Enforceability, Legitimacy, and the Right to do Wrong

I. THE DUTY-FIRST APPROACH: SOME DOUBTS

In what follows I’ll set out a force-first approach to the puzzle of enforceability. As I’ll argue, this approach not only gives us a satisfying solution to the puzzle; it also sheds light on the related “right to do wrong.” But before that, let me say something about the shortcomings of the duty-first approach beyond the brief observations above. To be clear: my intention at this stage is not to show that no version of the duty-first approach could possibly succeed, but to motivate exploration of the alternative force-first approach by drawing attention to the intuitive obstacles that a range of prima facie plausible versions of the duty-first approach face. I will, though, suggest a diagnosis that it will be the burden of the rest of the article to vindicate.

One plausible-seeming version of the duty-first approach distinguishes duties of non-interference from other duties and argues that features of the former justify their enforcement. In a recent development of this idea, Jessica Flanigan argues that duties of non-interference are enforceable because those who violate them cannot consistently complain about others’ interference with them in the form of enforcement. By contrast, there is no inconsistency in the complaints of a violator of other duties about interference to force her to conform. So all and only duties of non-interference are enforceable.⁵

Now, it is not clear that the consistency of a complaint against enforcing interference is enough to establish a claim against such interference. At best it would seem to pre-empt one reason (viz., inconsistency) to think that a claim that already obtained is forfeit; but no such claim can be presupposed without begging the question. Flanigan does adduce arguments from differential burdens on duty-bearers in further support of the unenforceability of duties other than non-interference. Yet these seem equally applicable to duties of non-interference.⁶

Even setting these doubts aside, the appeal to the distinction between duties of non-interference and other duties looks as if it must fail. For either it will treat promissory duties as duties of non-interference, and must therefore conclude that they are enforceable.⁷ Or it will deny that they are duties of non-interference, and therefore apparently cannot make sense of the enforceability of contracts. Neither of these options is attractive.

Another form of the duty-first approach appeals to distinctions not in the content but in the stringency of our duties. However, as Christian Barry and Emily McTernan argue, the solution to the puzzle of enforceability that this yields is also unsatisfying. Some associative duties, for instance, seem much more stringent than minor duties to respect property rights. Yet the former seem unenforceable and the latter enforceable.⁸ Even where two duties are alike in both content and stringency, as a contractual and a promissory duty might be, they may differ in enforceability.

Similarly, the distinction between duties that give us some discretion about how to discharge them and duties that don’t doesn’t furnish a satisfying solution to the puzzle. As Barry and McTernan note, we have a great deal of discretion about how to care for our children and no discretion about whether to cheat on a romantic partner or to keep our promises, yet aspects of the duty to care for one’s children seem to be enforceable, while the duty not to cheat and the duty to keep promises seem to be unenforceable.⁹

There are other forms of the duty-first approach that focus not on distinctions in the content, stringency, or type of duty but instead on the value of unforced conformity. But these don’t appear to offer satisfying solutions to the puzzle of enforceability either. Some influential defences of the right to do wrong suggest, for instance, that valuable autonomy calls for us to be able to do wrong without interference.¹⁰ But this isn’t clear: as critics have noted, there seems to be enough of a variety of morally permissible options for restriction to them to be perfectly compatible with duty-bearers’ giving distinctive shape to their lives in the ways that matter as far as the value of autonomy goes.¹¹ And anyway it is hard to see why valuable autonomy should require the freedom to violate the duties that seem intuitively unenforceable but not the freedom to violate the duties that are intuitively enforceable.

Alternatively, it might be argued that the expressive power of conformity with some duties depends on the possibility of non-conformity, so that enforcement would undermine the value of conformity. But although it may be true that the expressive power of conformity with intuitively unenforceable duties would be diminished by enforcement, since it would not then imply that the duty-bearer was attentive to the reasons that require it, the same is true of the expressive power of intuitively enforceable duties. After all, physical prevention and the threat of punishment diminish the power of non-interference to express respect for a person as an end in herself, as opposed to the fear of punishment or the assumption that it’s not an available option in the first place.¹² So this doesn’t seem an adequate account of the difference.

Perhaps it could be argued that the justification of some duties lies in the expressive power of unforced conformity, so that enforcement would be inconsistent with the values underpinning those duties; whereas that is not true of duties of non-interference, even if unforced compliance with the latter has expressive value too. But even if this kind of argument could be made to work, it looks as if it will once again produce the result that promissory duties, as well as duties of assistance and charity, the point of which is surely not expressive, are enforceable. And it does not seem that they all are.¹³

Some conclude on the basis of such problems that we must look to the wider social costs and benefits of enforcement to explain why some duties are enforceable and others aren’t, but even they concede that this isn’t intuitively the right kind of explanation, since it mislocates the complaint that duty-bearers have against the enforcement of unenforceable duties.¹⁴ Others conclude that the seemingly unenforceable duties are after all enforceable, even if there are sometimes good reasons—e.g. of uncertainty or reasonable disagreement or proportionality—not to enforce them.¹⁵ But it would evidently be preferable to find a response to the puzzle of enforceability that didn’t have these intuitive costs.

II. NATURAL RIGHTS AND ASYMMETRIES OF POWER

To reiterate, my aim in canvassing these doubts about specific duty-first approaches is to motivate interest in an alternative, not to refute them. Perhaps there are ways for duty-first approaches to get around the objections I’ve been highlighting. Nevertheless, I think many of the objections are traceable to a deeper and more serious structural problem. By attempting to determine enforceability by examining properties of independently binding moral duties or of forced conformity to them in isolation from questions of who is engaged in the enforcement, the duty-first approach overlooks the moral concerns generated by relations structured by force in themselves. By contrast, force-first approaches start from the moral logic of such relations. One way that the problem shows up in duty-first approaches is in an inability—exhibited by some of the accounts surveyed above—to distinguish enforceable duties from unenforceable counterparts that need not differ in respect of content and stringency, as in the case of contractual and promissory duties, or legally recognised duties not to interfere with a person’s property and legally unrecognised ones (such as might obtain between siblings within a family).

It may seem, however, that there is no real alternative. I said that the force-first approach starts not from the duties themselves, but from conditions on the legitimate wielding of force. Yet don’t considerations of enforceability feed into the conditions of legitimacy, so that unenforceability sets one of legitimacy’s limits, rather than the other way around? If so, the force-first approach will ultimately depend on prior conclusions that can be arrived at only by pursuing the duty-first approach.

Much contemporary political and moral philosophy, particularly in the context of a flourishing literature on self-defence, implicitly accepts a broadly Lockean account that would seem to fit this picture. On this view, we have certain natural duties not to harm or interfere with others, and what these duties protect is important enough to justify enforcing conformity with them as a matter of self- or other-defence. Legitimate political authority’s entitlement to use coercive force is just an instance of the enforcement entitlement so justified.

For instance, Locke himself describes a “Law of Nature” according to which “no one ought to harm another in his Life, Health, Liberty, or Possessions” on grounds that all are equally the workmanship of God.¹⁶ His reason for supposing that each person is naturally entitled to enforce this law of nature seems to be the importance of the duties (or of the security of the workmanship of God from violence) that the law prescribes.¹⁷ More recent Lockean accounts will typically replace the appeal to the workmanship of God with an appeal to the status of individuals as ends in themselves with distinct lives to lead, but nevertheless follow Locke in taking the importance of the value at stake to justify enforcement by anyone with the power to enforce.¹⁸

In Locke’s telling, coercive political authorities then gain their unique entitlement to use force—their monopoly on legitimate coercion—via the voluntary renunciation of individuals’ entitlements to punish others for violations of the law of nature that prescribes the natural duties.¹⁹ Punishment is conceived by Locke as a form of self- and other-defence, but renunciation of the entitlement to inflict it is rational in light of the “inconveniences” of the state of nature, which include a lack of impartiality in the application of natural laws to one’s own case and the fact that more generally most people are “no strict observers of Equity and Justice” anyway.²⁰ Legitimacy is conditional on the authority’s doing no more than enforcing the duties whose uncertain discharge in the state of nature rationalises the renunciation; if the authority goes beyond that mandate, the renunciation of individual enforcement entitlements is undone.²¹

Someone taking a Lockean view can pose the puzzle of enforceability, asking why the duties that Locke has in mind may be enforced whereas other duties may not. (Or, granting the widely accepted conceptual association between rights and enforceability, she might ask why those duties correlate with natural rights whereas others, such as promissory and imperfect duties, correlate with mere claims or nothing at all.)²² Given the association between enforcement and defence, it would seem that the only way to go about addressing the puzzle is via the duty-first approach, working out whether it’s only the duties Locke has in mind that are important enough to justify defence by force, or whether there are others too. The question how anyone could have standing to enforce any duty is off the table from the outset.

Locke’s own arguments raise doubts about this approach, though. He worries that since the law of nature is “unwritten, and so no where to be found but in the minds of men”, it cannot “determine the Rights, and fence the Properties of those who live under it” unless it takes the form of “promulgated standing Laws.”²³ If a political authority rules by “extemporary Arbitrary Decrees” instead, then the problems that attend the state of nature will not be solved by the civil state. Indeed they will be made worse:

It cannot be supposed that [men] should intend, had they a power so to do, to give to any one, or more, an absolute Arbitrary Power over their Persons and Estates, and put a force into the Magistrates hand to execute his unlimited Will arbitrarily upon them: This were to put themselves into a worse condition than the state of nature, wherein they had a Liberty to defend their Right against the Injuries of others, and were upon equal terms of force to maintain it … Whereas by supposing they have given up themselves to the absolute Arbitrary Power and will of a legislator, they have disarmed themselves, and armed him, to make a prey of them when he pleases.²⁴

Locke’s argument here is that submission to an arbitrary political authority would defeat the point of renouncing our natural enforcement entitlements. Upon reflection, though, the problem he is drawing our attention to cannot plausibly be supposed to arise only in the case of political authority. To see this, consider that—contrary to Locke’s assertion that in the state of nature we are “upon equal terms of force” to secure our rights—the logic of force and counterforce appears to imply that in the state of nature one person can secure rights by force only if someone else cannot. (If she could, then she could resist the force of the first person.) This second person is thus “disarmed, to be made a prey of” when the first person pleases. The question arises then why we should acquiesce in the state of nature to the strongest person’s enforcement of (unpromulgated) duties any more than we should acquiesce in an arbitrary political authority’s.

Of course, on Locke’s view we don’t acquiesce in the strongest person’s enforcement of unpromulgated duties in the state of nature. Acquiescence is a matter of renouncing enforcement entitlements, and in the state of nature we retain those. But once we recognise that the logic of force and counterforce effectively makes some the prey of others whether they renounce their entitlements or not, we might start to wonder whether the value of conformity to natural duties, however great, can really be enough to justify the enforcement entitlements. The asymmetries of power that mark the state of nature seem no less morally dubious than those that mark the relation of rule between an arbitrary political power and its subjects. We should hesitate to give the use of force against those who cannot resist it morality’s imprimatur, whether those using it are states or individuals.²⁵

III. KANT ON FORCE AND EQUAL FREEDOM

These doubts receive their most compelling and thoroughgoing elaboration in Kant’s political philosophy, and it is a Kantian version of the force-first approach that I will present.²⁶ For Kant, the moral logic of force conditions not only which duties may be enforced by whom, but also the moral authority of the duties themselves.

Kant begins not by identifying moral duties intuitively important enough to justify enforcement, but by asking how any instantiation of a person’s embodied agency through his choices can be consistent with the equal freedom of each other person likewise to instantiate her own agency through her choices—that is, with “equal external freedom.” In some cases this seems obvious. I can scratch my back, for instance, and that choice seems not to constrain you at all, leaving you no less free to scratch yourself. If that’s indeed correct, then the choice is consistent with what Kant calls our equal “freedom as independence from being constrained by the choices of others.”²⁷

Such consistency is not enough to show that a choice is not morally wrong, so this is not a general test for moral duty. Insulting someone, for example, meets the consistency test but not the test of compatibility with duty.²⁸ But it is enough to show that a choice to act is not wrong in one particular way: it is not inconsistent with a system of equal external freedom. And insofar as one can go ahead and make such choices even where this means physically precluding someone else’s performing some physically incompatible action, that is not inconsistent with a system of equal external freedom either. Kant says acts embodying such choices are right, though by this he does not mean that they are morally recommended or that failing to perform them is wrong (again, insulting someone is right in the relevant sense).²⁹

So on the assumption that my choice to scratch my back is indeed consistent with a system of equal external freedom, if I manage to scratch my back by virtue of my greater strength even as you attempt to pin my arm to my side, the force I thereby use against you inherits the scratching’s consistency with a system of equal external freedom as what Kant calls a “hindrance to a hindrance to freedom”—to equal freedom, that is. More generally, the threat and use of force to protect choices that are compatible with the equal freedom test are themselves consistent with it by the same token. So acts that are right in this sense are conceptually “connected with an authorisation to use coercion”,³⁰ and morality ratifies the boundaries of external freedom that the coercion secures in the form of moral duties correlative with rights on the part of the agent.³¹

On Kant’s approach, then, any entitlement to use force will not be based on the content or importance of independently binding moral duties that are enforced. On the contrary, the compatibility of the force and the choices it protects with equal external freedom is what validates the “juridical” system of rights as morally authoritative and therefore a source of moral duties. As Kant puts it, “ethical lawgiving … take[s] up duties that rest on another, namely an external, lawgiving [the coercive prescription that belongs to the realm of right] by making them, as duties, incentives in its lawgiving.”³²

That’s not to say that we can’t identify a priori the types of rights and correlative duties that belong to the realm of right. Indeed, Kant devotes a significant portion of the “Doctrine of Right” to doing so, distinguishing innate rights, acquired rights to things (property rights), acquired rights to others’ deeds (contract rights), and acquired “rights to persons akin to rights to things” (the relation that he thinks obtains between the head of a household and other members of it).³³ These exhaustively catalogue the possible forms that rightful private control of parts of the physical world can take; the distinctions between types of acquired right track distinctions between possible external objects of choice rather than distinctions in a list of independently binding moral duties.

But even though the kinds of duties that belong to the juridical realm and are therefore candidates for enforcement are in this way given a priori, what makes them enforceable, if they are, is the same thing that constitutes them as moral duties at all: that they belong to a system of external constraint that is compatible with equal external freedom and therefore authoritative. Where duty-first views derive conclusions about enforceability from reflection on the content, stringency, or value of independently binding moral duties (or unforced conformity with them), the Kantian approach starts with a principle governing the ways people may constrain each other’s external freedom and derives the duties from that. If a system of constraint is compatible with the principle, then the boundaries it imposes are ratified in moral duties whose enforceability is therefore already built in, rather than a further question.

IV. KANT’S PUZZLE

This might not seem to mark much of a breach with duty-first approaches if the rights and duties corresponding to the system of constraint that is compatible with equal external freedom were nevertheless settled and identified as enforceable prior to any consideration of questions of legitimacy, so that such questions concerned only who might have standing to enforce enforceable duties that individuals already had. But Kant also identifies a puzzle about how anyone could really have the rights and duties in the first place. Precisely because of the authorisation to use coercion and the associated moral ratification of correlative duties that rights involve, it turns out to be hard to see how this is possible. And the solution depends on distinctive considerations of legitimacy.

The puzzle comes out most clearly in connection with acquired rights. As we’ve seen, to have a right is for there to be a system of authorised constraint by force that protects the acts covered by the right from interference. Given that people don’t have acquired rights innately, coming to have such rights must involve the authorisation of such a system. However, for there to be a system of constraint by force is for someone to be able to successfully threaten and use force against others to impose it. And it follows from the logic of force and counterforce described earlier that if anyone can successfully wield force to constrain others, then she herself is not constrained. (If she encounters resistance, she can just use force to get her way.) Such a system of asymmetrical constraint is not on its face compatible with each person’s equal freedom as independence. So it seems impossible for there to be the system of authorised constraint that constitutes rights in the first place.³⁴

We want to say, of course, that one person’s constraining a second’s agency in this way is compatible with equal freedom when it limits the second person’s physical access to things that don’t belong to her, which she therefore has duties not to interfere with. But this is just to say that rights determine when force may be used to exclude others. And the point is that it is hard to see how anyone could have a right to anything external in the first place, precisely because rights authorise enforcement and enforcement by its very nature realises or reveals an asymmetrical distribution of constraint that looks incompatible with equal freedom as independence. Things can belong to someone in the sense that licenses forcefully excluding others only if the exercise of the power of enforcers to impose constraints on the freedom of others is compatible with a system of equal freedom. And that is what has been put in doubt.

Kant himself is clear that we can and indeed must be able to have acquired rights.³⁵ But he also sees that it’s a real question how that’s morally possible, given the concern about asymmetry of constraint. The difficulty is to show how the coercive boundaries set by exercises of power that purport to enforce acquired rights can be morally authorised rather than rejected as inconsistent with the equal freedom of all. Even if such an exercise of power happened to align with an independently binding moral duty on the part of the person subjected to it, that wouldn’t be sufficient to demonstrate its compatibility with equal freedom. As we’ve seen, acts contrary to moral duty can still be consistent with external freedom, and that includes interactions with external objects and people, such as some failures to assist (I gratuitously decline to throw you the life ring) and some cases of insisting on one’s rights (I expel you from my vast estate, forcing you to take the long way around and so miss your dying spouse’s last moments).³⁶

In one sense this approach to the question of enforceability does start from duty, since it begins with the question how one person’s exercise of the power to use force against others could be compatible with the equal freedom of all and so in that way not contrary to duty. But that does not make it a duty-first approach to enforceability.³⁷ For Kant, since acquired rights involve the authorisation of enforcement, whether acquired rights and their correlative duties are morally possible depends on whether the power to enforce them can be consistent with equal freedom. We have the moral duties only if the force is legitimate. So to find out which of our duties are enforceable, we don’t start from a list of independently binding moral duties. (The form of the duties of private right is given independently, but that does not confer moral authority on them as independently binding duties that could be the starting point for investigations into enforcement.) We start with an inquiry into legitimate force.³⁸

V. DEFENSIVE FORCE IN THE STATE OF NATURE

As is well known, Kant thinks that only in a “rightful condition”—instituted by the state—can the use of force connected with acquired rights be authorised. The key idea is that the state instantiates a public will that eliminates the problems of asymmetrical constraint associated with private force’s governing access to external things in the state of nature. What is distinctive about a public will is that it is “omnilateral” in virtue of being constituted independently of any private will or aggregation of wills. Force used or authorised by it does not asymmetrically subordinate the agency of some individuals to the agency of others in the way that private or unilateral force does. This is how ownership of external things is possible.

I will say more about this and its implications for the puzzle of enforceability in a moment. But first let me address a worry about where the Kantian argument leads. Surely on a Kantian account (as on any other) we have duties not to kill or assault each other, duties that are naturally enforceable (i.e. independently of the state) precisely because of the importance of the freedom that they are duties to respect? Yet since the successful use of force outwith the state implies (according to the account) the asymmetrical power of one agent to set the boundaries of the freedom of another in a way that is inconsistent with equal freedom, such enforcement seems to be ruled out by that account. So it appears that the Kantian cannot say that these duties are naturally enforceable. Isn’t this a fatal blow?³⁹

Of course the Kantian grants the duties, which Kant himself takes to be correlative with innate rights. The Kantian also takes forceful resistance to their violation to be permissible outwith the state. (Like acquired rights, innate rights authorise the use of force.) But the permissibility of using force to resist violations of duties not to kill or assault here is not explained by evaluation of the independently binding moral duties correlative with the rights any more than it is in the case of acquired rights. The question how a system of forceful constraint is morally possible—and so how the duties correlative with rights could be morally binding—arises here too. The difference is that in the case of innate rights the state is not needed to solve it. Forceful resistance is rightful because and insofar as the use of force to maintain life and bodily integrity in the face of attempted killing or assault prevents the intrusion of one person’s agency into the very embodiment of another’s, given that there are no naturally overlapping physical embodiments of distinct rational agents.⁴⁰ Force that resists physical assault, therefore, does no more than preserve the structure already presupposed by equal freedom.⁴¹

It is true that the force is therefore automatically morally authorised, unlike the force associated with acquired rights, and so respect for the boundaries of external freedom that it maintains is likewise automatically ratified by morality as a duty. But what authorises it is nevertheless the principle permitting constraint of others by one’s external actions (including by force) only insofar as it is compatible with equal external freedom, not an appeal to independently binding moral duties. So, this part of the Kantian account, like the part dealing with acquired right, is force-first. The duty-first approach is not being smuggled in.⁴²

VI. LEGITIMACY CONDITIONS ON FORCE

As I said a moment ago, on Kant’s view the state instantiates a public will that eliminates the problems of asymmetrical power associated with private force’s governing access to external things in the state of nature. Force used or authorised by the state to govern that access does not subordinate the agency of some individuals to the agency of others in the way that private force does.⁴³

The need for a distinct public will that has this implication imposes formal conditions on the constitution of a legitimate state. For our purposes, the most important conditions are those associated with the rule of law. In particular, as in Locke’s account, this requires promulgated, standing law by which official decisions are bound, and to the formulations of which they are accountable in their decision-making (this is one basis for the rule-of-law condition of clarity). It also requires equality before the law and consistency, openness, and impartiality in the law’s administration, among other things.⁴⁴ If these conditions do not hold, then the law does not rule, and state power is the instrument of private individuals.⁴⁵ The state’s use of force is then not compatible with each person’s equal freedom as independence, and resistance is legitimate as it would be to private force in the state of nature.

Kant also stresses the separation of powers, and there are some substantive conditions on the constitution of a legitimate state. Most centrally, it must recognise the equal status and the basic liberties of all citizens, thereby respecting innate rights, which, as we saw, follow from equal freedom in a world of non-overlapping physically embodied rational agency.⁴⁶

In addition, Kant thinks the state ought to act only for the sake of bringing about the equal freedom of each under law, since that is the sole justification for its power in the first place.⁴⁷ It effectively abuses that power if it acts for any other reason. Hence the state may tax citizens to pay for its own defence and for the system of courts, for example, and Kant also argues that it may tax citizens to pay for the basic needs of people whose free agency will be undermined otherwise. But it would exceed its mandate if it were to tax them for the purposes of promoting any particular conception of the good, or use its power to bring about good states of affairs simply on the basis that they are good.⁴⁸

However, by contrast with the rule-of-law conditions and the basic liberties, the restriction to act only for the sake of bringing about equal freedom is not a condition of the state’s legitimacy. Legitimacy is a matter of the constitutional conditions under which power is wielded. The rule of law and respect for the basic liberties are necessary conditions of the required omnilateral constitutional form. Thanks to that form, an exercise of power under the constitution is not subject to an objection from inconsistency with a system of equal freedom, even if it lacks adequate justification. So even a state that violates the additional restriction can be legitimate insofar as resistance to it is effectively an attempt to take power by private agents (however well motivated). Private agents cannot wield force beyond that necessary for defence against threats of assault or killing consistently with the equal freedom of all.⁴⁹

VII. THREE DISTINCTIONS IN ENFORCEABILITY

The duty-first approach to the puzzle of enforceability looks to the nature and value of independently binding duties or forced conformity to them to determine whether they are enforceable. The Kantian force-first approach takes a different path. Whether a duty is enforceable turns out to be a question about legitimate authority—that is, how and by whom force may be wielded—where the answer to that question is not given via the duty-first method. It is given by determining whether the use of force to get people to act in ways that coincide with the duty’s requirements can avoid relations of subordination inconsistent with their equal status. More specifically, given the necessity of the state for the possibility of acquired rights, it is given by determining whether a state could force people to act in ways that coincide with the duties’ requirements, consistent with the constitutional conditions necessary for legitimacy.

This approach generates two key distinctions that frame its answer to the puzzle of enforceability. The first is the distinction between duties that are in principle enforcement-apt and duties that are in principle enforcement-inapt.⁵⁰ Duties that are in principle enforcement-inapt are duties whose requirements no state could legitimately force anyone to conform to consistently with the necessary constitutive conditions. Other duties are enforcement-apt. This distinction falls out of the more general distinction between laws that no state could enforce consistent with the necessary constitutive conditions and laws that a state could enforce consistent with those conditions. As I noted above, legitimate legal requirements can require conduct other than that which is independently morally required, after all.

The second distinction is the distinction between duties that can legitimately be enforced within a given state’s jurisdiction under prevailing law—call them “actually enforceable” duties—and duties that cannot. The former are duties that the state may in actually obtaining circumstances force people to conform to, because it has legitimately authorised enforcement of legal requirements to act in ways that coincide with those required by the duties. The latter are duties that the authority may not in actually obtaining circumstances force people to conform to. This obviously includes duties that are on the “enforcement-inapt” side of the first distinction, but it also includes those duties that are on the “enforcement-apt” side of that distinction but call for acts that are not required by any legal requirement that happens to have been imposed by the public authority as things stand.

There is also a third distinction to be made with regard to each actually enforceable duty. This is the distinction between agents who may legitimately enforce and agents who may not legitimately enforce the duty. From the fact that a duty is actually enforceable it does not follow that anyone may threaten or impose force to secure conformity with it. In all cases apart from rights against assault and killing, the Kantian account implies that only a public will or someone authorised by it may do this.

It is striking that the duty-first approach struggles to find an application for the third distinction. If a duty is enforceable simply because the value of its performance outweighs the loss of autonomy, for instance, that would seem sufficient to justify anyone’s doing the enforcement. Even if further considerations make it impermissible for individuals or at least third parties to engage in enforcement, it will be hard to explain why the duty-bearer should be wronged by it.⁵¹ But on the face of it, unauthorised individual enforcement even of actually enforceable duties can and typically does wrong duty-bearers. “Unauthorised individual enforcement” is, after all, another term for vigilante justice. For instance, it can wrong someone to appropriate an object she has stolen even if one’s intention is to return it to its owner; all the more so to threaten or impose even a milder punishment than her legitimate state would impose.⁵² What is wrongful here is not the frustration of her intentions, but the fact that the enforcement subjects her to a power not authorised by a public will, violating the very structure that gives enforcement its legitimacy in the first place.

VIII. IN-PRINCIPLE ENFORCEMENT-INAPT DUTIES

For the purposes of addressing the puzzle of enforceability, the first distinction is most relevant. The puzzle is to determine which moral duties may legitimately be enforced, and we might think that some duty could (and even should) in principle legitimately be enforced without accepting that a given instance of enforcement, even by a legitimate state, is itself legitimate. So what we want to know is which duties are in principle enforcement-apt, and which not.

The answer given by the Kantian approach is that any duty whose requirements coincide with laws that could be enforced by a state meeting the constitutional conditions for omnilaterality is in principle enforcement-apt. That clearly includes the duties of respect for others’ freedom whose requirements coincide with what agents may legitimately use force to prevent others from doing in the state of nature—duties to refrain from assault and killing. But it may also include many more, since the range of laws that a legitimate state may impose is wide. As we have seen, Kant’s own argument for the state appeals to what he takes to be the moral necessity of rights to external things, which cannot be authorised in the state of nature. But our question is about what the state may legitimately use force to make people do, rather than what it is its proper purpose to make people do. And, as we saw above, the former category is wider than the latter. It is easier to get to grips with the question, therefore, by asking which if any moral duties are in principle enforcement-inapt.

One class of moral duties that will be in principle enforcement-inapt are imperfect duties such as the duty of beneficence or the duty to be considerate. As I interpret them, these duties require duty-bearers to have certain ends or concerns, leaving significant latitude for the duty-bearer’s judgment in choosing which steps to take in pursuit of the concerns given details of the circumstances (including the bearing of other imperfect duties).⁵³ You ought, for instance, to be considerate, but what that calls for in any given context will depend not only on considerations internal to considerateness (maybe helping her out here will actually undermine her in front of her colleagues) but also on considerations internal to other imperfect duties (it is not typically inconsiderate to be late for an appointment if the reason was that you were helping to calm someone suffering a panic attack).

So enforcement of an imperfect duty so understood could be only of performances judged essential in given circumstances to a particular duty-bearer’s living up to the demands of the duty. Now, conformity to such a law wouldn’t be sufficient for compliance with the relevant duties, since it does not imply that one has the necessary end. But insofar as duties not to kill, assault, and steal are (fully) discharged only by refraining from killing and so on out of respect for others as ends in themselves, conformity to laws prohibiting these things wouldn’t be sufficient for compliance with those duties either. Yet I think we want to say that these duties are enforceable, and that the enforcement of such laws counts as enforcing the duties.

The deeper problem with imperfect duties is that since no specific act is necessary in general for the duty to have been discharged, the enforcement of a promulgated, consistently administered law prescribing such an act (or acts) could not count as the enforcement of an imperfect duty (except coincidentally, in the way that a law requiring me not to consume alcohol could count as the enforcement of my promise to my mother not to consume alcohol).

Couldn’t the law simply require that people adopt the end of considerateness? It’s not contrary to the rule of law to apply its requirements to mental states if mens rea tests are consistent with the rule of law. But for such tests to be consistent with the rule of law there must be canons of interpretation, evidence, and application that secure the distinction between enforcement of public law and enforcement of private judgment. Yet if such canons are in place, then it will be possible for a person to fail in the imperfect duty of considerateness and yet comply with the law, or to be justifiably found legally liable and yet not have failed in the duty of considerateness. This is due to the much greater gap, as compared with the case of mens rea, between what could be adduced in accordance with rule-of-law respecting canons of interpretation, evidence, and application and the necessary and sufficient conditions of the inner states involved in adopting the ends that the duty requires.

So we can grant that the enforcement of laws requiring specific performances typically or conventionally regarded as ways of discharging an imperfect duty is compatible with the conditions of legitimacy. To take the example of a different imperfect duty, a law requiring verbal expressions of thanks for benefits received need not in principle be incompatible with such conditions, and this might be thought a way of enforcing the imperfect duty of gratitude. But even perfect enforcement of perfectly formulated such laws, applied perfectly and in accordance with appropriate canons of interpretation and evidentiary standards, would be compatible with almost total failure on the part of subjects to live up to the moral duty of gratitude. Cases of real ingratitude falling outside the law’s formal categories would go unaddressed. Written thanks can be insincere or even sarcastic, and so are compatible with ingratitude. Meanwhile, real gratitude not expressed in the legally specified ways would go legally unrecognised. So here we have a point of real contrast with the duties not to kill or assault others.

The same is true even in the case of imperfect duties not so clearly involving any distinctive attentional or motivational aspect whose expression seems to be fundamental to discharging the duty. Consider the imperfect duty of beneficence. You might think that it would be easy to enforce this duty via something like a tithe. But it seems clear that a person who pays the tithe can fail in her duty of beneficence even so—e.g. by turning her back out of a sense of entitled irritability on a person whom she could easily and at no cost to herself help with some heavy load. Again, no law compatible with the relevant legitimacy conditions could close off every opportunity for culpable non-beneficence, any more than for ingratitude.

Thus in the case of imperfect duties there will be overlap between what the law can require and what the duty requires, but the overlap cannot be complete. And so imperfect duties fall on the in-principle enforcement-inapt side of the distinction.⁵⁴

IX. THE RIGHT TO DO WRONG

Since imperfect duties are in principle enforcement-inapt, they furnish instances of the right to do wrong. If the Kantian approach is sound, there will be cases in which conformity with imperfect duties cannot legitimately be enforced, because such enforcement cannot be compatible with the rule of law, and the rule of law is a condition of legitimacy. Enforcement (by anyone) would therefore violate basic rights of the duty-bearer that are not only compatible with but themselves conditions of legitimacy.

Few will treat this as a compelling defence of the right to do wrong, though. It seems to depend upon ways of failing to discharge a duty that are only of peripheral interest to those concerned with the possibility of such a right. People interested in the right to do wrong are generally concerned with the possibility of act-types or omission-types that are generally wrong but protected by rights against interference. Most would probably accept that there is in effect a right to entertain wrongful thoughts and even form wrongful intentions or adopt bad ends, since force can’t stop people having such thoughts and intentions and ends (and other internal states) anyway.⁵⁵ But that wasn’t what they were interested in. So, insofar as the defence of in-principle unenforceable imperfect duties relies on their being duties to have or not have certain internal states, it will seem not really to target the cases of interest. Or, as some might prefer to put the point: imperfect duties as I’ve interpreted them aren’t really duties at all, but requirements of virtue, and a right to be unvirtuous is not philosophically puzzling.

However, the pattern of reasoning that led to the conclusion that imperfect duties are in-principle enforcement-inapt also applies to certain perfect, directed duties, and the question whether and why non-conformity with these is protected by a right to do wrong certainly is of central interest in debates about that right.

Consider the case of promises. On standard accounts of promising, promises create duties of performance and correlative claims to performance in the promisee. What does the Kantian force-first approach say about the enforceability of these duties?

Laws requiring performance of promises could certainly be legitimately made and enforced. However, as in the case of imperfect duties, consistency with the constraints of the rule of law would require specification of the conditions under which a promise is legally understood to have been given and associated canons of interpretation, evidence, and application in order to secure the distinction between public and private enforcement. And such specifications and canons must inevitably leave room for promissory duties that do not qualify as such from the legal point of view. Perhaps I make you a promise, and we both know it, but I do it in a way that doesn’t meet the conditions necessary for a promise to have been made from the legal point of view. The state cannot legitimately enforce such promises, but there’s no reason to think that they’re not real, that I don’t wrong you by failing to do as I said I would do. Our practices of promising rely on all sorts of social and linguistic cues beyond those codified in contract law.⁵⁶ Both on Kant’s account and intuitively I can fail to treat you as an end in yourself by breaking my legally unrecognised promises to you. The requirements of Kantian Right are not coextensive with those of morality: not all moral wrongs are wrongs of justice.⁵⁷

Consider also cases in which—without actually making any promises—I intentionally and maliciously induce expectations in you that I go on intentionally to disappoint. These too are wrongs that the state cannot legitimately use its coercive power to prevent. As with the case of imperfect duties, it seems that no matter how carefully the law is drawn up, there will be ways to induce expectations in another person so that one gains a moral duty to fulfil the expectations, but not so that the duty has a legitimately enforceable legal counterpart. And so at least some perfect duties will be in-principle unenforceable too—or, to put it more carefully, there must in principle be some perfect duties that are enforcement-inapt.

The same argument can be made in connection with duties of fidelity between lovers. It is a commonplace that such duties are in principle enforcement-inapt. But this seems too quick to me. It is not obvious that the enforcement of laws giving legal recognition to a relationship between lovers and imposing legal duties of fidelity on them, on the model of a contract, must be illegitimate. Marriage might be so understood. What does seem obvious in light of the foregoing argument is that any institutionalisation in law of the relationship between lovers and the duties of fidelity they owe each other must, if it is to be compatible with the constraints of the rule of law, leave room for relationships between lovers and breaches of their duties of fidelity to each other that escape legal recognition. In the simplest case, if marriage is the only form in which the law recognises the relationship between lovers, then unmarried lovers must go unrecognised as such, and so their duties of fidelity will be unenforceable. In each of these cases enforcement of the relevant duty would be illegitimate—and not only enforcement by the state, but enforcement by anyone: it would be a violation of the duty-bearer’s rights.

It’s important to stress here that on the Kantian approach those rights are not merely legal. If they were it could be argued that enforcing the relevant duties does not violate any of the wrongdoer’s moral rights, and if that were true there would be no moral right to do wrong—only a legal one, which would be once again of little interest. But they are moral rights, because morality ratifies legitimately imposed legal distributions of external freedom as the only means of respecting the freedom and equality of each. This is not to say that whatever is legally permitted is morally permitted too. But within the domain of force, as Kant says, ethics takes up legitimate legal duties as moral ones. This is what gives us the structure of moral rights to do wrong: the conduct is morally wrong, but the legal permission to engage in it is morally valid, and so the use of force to prevent it is morally impermissible. Hence there are rights to do wrong.

X. BARRY AND McTERNAN ON THE KANTIAN SOLUTION

In their discussion of the puzzle of enforceability, Barry and McTernan consider the possibility of a Kantian solution of the kind I have been elaborating, but they reject it for the following reason:

[W]e take it that an explanation for which duties we take to be enforcement-apt must permit us sometimes to claim that a legitimate state has incorrectly determined that some duty ought to be enforced, as, for instance, where spousal fidelity is legally enforced. And some duties seem enforceable even in the absence of legal recognition: a state’s failure to legally require that same-sex couples be protected from harassment would not render duties to refrain from such conduct enforcement-inapt.⁵⁸

The Kantian approach can meet these objections. It certainly has the resources to criticise the legitimate imposition of unjust or otherwise objectionable law: that is the point of the distinction between the rule-of-law constraints and basic liberties that are necessary conditions of legitimacy and the additional conditions that rule out paternalistic lawmaking, for instance. Meeting these additional conditions is necessary for a state to be (fully) just, but not for it to be legitimate. Laws validating marriage contracts might be an example of this.⁵⁹

The Kantian approach also makes perfect sense of the idea that some duties are enforceable even in the absence of legal recognition. Evidently it can say that we may permissibly resist violations of duties to refrain from assault and killing, as we’ve seen. And of many other duties it can say that such duties are in principle enforcement-apt insofar as laws requiring acts or omissions coinciding with those required by the duties could be compatible with the conditions of legitimacy that the Kantian approach sets, even if in the absence of such laws they are not what I called “actually enforceable.”

Barry and McTernan might want to say something stronger: that duties to refrain from harassment of same-sex couples—where the harassment in question is less than a straightforward violation of the basic liberties that are a condition of legitimacy on the Kantian approach—are actually enforceable even in the absence of such laws. (And whatever persuades them of that may also persuade them that they’re actually enforceable not only by the officials of a legitimate state but by anyone, even in the presence of such laws.) But this does not seem so intuitively obvious as the claim that these duties are in principle enforcement-apt; at any rate it does not seem obvious enough to render the Kantian approach’s inability to accept it a fatal objection. (To repeat: “unauthorised private enforcement” is another term for vigilante justice.) And indeed the concerns about asymmetries of force that underpin the Kantian view, as described in Sections III–V above, seem to me to provide good grounds to reject it.

XI. CONCLUSION: THE PRIMACY OF POLITICAL PHILOSOPHY

The puzzle of enforceability was to determine which moral duties are enforceable, in the sense that duty-bearers are liable to the use of force to conform to the duties. I have argued that we cannot answer this question without first asking who may legitimately use force, and under what conditions. This turns out to depend on considerations that are excluded by duty-first approaches to the puzzle.

The considerations appealed to by duty-first approaches may of course be relevant to state officials’ deliberations about how to use the power they legitimately wield. For example, the value of unforced conformity to certain duties, or the justification of those duties, might make it in some way morally objectionable to force duty-bearers to conform to them. More general considerations of social benefit may also play a role. But if officials do legitimately impose law requiring conformity anyway, the enforcement inherits that legitimacy, and the duty-bearers are liable to it. And that was what the puzzle concerned.

The core insight of the force-first approach is that the use of force places agents in relationships of morally significant asymmetry that raise special concerns of their own, quite apart from whether the force coincides with what we are independently duty-bound to do or not. Addressing these concerns calls for philosophical resources internal to the analysis of the asymmetrical relationships of force. Because this makes it unlikely that the solution to the puzzle of enforceability could lie in isolated reflection on the importance or content of independently binding moral duties whose enforcement we are inquiring into, duty-first approaches seem destined to fail. Moreover, once we appreciate the insight, we are in a position to give a more satisfying explanation of the right to do wrong.

I have developed the insight in a Kantian framework, in which certain conditions on the wielding of force must be met if the boundaries of freedom structured by it are to be morally authoritative in the first place. But the underlying thought is more general. Republican, realist, and relational egalitarian frameworks can agree that the solution to the puzzle of enforceability must be determined in the first instance by addressing moral concerns about relations of power and force, rather than by reflecting on the content and stringency of independently binding moral duties, or on the value of duty-bearers’ unforced conformity with them.⁶⁰ If this is right, then any satisfying solution to the puzzle must start with political philosophy, rather than end in it.

ACKNOWLEDGMENTS

This article was originally written for the Rights and Wrongings seminar that Sandy Steel and I convened in Oxford in Trinity Term 2025. I’m grateful to Sandy for his commentary on it and to the participants—especially Angelo Ryu, Trenton Sewell, and Hugo Till—for their comments, as well as to colleagues attending Oxford’s Centre for the Study of Social Justice seminar the following term for theirs, and to Luke Davies for his. I am also extremely grateful to the editors of Political Philosophy and to its two anonymous reviewers for critical comments that helped to improve the article greatly, as well as for their notable generosity and patience.

COMPETING INTERESTS

The author declares that he has no competing interests.