must pursue a sound or worthwhile
conception of the good.22
Thus far I have said that to live well one must have the attitude of self-respect, and I have
argued that an implication of having this attitude is that one must be committed to pursuing a
15
sound conception of the good. The third claim in the argument presented above adds that a
person with self-respect must also care about his character. There is a tight connection between
having a good character and pursuing a sound conception of the good, at least on the broad
construal of character that I am are here assuming. Character traits are stable dispositions to
respond, either well or poorly, to moral and prudential reasons. A person whose character is
corrupted or debased will be systemically unresponsive to reasons that inform a sound
conception of the good.23 He may try heroically to go against the grain of his bad traits to
engage with valuable pursuits and avoid base ones, but even if he occasionally succeeds, he will
almost certainly fail in the main unless he manages to refashion his character in a way that makes
it more responsive to the reasons that apply to him. That is why loving parents, to take an
obvious example, are so concerned with the character formation of their children. They want
them to live well, and they know that doing so requires that good traits be cultivated and bad
ones rooted out or weakened. This fact also explains why the citizens of a political society may
sensibly be concerned with the character of the moral and cultural environment in which they all
live, since they may sensibly believe that this environment plays an important causal role in the
character formation of those who live within it.
There is, then, a strong case for the claim that having a good character is vitally important
to living well and, therefore, to having self-respect and meriting it. But could a person have a
good character without caring about it? It is a commonplace that we are not fully responsible for
the beliefs, desires, traits and dispositions that constitute our character. We can play a role in
shaping our character, but much of what makes us the kind of person that we are did not arise
from our conscious choice and is not subject to our control. So it is certainly possible for one to
have an exemplary character without having thought at all about the matter. One may be so well
16
formed that one does not even consider the problem of cultivating virtue and suppressing vice in
oneself. But could one have a good character and be indifferent about this fact? Possibly, but
given the tight connection between having a good character and living well, a person who was
indifferent to whether or not he had a good character would manifest an attitude inconsistent
with self-respect. For that attitude would, barring the very exceptional case24, manifest a
disregard for living well. Accordingly, if people are to live well, they must not only have a good
character, but also care that they have such a character. And if they have this concern, then they
will not be indifferent to the character of the ethical environment in which they and others live.
They will want that ethical environment to be supportive of, not inimical to, the development and
maintenance of good character.
Having now defended the second and third claims of the argument presented above, it is
appropriate to return to the first claim. That claim ties living well to the exercise of political
power. And it sweeps broadly, and many will think much too broadly. It maintains that it is a
proper function of the criminal law to help its subjects live well, and living well requires more
than being autonomous, being respected by others, and having a decent level of well-being. It
also requires self-respect and a good character. We can use an old fashioned term – “soulcraft” –
to refer to the use of the law, including the criminal law, to foster an ethical environment that
promotes self-respect and good character. There may be special reasons for thinking that even if
the criminal law generally should be used in the service of helping its subjects live well, it should
never engage in soulcraft. Thus the argument presented here, if successful, only establishes a
presumption. That presumption – what I have referred to as the Presumption – may be undercut
or overridden by the special reasons that rule out this type of political action.
17
III. Authenticity and Independence
No attempt will be made to establish that no such special reasons exist. It would be
foolish to make such an attempt, since it would be foolish to try to anticipate and respond to
every argument that might be pressed in support of such reasons. Instead, I will consider a range
of arguments that take very seriously the idea that political and legal authorities should be
concerned with enabling those who are subject to them to live well, but then reject the legal
moralism that we are now considering. These arguments pose an interesting challenge to the
Presumption, since they do not deny its underlying rationale. Appreciating why they fail will
add a further measure of support to the Presumption.
Before turning to these arguments, a preliminary point is in order. I noted above that the
enforcement theses distinguished in section I might be thought to come into conflict. In
particular, those who maintain that the criminal law should be used to protect the autonomy of
those subject to it may think that a proper respect for autonomy excludes the measures we are
now considering. This is an important objection, and it is one that critics of legal moralism have
often pressed. I have argued elsewhere that a due regard for autonomy does not preclude all
efforts by the state, including its use of the criminal law, to sustain a moral and cultural
environment that enables its subjects to live well, or facilitate their efforts to do so.25 I will not
repeat those arguments here. I will assume that soulcraft, or at least intelligently executed
soulcraft, is not inimical to a proper regard for autonomy.
In making this assumption, I am in at least partial agreement with Ronald Dworkin, who,
as we will see in a moment, advances arguments against the Presumption that I want to consider
here. Dworkin distinguishes the value of authenticity from autonomy, where autonomy requires
that people have access to a sufficiently wide range of options. He writes:
18
A person’s autonomy is not threatened, on this view, when government
manipulates its community’s culture so as to remove or make less eligible certain
disapproved ways of living, if an adequate number of choices remain so that he
can still exercise the power of choice.26
If the “disapproved ways of living” are rightly disapproved, and if the government uses
proportionate means to discourage them, then Dworkin’s claim is substantially correct. But
Dworkin is keen to conclude from this not that these governmental measures can be justified, but
rather that they should be rejected out of hand because they offend the different value of
authenticity. Authenticity demands that we strive for independence in our ethical lives. It is
“damaged when a person is made to accept someone’s judgment in place of his own about the
values or goals his life should display.”27
Authenticity, as Dworkin presents it, is probably better described as a component of
autonomy rather than a value distinct from it. But however it is best described, Dworkin is right
that authenticity has had an important “life in our literature and in much of our most influential
philosophy.” It is, he claims further, an integral component of self-respect.28 If this last claim is
right, and if authenticity is indeed damaged whenever a person is made to accept someone’s
judgment in place of his own about the values or goals his life should display, then we have a
nice rebuttal to the Presumption. The rebuttal can grant the third claim in the argument we
advanced, but then use the second claim to reject the first. If a sense of self-respect is necessary
to live well, and if efforts by political and legal authorities to help citizens live well necessarily
undermine their self-respect by offending authenticity, then our argument is divided against itself
and the Presumption, to the extent that it rests on that argument, is called into doubt.
19
We have ample reason, accordingly, to take a closer look at the value of authenticity and
the particular implications Dworkin claims follow from it. Each person, on Dworkin’s view, has
a special responsibility for the success of his own life. We cannot live well if we abjure all
responsibility for our decisions. To realize authenticity, however, we must do more than assume
responsibility for our lives. We must also make decisions that follow from and reflect our
convictions and character. These convictions and character, moreover, must be genuinely our
own. They must not be the product of coercion or manipulation from others. “Authenticity
demands that, so far as decisions are to be made about the best use to which a person’s life
should be put, these must be made by the person whose life it is.” Dworkin refers to this last
aspect of authenticity as “ethical independence;” and he argues that governments, and the laws
they pass and enforce, violate the ethical independence of their subjects, and thereby damage
their self-respect, whenever they restrict their freedom on the grounds that some controversial
views about how to live well are better than others.
The soulcraft we have been considering, even if it intelligently favors the good over the
bad, is, on this view, self-defeating. In aiming to help citizens live well, the practitioners of
soulcraft will compromise the authenticity of those they aim to help, an authenticity that is
necessary for living well.
Let me begin by assuming that Dworkin is right that authenticity is a necessary ingredient
of self-respect. On Dworkin’s view, to have self-respect, one must take one’s life seriously, one
must care that one live well rather than poorly. This is implied by the view of self-respect I have
defended, but I have argued that merited self-respect, as opposed to the psychological attitude
that Dworkin identifies, requires in addition that one be committed to pursuing a sound
conception of the good and that one not be indifferent about having a good character. So, if I am
20
right that merited self-respect is the relevant value, and if Dworkin is right that authenticity is a
necessary ingredient of self-respect, then the issue of how the state might help its members live
well becomes a quite complicated matter. On the one hand, the failure of political authorities to
undertake soulcraft measures could make it harder for their subjects to pursue sound pursuits and
develop good character traits. On the other hand, undertaking these same measures could
compromise the authenticity of those subject to them, thereby frustrating their efforts to live
well.
Seen in these terms, an argument that appeals to self-respect neither clearly supports nor
opposes the case for soulcraft. Perhaps both concerns – promoting merited self-respect and
honoring authenticity – need to be balanced against each other to reach verdicts about the
advisability of soulcraft in particular cases.29 The tension between the two concerns, however,
will diminish to the extent that honoring authenticity does not have the stark political
implications that Dworkin claims. And I now want to argue that this is indeed the case. Whether
or not authenticity is a necessary ingredient of living well, it is not threatened by every political
measure that falls under the description of soulcraft.
Authenticity demands ethical independence. Interpreted incautiously, the idea of ethical
independence looks absurd. No one escapes the influence of others. We are all, in part, the
products of the ethical environments in which we live. Dworkin grants as much, as he must. His
point is that we can distinguish two kinds of ethical environment. The first is “created
organically by the [free and uncoordinated] decisions of millions of people;” the second is the
result, in part, of political action designed to shape or influence its character. Ethical
independence, on Dworkin’s view, is protected by the first kind of ethical environment and
destroyed by the second.
21
This account of the distinction between the two kinds of ethical environment, however,
needs refinement. So long as anarchism is renounced, political action that is justified by moral 30,
as opposed to ethical, considerations will be unavoidable. Further, political action so justified
undoubtedly affects the character of the ethical environment that results from it, and it affects it
in ways that have serious consequences for people’s efforts to live well. Dworkin’s view, then,
is best understood as holding that morally justified political action that shapes the character of
the ethical environment of the society in which it is undertaken does not compromise the ethical
independence of the members of the society, since its impact on the ethical environment is a
mere byproduct of what morality demands. Only when political action is undertaken for the
purpose of shaping or influencing the character of the ethical environment of the society does it
compromise the ethical independence of its members.
Thus the distinction that Dworkin needs – the distinction between the two kinds of ethical
environment, one supportive of, the other hostile to, ethical independence – rests crucially on
two further distinctions: the distinction between morally justified state action and state action
justified by ethical concerns and the distinction between purposeful and by-product effects on the
ethical environment. Both of these latter two distinctions are problematic in the present context,
however. For, to begin with, the proponent of soulcraft maintains that citizens are morally
permitted, and may have a moral duty, to help bring about or sustain an ethical environment that
facilitates the realization of self-respect for those shaped by it.31 When political authorities
competently undertake political and legal measures designed to shape the ethical environment,
they are helping citizens to comply with this moral demand. Political morality – what we owe to
one another in politics – on this view is not confined to protecting liberties and ensuring a fair
22
distribution of resources. It extends further to encompass political and legal efforts that seek to
help citizens live well.
The manner in which Dworkin employs the distinction between morally justified state
action and state action justified by ethical concerns, accordingly, begs the question against the
proponent of soulcraft. The issue in dispute concerns the scope of our moral duties in politics.
That issue cannot be settled by stipulation. The second distinction, that between purposeful and
by-product effects, of state action is also suspect in the present context. A political society will
have an ethical environment and that ethical environment will be shaped by its political and legal
decisions. These decisions can affect the shape or nature of that environment. Let us say that a
political society countenances certain of these effects when they are the foreseeable, and
foreseeably avoidable, result of its political and legal decisions. So understood, a political
society can countenance a certain ethical environment either by deliberating undertaking action
or by failing to do so. Countenanced ethical environments are not aptly described as those that
arise organically from the free and uncoordinated decisions of individuals. They are the products
of deliberate political action. But the question of what kind of ethical environment political and
legal authorities should countenance is precisely the question posed by the proponent of
soulcraft. An example may help to illustrate the point. If (i) legal measures could be taken to
effectively discourage prostitution, and (ii) this is known by state officials, and (iii) one salient
effect of not undertaking these measures is that attitudes to sex in the society generally will be
changed in ways that are detrimental to good character, and (iv) this detrimental effect is
foreseen by the state officials, then it will be countenanced by them if they decide not to
undertake the legal measures in question.
23
This decision may be the right one to make, but it is not a decision that can be adequately
defended by the claim that the effect was a mere byproduct of political action. The state officials
must defend the claim that the effect was rightly countenanced. To claim that it is merely the
organic result of free action, and so not subject to moral assessment and justification, would be
like claiming that severe poverty that was caused, but not purposefully caused, by bad economic
policy is an organic effect and so not one for which state officials are responsible. With
soulcraft, as with economic policy, once the effects of state action are foreseen, and once it is
realized that these effects are avoidable, then the countenancing of the effects by state officials is
a political decision that demands justification.
In pressing this claim, it is not necessary, nor would it be sensible, to insist that there is
no moral difference between deliberately imposed, and merely countenanced, effects of state
action. Other things equal, a state that imposes poverty on a group by enforcing discriminatory
legal rules is more unjust than one that allows a group of the same size to be in the same state of
poverty, when this outcome could be avoided without great cost. The difference here plausibly
lies in the attitude expressed by the actions or omissions of state officials. Direct imposition of a
harm often expresses more contempt or disrespect than merely allowing it to transpire. But this
point, important as it is, does not show that when the effects of state action are foreseeable and
foreseeably avoidable, that they no longer stand in need of justification and can be considered the
mere organic offshoots of the free activity of individuals.
Thus Dworkin’s account of the distinction between an ethical environment that is
supportive of, and one that is hostile to, the ideal of ethical independence looks hard to sustain.
And without a defensible distinction of this kind, his case against soulcraft collapses. A couple
of supplementary considerations can now be addressed. Living well, Dworkin claims, consists in
24
responding well to the challenges one’s life presents, but “a challenge cannot be more
interesting, or in any other way a more valuable challenge to face, when it has been narrowed,
simplified, and bowdlerized by others in advance.”32 When the law screens out bad options from
the set of options that people confront in deciding how to live, it decreases the value of their lives
by distorting the challenges they confront. This is not a good argument, as I explain below. But
it is instructive that, if accepted, the argument would put further pressure on Dworkin’s
distinction between the two kinds of ethical environment. Without state action, the challenges
that people confront, after all, could become less interesting and less valuable as a result of the
uncoordinated decisions of millions of people. Presumably, then, a state that was concerned with
enabling its citizens to live well would have reason to deliberately undertake action to help
ensure that the challenges they confront were sufficiently robust.33 Under some circumstances,
the necessary state action might include support for worthless options of various kinds, the very
options that intelligent soulcraft aims to discourage. Reflection on this possibility, however,
reveals the weakness of the argument. Even if – which I doubt – a state could foreclose
effectively so many disvaluable options that the lives of its members become flat and
uninteresting and thus less valuable, this would establish merely a limit to soulcraft, not the
blanket prohibition that Dworkin envisages. Sound measures, in this unlikely scenario, would
need to stop short of the full program, but not be abandoned altogether.
A final consideration should be mentioned. A friend of Dworkin’s argument for ethical
independence might claim that deliberate state action that seeks to shape the ethical environment
is disrespectful to citizens in a way that mere “organic” changes to the ethical environment are
not. It is disrespectful insofar as it presumes that some citizens are not fully capable of making
their own decisions about how to live well. Soulcraft, in short, disrespects people by denigrating
25
their status as rational decision makers. It is instructive to compare this claim to the different
claim that when the state enforces a controversial conception of justice it thereby presumes that
some people are not fully capable of making rational decisions about justice. Few will press this
latter claim, however. It is no insult to one’s standing as a rational decision maker to treat one in
ways that presume that one’s decision making capacities, when applied to matters of justice, are
not perfect, but subject to error. Should we not say the same thing about the task of living well?
The natural reply is that we have no choice but to enforce justice. We have a moral duty to do
so; but with soulcraft we have a choice. We can let the ethical environment emerge organically,
as Dworkin recommends. The problem with this reply is that it assumes that citizens have no
moral duty to engage in soulcraft, but this assumption is challenged by the Presumption. If the
Presumption is vindicated, then the reply loses its force. Of course, the Presumption may not be
vindicated. It may be defeated by considerations that I have not considered; but, if so, then it
will be these considerations, and not the disrespect claim, that explain why soulcraft should not
be undertaken. The appeal to disrespect cannot stand alone, but must ride piggyback on these as
yet unidentified considerations.
IV. Beating a Retreat
As the previous sentence suggests, I do not wish to claim that I have established more
than I have. I have defended a presumption in favor of the legal enforcement of morals, a
presumption that places me squarely on the Devlin side of the Hart/Devlin debate, and I have
offered an argument in support of this presumption, an argument that runs through the value of
self-respect, rightly understood. But I have not shown that every argument that could be pressed
against the presumption must fail. I have said nothing here, for example, about libertarian selfownership, which obviously conflicts with both legal paternalism and legal moralism. And I
26
have said nothing about the currently fashionable political liberal argument for state legitimacy,
an argument that may speak against the soulcraft I have been discussing. And so some may be
willing to grant what I have argued, but believe that I have not targeted the best case for rejecting
the view expressed by the Presumption. But others, I suspect, will have a different reaction. The
case against soulcraft, they will say, should not be conducted at the level of high principle. We
should reject soulcraft because we have learned from hard experience that it cannot be done well.
State officials, at least in the modern political societies with which we are familiar, are either too
unenlightened or too corrupt to carry out the task well. And the dangers of abuse are sufficiently
serious that good sense recommends taking the whole issue off the political agenda. We should
allow the ethical environment to emerge organically, not out of high principle, but out of prudent
distrust.
My own sense is that these concerns, while important, are exaggerated. The concerns are
important since the criminal law is not in general a good means for helping people to live well.
Noncoercive efforts to shape the ethical environment are preferable to those that rely on force
and threats. And even when legal coercion is necessary, it is better if it does not punish, but
instead makes it more difficult for people to engage in the disvaluable activity.34 It is also true
that modern governments, in many circumstances, are not competent, and/or not motivated, to
make the kind of judgments necessary to execute intelligent soulcraft. But all of these claims are
generalizations, and they likely admit of important exceptions. Prudent distrust does not, I
believe, support an absolute ban on soulcraft, even if it points to considerations that any
proponent of soulcraft must take very seriously.35 But even if I am wrong on this last point, and
prudence does indeed counsel an in-practice ban on soulcraft36, it is worth emphasizing just how
significant a retreat this response would be for the opponent of legal moralism. In retreating to
27
prudential and pragmatic considerations, he or she would concede the in-principle permissibility
of the legal enforcement of morals, as Delvinites conceive it. This kind of opponent would be
like the person who rejects religious persecution or torture, not on the grounds that it is evil, but
because it simply will not work. Liberals from Hart to Feinberg to Dworkin have wanted more.
They have sought to erect a strong barrier of moral principle against the kind of legal moralism
that I have been discussing. This paper has engaged the debate on this plane. That is why it has
neither identified particular types of conduct that are proper targets for criminal penalties nor
discussed the dangers of granting legal officials, in this or that circumstance, the moral license to
impose such penalties.
1
An earlier version of this paper was presented at the 4 th annual New Orleans Invitational Seminar in Ethics in
March 2013. Thanks to participants at this event, and especially to David Lefkowitz who served as a commentator
on the paper, for comments and criticisms. Thanks also to Massimo Renzo for excellent comments on the
penultimate draft.
2
H.L.A. Hart, Law, Liberty and Morality (Stanford University Press, 1963). Devlin’s contributions to the debate are
collected in his The Enforcement of Morals (Oxford University Press, 1965). In large measure, the debate between
Hart and Devlin was provoked by the 1957 “Wolfenden Report” commissioned by the British government on
“Homosexual Offenses and Prostitution.”
3
Hart, Law, Liberty and Morality, pp. 23-24.
4
For a recent contribution to the debate in this vein see R. George, Making Men Moral (Oxford University Press,
1993).
5
The existence of critical morality can be rejected, of course. Various relativistic views of morality will find no
place for it. This paper is premised on the falsity of such views.
6
Henceforth, when I speak of Hart or Devlin’s side in the enforcement debate, I mean to refer to opposing views on
the proper legal enforcement of critical morality.
7
Generally, I will use the terms “legitimate” and “permissible” interchangeably. When I have moral requirements
in mind, I will speak of moral duties.
8
Naturally, many will deny that any objects or activities are excellent in this sense. These people can still consider
the implications these objects would have for the enforcement of critical morality on the assumption that they have
the value here attributed to them.
9
See, for example, A. Ripstein, Force and Freedom (Harvard University Press, 2009).
10
Raz suggests, for example, that some forms of injury, such as the infliction of pain, may not diminish a person’s
prospects or adversely affect his options, but they are still often regarded as harms. See The Morality of Freedom
(Oxford University Press, 1986), pp. 413-414.
11
Hart, Law, Liberty and Morality, pp. 32-33.
12
J. Feinberg, Harm to Self, p. 16.
13
To avoid misunderstanding, by “defective will” I mean a will that is not sufficiently voluntary or autonomous.
The voluntariness-reducing factors catalogued by Feinberg all point to defects of the will in this sense. See
Feinberg, Harm to Self, pp. 150-153.
14
Living well is sometimes identified with moral rectitude. But this is not how I understand the idea. On my
understanding, the moral saint may fail to live well. Living well requires one to lead a good life, and the best life is
not identified with either the life highest in well-being or the most virtuous life.
28
15
For a statement of this principle see Raz, The Morality of Freedom, p. 194.
The discussion of “authenticity” below engages with some of the concerns that motivate the autonomy objection
to legal moralism. But a full consideration of this objection will not be undertaken here.
17
Might it be a proper function of the criminal law to protect, but never to promote, the well-being of those subject
to it? It might be said that, from the standpoint of the law, there is reason to prevent people from falling below a
baseline level of well-being, but no corresponding reason to raise their well-being above the baseline level. I doubt
that this view could be sustained, however. We would need some way to specify the baseline level of well-being
that was determined independently of the law and the criminal law, and it is not at all clear how this could be done
in a compelling way. Hence, I speak here both of protecting and promoting well-being.
18
T. M. Scanlon, What We Owe to Each Other (Harvard University Press, 1998), p. 119. The claims in this
paragraph and next are indebted to Scanlon’s discussion of well-being and its limits. See especially pp. 126-143.
19
How is this possible? If succeeding in a rational aim advances a person’s well-being, then how could a person
know that to succeed in the aim he must sacrifice his well-being to some extent? The answer is that, while the
success of the aim will advance the person’s well-being in virtue of the fact that success in one’s rational aims
contributes to well-being, it also may impose costs on the person and these costs could exceed the well-being gains
that accrue from succeeding in the aim. In short, the person’s well-being is advanced along one dimension, but
diminished more significantly along others.
20
J. Rawls, A Theory of Justice (1999 edition), p. 386.
21
Rawls claimed further that self-respect rests on confidence in one’s ability to successfully carry out one’s plans
and intentions, as set by one’s conception of the good. Ibid., p. 386. However, it is less plausible to hold that this
kind of self-confidence is necessary to living well. Despite having low confidence in one’s ability, a person may
have worthy plans and she may succeed in them despite her lack of confidence in her ability to do so. Such a person
could live well.
22
The attitude of self-respect can be distinguished from the value of self-respect. The former is judgment-dependent.
The latter is fact-dependent. I have claimed that if one has the attitude of self-respect, then one will accept that its
value is fact-dependent, although I have allowed that, consistent with the attitude of self-respect, one might think
that deluded self-respect, while less valuable than nondeluded self-respect, remains valuable to some degree.
23
A person with good character would be disposed to respect excellence. But it does not follow from this that a
person who was disposed to respect excellence, and who responded well to the impersonal reasons grounded by the
value of excellent natural and cultural objects, would have a good character. Such a person might have many of the
moral and prudential vices associated with bad character, but still respond very well to the reasons of excellence.
And it is possible – think here of William’s discussion of Gauguin – that such a person could successfully pursue a
sound conception of the good, one oriented toward the achievement of excellence. If so, then the link between
having a good character and pursuing a sound conception of the good would be tight, but not water-tight tight.
24
For example, the kind of case mentioned in the previous note.
25
See my “Moral Environmentalism” in Paternalism: Theory and Practice, eds. C. Coons and M. Weber
(Cambridge University Press, 2013).
26
R. Dworkin, Justice for Hedgehogs (Harvard University Press, 2011), p. 212.
27
Ibid.
28
Ibid., p. 209.
29
Dworkin rejects this kind of balancing, but his reasons for doing so are obscure. For criticism of Dworkin on this
point, see R. Arneson, “Cracked Foundations of Liberal Equality” in Dworkin and his Critics, ed. J. Burley
(Blackwell, 2004).
30
For Dworkin morality refers to how we ought to treat others, while ethics refers to how we ought to live ourselves.
I do not myself accept this way of drawing the distinction, but I need to employ it here to articulate Dworkin’s
argument.
31
There is a gap between defending a moral permission to use the criminal law for a certain purpose and defending
the claim that there is a moral duty to do so. But often, if there is no moral duty to do so, then there will be no moral
permission to do so either. The reason for this is that the use of the criminal law is a grave matter. To justify its
permissible use, the moral considerations for using it must be compelling; and when these considerations are strong
enough to justify a moral permission they often will be strong enough to justify a moral requirement.
32
R. Dworkin, Sovereign Virtue (Harvard University Press, 2000), p. 273.
33
Curiously, this is a point that Dworkin himself appeals to in defending the state’s role in funding the arts. See
ibid., p. 274.
16
29
34
Criminal sanctions can stigmatize offenders and impose hard treatment on them. For this reason, some who are
otherwise sympathetic to legal moralism may insist that soulcraft should never be pursued by the criminal law.
Politics, they will grant, is properly concerned with soulcraft, but they will insist that we cannot help people live
well by threatening or punishing them. Yet the case for legal moralism looks not only to the effects of sanctions on
the offenders, but also to the wider effects on the ethical environment of the society in which it is undertaken. Even
if the criminalization of a targeted activity would not help those who were prosecuted to live better lives, it might be
justified all things considered because it would be effective in dissuading others from taking up pursuits that would
ruin or mar their lives. On this point see P. Devlin, The Enforcement of Morals, p. 110.
35
Compare these remarks with the similar view expressed by G. Sher, Beyond Neutrality (Cambridge University
Press, 1997), pp. 70-71.
36
An in-practice ban on soulcraft could be the prudent thing to do in some circumstances. Principles can be
distinguished from what we might call general policies. The former apply to a very wide range of circumstances.
By contrast, the latter are tailored to a more specific set of circumstances. General policies could exclude a whole
category of state action, such as soulcraft, on the grounds that doing so, in the specified set of circumstances, would
be the best response in light of all the relevant values. Nothing I have said in this paper implies that there could not
be sound general policies that exclude soulcraft in some circumstances.