II. Plural Grounds of Legitimacy

Pluralists want to separate political obligation from authority. Therefore, although they reject the idea of authority, they do offer an account of legitimate political obligation. Their argument, in short, is that the grounds of legitimacy are plural rather than singular. I put to one side for now consideration of whether, pace pluralism, these may also be grounds of legitimate authority (see Section VI). For now, I focus on the way in which a pluralist approach to legitimacy is at odds with much that is written on this subject. The traditional and dominant approach is to examine whether political obligation is justified by one consideration taken in isolation: either fairness, or gratitude, or consent, or justice, or membership, and so on. To take just one example, at one point, Alan Simmons concludes that “actual consent is the only possible ground of a moral duty to obey the law” (Simmons, Reference Simmons, Wellman and Simmons2005, p. 120 n. 10). The pluralist approach, in contrast, arises from the conviction that there are a variety of reasons to draw on when evaluating the legitimacy of political obligation.

What are the grounds of legitimacy, according to pluralists? Although not an exhaustive list, the following play a prominent role in pluralist arguments: one is a consideration of what is “rational,” and so it appeals to ideas of “self-interest.” A second focuses on the idea of “reciprocity,” and this is evident in arguments about “gratitude” and “fairness.” Another concerns what is “reasonable” to demand from others when considering justice. There is also the requirement to promote the “common good.” And, finally, there is the argument that membership in a community can itself give rise to obligations (Gans, Reference Gans1992, pp. 57ff, 83ff; Klosko, Reference Klosko2004, pp. 812–813; Wolff, Reference Wolff1995, pp. 8–9). Those committed to singularity of ground might accept the diversity of putative grounds of legitimacy. But, if they do, their approach is to examine each one in turn, and determine whether any one of them can, on its own, justify political obligation. The pluralist approach, in contrast, reflects a conviction that the case for political obligation can be strengthened by letting the variety of different justifying considerations work together in a number of different ways (Gans, Reference Gans1992, p. 43; Klosko, Reference Klosko2004, p. 801, Reference Klosko2005, p. 100, Reference Klosko2011, p. 515).

How precisely does this happen? At times, pluralists call on different reasons to justify the obligations characteristic of a number of distinct areas of public life. For instance, in order to justify the obligation to support institutions like the police, the army, and the judiciary, pluralists appeal to mutual self-interest (Wolff, Reference Wolff1995, p. 20); when they want to justify the obligation to provide public goods for universal consumption, such as clean water, they turn to reciprocity, fairness, or the public good (Klosko, Reference Klosko2019, p. 80; Wolff, Reference Wolff1995, p. 20); and to justify the obligation to aid those most in need, different considerations are called for, including distributive justice (Wolff, Reference Wolff1995, pp. 20–21).

At other moments, however, they want to combine different reasons together in order to justify one and the same obligation. For example, in order to justify the obligation to support public funding for opera and for public parks, the idea of the common good is combined with that of fairness to ensure that the benefits and burdens are not unfairly distributed (Klosko, Reference Klosko2004, pp. 812–813; see also Gans, Reference Gans1992, p. 62).

In addition, while it is conceded that some obligations will apply to all members of a polity in the same way, pluralists believe this need not be true in all situations. A political obligation need not be universal over some range of persons; and citizens need not have the same type or level of political obligations (Gans, Reference Gans1992, p. 76; Klosko, Reference Klosko2011, p. 516; Wolff, Reference Wolff1995, pp. 10, 15–16).

By proceeding in this way, the argument goes, it is possible to justify a wide range of political obligations. This stands sharply opposed to an approach restricted to a single ground of obligation, and which asks how that single ground will justify each and every obligation, and the obligations that apply universally to all citizens of all states. To return again to Simmons, at various points in his work, he insists that there is only one possible justification for the duty to obey the law, but also that such a duty must hold for “most (or at least many) citizens” (Simmons, Reference Simmons1979, p. 55). It could be argued that Simmons’ approach is, as a result, ill-suited to the task of making sense of certain types of obligations, such as the obligation to support the provision of public goods. And that is why, for pluralists, one should not strive to identify the one, single ground of legitimacy, for doing so will only lead to the implausible conclusion that either one has very few obligations or that very few are obligated. A better theory results from allowing multiple principles of obligation work in combination in the various ways described above.

In this article, I do not call into question the pluralist argument that both obligations and their justifications are varied. But pluralists also presume that this commitment to pluralism is incompatible with the idea of authority. I return to that point in the final section. Before that, I consider the independent reasons pluralists offer for rejecting authority.

III. The Mode of Operation of an Authoritative Directive

What pluralists object to is the mode of operation of an authoritative directive. So, let us begin with what those who defend authority take to be the way in which it operates as a reason. There are two features to note. The first is that an authoritative directive is a content-independent reason. If one receives an order from someone with the authority to issue that order, one is expected to act as commanded, independent of the quality of the action (Hart, Reference Hart and Raz1982/1990, p. 101). The same is true of the law, if the law has authority. Then, one has a duty to obey the law because it is the law and independent of the quality of the action demanded in the given instance (Simmons, Reference Simmons, Sanders and Narveson1996, p. 24). The second feature of an authoritative directive is that it is an exclusionary reason. It is not only a first-order reason to act in a particular way: a reason of some comparative weight or importance. It also a higher-level (second-order) exclusionary reason: a reason to put to one side (i.e., exclude) the weight of the reasons to act or not to act in this way (Raz, Reference Raz1979/2009, pp. 17–18).

Authority, therefore, involves deference (Hart, Reference Hart and Raz1982/1990, p. 99; Raz, Reference Raz1986, p. 7; Simmons, Reference Simmons2016, p. 13; Wolff, Reference Wolff1970/1998, pp. 17–18). As Raz says, one is free to think what one likes about a legitimate directive (Raz, Reference Raz1979/2009, p. 7, 1986, p. 8). Nonetheless, what a directive demands is that one not act on one's own judgement of the balance of reasons. This very fact is evidence that consensus is not presumed at the outset. Indeed, an authoritative directive is intended as a way to deal with disagreement. An authoritative directive is a reason to act whether or not one agrees with it. However, that also explains why there is serious concern that authority will come at some cost to freedom and rationality, for it demands that one not act as one otherwise would have done, and not as the weight of reasons demands.

What we have here is a distinction between two different ways in which reasons operate. A legitimate directive defeats other reasons by excluding them. It is unlike, say, a moral ought, which, as a first-order reason, defeats other reasons (if it does) by outweighing them. It does not follow that a directive is a conclusive reason, however. For example, if one ought to provide charitable donations to those unable to afford basic goods for themselves, one should act as suggested if the reason for doing so (say, the natural duty to help another) outweighs other reasons, as when one can make the donation with little cost to oneself, and is confident that the donation will be used efficiently and justly. One's reasons for (and against) donating are first-order reasons. In contrast, if one is legally obliged to provide financial support to the most needy (say, through the payment of one's taxes), the intention is that the law will prevail even if it conflicts with other, weightier reasons, including when the current system of taxation and public spending is known to be less than optimally efficient in helping the most needy (Raz, Reference Raz1979/2009, p. 23). As I said, none of the above entails that a directive always defeats any conflicting first-order reason, and indeed, Raz is explicit that at least some first-order reasons can defeat a conflicting directive (Raz, Reference Raz1979/2009, p. 22, Reference Raz2006, p. 1023). However, what pluralists object to is the very idea of the law operating as an authoritative directive in the first place.

Although we will have occasion to return to the arguments of other pluralists, I start with Klosko, as his work is the clearest expression of this pluralist position. And what he says is that, although often there are reasons to obey a law, one should never defer to it. That is, one should not act as the law demands independent of the content of the action and regardless of the reasons for and against acting in this way. One reason to obey a given law, he maintains, is that it corresponds with and satisfies an already existing moral demand. For example, one has an independent moral duty to treat others fairly, and that is why one should obey laws that are themselves fair (Klosko, Reference Klosko2011, pp. 498–499, 504). Similarly, if citizens should pay their taxes, this is not because the law demands it, but because and insofar as the funds collected in this way are used efficiently and for legitimate purposes: for example, projects that will aid the most needy, as above, or those that promote the public good, such as publicly funded parks and museums, and so on (Klosko, Reference Klosko2011, pp. 512–513). Of course, there is a difference between one's acting in a certain way simply because fairness, for example, demands it, and one's obeying a law because the law is fair. One difference concerns the coordination of effort in the latter case, as we will see again below. Nonetheless, Klosko's point is that one cannot be required to obey a law independent of the law's content, here its moral content. As he puts it, “the requirements to obey pass through from CD [content dependent] considerations” (Klosko, Reference Klosko2011, p. 506).

Is it the case that, as Klosko contends, there is no duty to obey a law independent of the weight of reasons to act in the way the law demands? A possible counter-argument can begin from the observation that it is already commonly accepted that laws operate as authoritative directives. For instance, we appreciate that a red traffic light is not a request that one stop one's car at a certain moment (Raz, Reference Raz1979/2009, p. 25). It ceases to function as a law if citizens are free to make up their minds on each occasion whether instead it is better (for example, safer) to run the lights. However, this argument will not convince those who believe authority to be incompatible with rationality. For all the above argument shows (it could be argued) is that authority is claimed by certain agents, and perhaps in addition that there is, or might be, a social purpose served by such claims to authority, including the interests served by a coordination of effort, here concerning road safety (Klosko, Reference Klosko2011, p. 512). As some critics of authority say, such rules can be treated “as if” they were authoritative directives for as long as there is sufficiently strong reason to do so (Alexander, Reference Alexander1990, p. 6; Regan, Reference Regan1989, p. 1009). But none of this entails that it is rational to defer to an authoritative directive, and that the mode of operation of an authoritative directive is rational. Indeed, pluralists maintain that an authoritative directive does not function as a reason should.

Those who claim authority are saying that others should obey the directive they issue because it has been issued by someone (or an institution, or an office, and so on) with the authority to do so. What its critics do not accept is the idea that, although one may have reasons to act as a directive demands, the authority of the directive itself provides a different, independent reason. Rather, if there is a law against theft, for example, one should refrain from stealing not because it is illegal, but because of the quality of the action and the strength of the reasons not to act in this way. It seems peculiar to explain one's refraining from stealing by saying theft is illegal. If an action is morally wrong, its being illegal should add nothing to one's deliberations (Klosko, Reference Klosko2011, p. 506). To claim otherwise is to in effect double count reasons in favour of the authoritative directive when they are weighed in the balance against other, conflicting reasons. And this is objectionable on grounds of rationality. That is, if a law against theft is understood as a content-independent, exclusionary reason, this involves an illicit combination of two reasons not to steal: the fact that theft is morally wrong and the fact that theft is against the law (Essert, Reference Essert2012, p. 66). The point of the pluralist objection is two-fold. It is to show that, on the one hand, the idea of authority is incompatible with rationality, because it involves double counting of reasons; and, on the other, if political obligation has nothing to do with authority, then at the very least it will not be vulnerable to this charge of irrationality. Also, what is not at issue here is the question of whether one has prudential reasons not to break the law, insofar as one has good reasons to believe that one's illegal activity will be detected and punished. This does indeed provide a reason to refrain from theft that is independent of the fact that theft is morally wrong. But whether or not there is such a reason to obey is not what is at issue. The fact that theft is illegal entails that one is legally obliged not to steal, and if one does steal, one can expect to bear the punishment meted out to those who violate that legal duty. That is true by definition. The question, for us, is different. It is whether there is a moral obligation to obey such a law simply because it is the law. And, for Klosko, there is not. The fact that theft is illegal adds nothing to our reasoning (that is, it adds no moral reason), and if we think it does add something, doing so involves an illegitimate double counting of reasons.

There is, however, a two-part response to this pluralist critique, the first part being that it does not adequately account for the way in which a law operates as a reason for action. In fact, there are strong grounds to believe that a law operates as a content-independent and exclusionary reason. As pluralists note, one has independent moral reasons not to drive dangerously, but that by itself is not a reason to obey any specific speed limit, say, the limit of 50 miles per hour. One has a reason to obey this speed limit only insofar as it has legitimacy. That is, once instituted, the law itself becomes a reason for action in its own right, and that is why one is obliged to obey this specific speed limit even though there is something arbitrary about which limit is chosen. Pluralists like Klosko (Reference Klosko2011, p. 506) accept this much, but the implications of doing so are hard to reconcile with the pluralist rejection of content independence and exclusionary reasons. For, it suggests that the weight of one's reasons not to obey the law are one thing, and the obligation to obey the law is another. The pluralist position is “that laws without valid purposes lack moral force,” and hence, for example, in a situation where one drives at five miles per hour over the speed limit, the law does not have “any moral force at all” (Klosko, Reference Klosko2011, p. 521 n. 40). This suggests that the reason to obey a law is nothing more than the weight of the reasons to act in this way (Alexander, Reference Alexander1990, p. 6; Regan, Reference Regan1989, p. 1009; see also Raz, Reference Raz1989, p. 1178). Any given law will have a certain weight or importance, it is true, and breaking the speed limit by a small amount perhaps should be thought to have very little moral significance. However, it does not follow that, in this instance, the law has no moral force. If one knowingly breaks the speed limit — if, say, one habitually drives five miles per hour faster than the law permits in a residential area — one is guilty of some wrong, even if it is not a very grave wrong. We need some way to capture the wrongness of this behaviour. And we can do so if we accept that a law, when it has legitimacy, operates in a way that is different from a first-order reason. It is a reason for action in its own right, in the sense that it is a reason independent of its content and regardless of the weight of reasons not to act in this way.

The second part of the response to this pluralist objection is that it misrepresents the way in which an authoritative directive operates. In particular, the idea of exclusion does not involve a double counting of reasons. It is not the case that an authoritative directive to φ — say, the legal requirement to pay one's taxes — is based solely on reasons to φ, and that one then adds together or combines both the directive itself and the reasons for the directive. For a start, as Raz argues, a directive is based on a balancing of reasons, that is, a balancing of reasons to φ and reasons not to φ (Raz, Reference Raz1986, p. 41). Regarding the requirement to pay one's taxes, considerations of justice and efficiency (among others) might sit on one side, but these are balanced against other considerations, including, among other things, rational self-interest (understood narrowly, say, as the reasons one has to promote one's own material interests). If a tax law does have authority, it represents a judgement of how much restrictions on rational self-interest (say, restrictions for those who will not benefit from a redistributive tax policy) are justified in order to promote an efficient and just system of taxation. In addition, the obligation excludes both reasons to φ and reasons not to φ, and so neither should figure in one's deliberations. Both reasons for and against the directive are reasons that should have been considered, and both are in turn “pre-empted” or excluded by the directive (Raz, Reference Raz1986, pp. 46–47). That is why this particular objection to authority fails. One's reason to obey a tax law, to stick with this example, is not a combination of the fact that it is illegal not to do so and whatever other reasons seem to be on the side of that law (say, that the law is just and efficient) while discounting the reasons that under different circumstances could have led to a different tax law (including concerns around rational self-interest). Therefore, the process of exclusion does not lead to a double counting of reasons.

What I have tried to do, above, is counter the pluralist contention that an authoritative directive is incompatible with rationality. An authoritative directive is a genuine, valid reason, and cannot be summarily dismissed as an illicit double counting of reasons. It is now time to return to an issue raised earlier. A conclusive reason is a reason that defeats all reasons with which it comes into conflict (Raz, Reference Raz2011, p. 109). It may be assumed that if a law has authority (if one is to defer to the law because it is the law), then it is conclusive. But that is not the case. What I have argued so far is that an authoritative directive defeats (some) reasons by excluding them. But I have also noted that something's being an authoritative directive has no implications for its weight as a reason. If reasons operate both by excluding and by outweighing, then it is possible for a directive to be legitimate and yet still be outweighed by other reasons. On the one hand, this shows that proponents of authority do not make what would be the highly implausible claim that a legitimate directive must be the final word in any controversy. On the other hand, it suggests that one can be in a situation where one is morally obliged to obey a law even when there are weighty moral reasons not to obey. As we shall see, moral conflicts of this kind do arise, and authoritative directives can be the focal point of such conflicts. But critics of authority worry that this will have the following perverse implication: if one ought to obey a directive regardless of the weight of reasons not to do so, one could find oneself in a situation where one is obliged to act in ways that are deeply immoral. And then, arguably, a political community becomes analogous with a criminal gang, where membership leads to obligations despite the immorality of the group's aims (Gans, Reference Gans1988, p. 97).

When responding to this critique, the first thing to note is that the requirement to defer only arises when authority has legitimacy. One is not required to defer to those with merely de facto authority, i.e., those with the brute power to issue commands and enforce compliance. In immoral societies, at times one should comply with the laws, say, laws against theft, murder, and rape, but that is because the laws simply require what morality also demands. It is only in certain regimes where deference is justified. Indeed, those who defend authority do so on the ground that only certain regimes have authority, for otherwise one would be required to obey deeply immoral commands (Christiano, Reference Christiano2004, p. 279). Of course, this leaves an important question unanswered: what is it that justifies authority? I return to that question in the next section.

Nonetheless, insofar as moral conflicts are real, and do arise, a worry remains. Even if an authority has legitimacy based on certain moral considerations (say, among other things, its democratic credentials), there are other moral considerations in respect of which it may be open to critique (say, it is overly restrictive of individual freedom). A legitimate authority will then make demands that are, in some sense, immoral, and one is expected to defer. Two things can be said about this argument. The first is that such conflicts are real, and often are deeply troubling, as we shall see. Second, however, it does not follow that the idea of authority can leave one caught in a situation where one ought, all things considered, act in ways that are deeply immoral. What those who defend authority insist on is that obligations are “inconclusive,” as their “finality or imperativeness” is “very much a function of the context within which the […] obligations are exercised” (Simmons, Reference Simmons, Sanders and Narveson1996, p. 23). An authoritative directive is one reason among others, a reason to not act on other reasons, but it might be defeated by a conflicting reason, if the conflicting reason is weightier. And this might happen when authority has legitimacy, and yet makes unjustified demands. It is then possible (although not certain) that the authoritative directive will be defeated. Nothing in the concept of authority precludes that possibility, given that an authoritative directive is not a conclusive reason.

Consideration of civil disobedience can illuminate this point. Cases of civil disobedience might at first be taken to support the pluralist position, insofar as they appear to show that it is up to each individual to decide whether or not to obey a law, and do so guided solely by their own conscience. However, while one is expected to make such a decision in certain cases, it is noteworthy that those scenarios are not the norm. As John Rawls makes clear in his account, for a start, civil disobedience is not the appropriate response to either a completely unjust or a completely just regime. It is appropriate, rather, in a largely just regime, when one is faced with an unjust law, and one has exhausted all available formal routes to remedy the injustice (Rawls, Reference Rawls1971, pp. 364, 366, 373). Rawls also makes clear that civil disobedience is not appropriate in the normal course of events. When a law has legitimacy, one is duty bound to obey even when one is opposed to it, including when one thinks it unjust (Rawls, Reference Rawls1971, p. 351). As already mentioned, authority presupposes dissent. It is assumed that some will disagree with the law that nonetheless should be obeyed. Civil disobedience becomes appropriate when deference no longer is. But that assumes deference is appropriate prior to reaching that point.

Therefore, we can retain the idea of authority and yet avoid the kind of scenario where one is obliged, all things considered, to act in ways that are deeply immoral. However, pluralists advance a further objection. A regime that issues rules and orders that are intended to function as content-independent, exclusionary reasons will be, the argument goes, highly authoritarian. These are regimes, like those seen in medieval feudal times, or Thomas Hobbes’ Leviathan, where laws and commands bind regardless of their content (Klosko, Reference Klosko2011, pp. 504–505). Those who accept authority of this sort become “too obeisant” (Moore, Reference Moore1989, p. 858). Authority, therefore, is incompatible with freedom.

It is true that a legitimate rule demands that one not act on one's own judgement of the balance of reasons. However, there is no necessary incompatibility between rule following of this kind, and at least one conception of freedom (Raz, Reference Raz1989, p. 1178). We find such a conception in Jean-Jacques Rousseau (but also in Immanuel Kant, John Locke, and many others). To be “governed by appetite alone is slavery,” Rousseau maintains, “while obedience to a law one prescribes to oneself is freedom” (Rousseau, Reference Rousseau and Cranston1762/1968, Book I, Chapter 8). On this understanding of freedom, if citizens freely contribute to the creation of their own laws, and craft those laws on the basis of their commitment to moral principles (for Rousseau, it is the general will that must be foremost), they suffer no loss to their autonomy when they obey those laws. What that means is that, once again, authority cannot be rejected summarily, here on the grounds that it is incompatible with freedom. One aspect of freedom, namely rule-based autonomy, is compatible with authority.

Admittedly, this argument is rejected by a priori philosophical anarchists, and they also draw on Rousseau in doing so. For anarchists of this stripe, obligations have legitimacy only if based on the consent of each, making them compatible with autonomy, whereas authority involves enforcing rules on those who disagree (Huemer, Reference Huemer2019, p. 21; Wolff, Reference Wolff1970/1998, pp. xxvii, 6, 12, 51). But pluralists cannot find any succor in this argument. Indeed, pluralists accept that obligations can be imposed on those who disagree, so long as the obligations have legitimacy (Gans, Reference Gans1986, p. 388; Moore, Reference Moore1989, p. 895; Wolff, Reference Wolff1990–1991, p. 154). The argument pluralists are making is that there is a loss to freedom when one is obliged to obey a law regardless of the weight of reasons not to act in this way, but there is no such loss when one is obliged to obey a law that is supported by the weightier reason. In the latter case, pluralists assume, being obliged to obey simply re-affirms the balance of reasons earlier taken to support the legitimacy of that law, as when one is obliged by the principle of fairness to bear one's own fair share in providing public goods (Klosko, Reference Klosko2011, p. 517). But what such an argument fails altogether to address is the reality that for many of those who disagree with the law, the balance of reasons does not support that law. For them, there are weighty reasons not to obey the law. If pluralists accept that laws can be imposed on those who disagree without any loss to their freedom, they cannot object to authoritative directives (as making one “too obeisant”) simply because authoritative directives are imposed on those who disagree.

Nonetheless, this is not the end to the matter. Even if being compelled to follow rules is compatible with one aspect of freedom (i.e., autonomy), nonetheless, it will infringe another, namely negative freedom, the freedom from intentional constraint. And that violation of freedom does matter, and may, on a given occasion, be so serious as to outweigh the duty to obey. I cannot pursue that matter further here (see Fives, Reference Fives2022, Reference Fives2023) except to note the following: for a start, it is noteworthy that pluralists do not raise this particular objection to authority. But those who accept that authority can have legitimacy could also worry about the various ways it violates negative freedom. This line of argument about authority is, therefore, very different from the one pluralists offer. On the one hand, based on what we have seen so far, it can be said that there is no overwhelming reason to reject authority (namely, out of a concern with freedom, morality, or rationality). On the other, in part because of the concern for negative freedom, the prospect that opens up before us is one of ongoing and open-ended moral conflict. And yet, this is to be expected, given the moral problem of authority. It might be said that the pluralist argument, therefore, fails in two different ways. First, pluralists have not shown that authority can be dispensed with. However, second, they also have not shown that the moral problem of authority, and the moral conflicts to which authority gives rise, can be somehow avoided.

IV. How Pluralists Also Employ Content Independence and Exclusion

What I have tried to do so far is offer a response to the pluralist critique of authority. My argument has been that their objections can be answered once we have a clearer understanding of what authority is and how an authoritative directive operates as a reason. Now, I want to consider what it is that pluralists themselves propose as an alternative to authority. What we find is that all of the pluralists considered here at some point utilize the idea of content independence, and that Gans in addition employs exclusion, and while they wish to restrict the scope of these concepts, those efforts, I argue, fail.

For a start, all pluralists accept that a legitimate law operates as a content-independent reason. For Gans, it is not the content of the law that matters, but the fact that the legal system is a valuable social tool that citizens should work to preserve. So long as the legal system is “morally sound,” one is obliged to obey a law, even when this law is, in some other sense, “inadequate” (Gans, Reference Gans1992, pp. 36, 70). Jonathan Wolff agrees with Gans on this point. Given that not everyone will consent to such a duty, Wolff contends that we need to justify the duty to obey the law simply because it is the law (Wolff, Reference Wolff1990–1991, pp. 155–156). Klosko's approach is different, as he rejects the idea that there is an obligation to obey the law simply because it is the law. Nonetheless, he accepts obligations to obey laws that are independent of the content of the particular law in question. For example, there is such an obligation insofar as the law is justified by the appropriate moral principle, fairness, say, but independent of what the specific content of that law turns out to be. Hence, one might think that the law is less than adequately fair and yet one is still obliged to obey (Klosko, Reference Klosko2011, p. 501, Reference Klosko2019, p. 103). Of course, these are two quite distinct notions of content independence: the former argument accepts where the latter rejects that something being a law counts as a reason to obey the law. Despite this difference, it is clear that Klosko, Jonathan Wolff, and Gans do accept that the duty to obey the law operates as a content-independent reason.

Gans goes further again, in effect accepting that a legitimate law operates as an exclusionary reason. He begins from what he sees as the “need to distinguish them [duties to obey the law] from actions which are morally desirable only,” and so characterizes the former as a “limited practical must” (Gans, Reference Gans1992, pp. 19–20). They are “act-types whose performance is a must as long as the reasons against it are of certain types” (Gans, Reference Gans1992, p. 20). He does not use the word “exclusion” here, but that is what the operation of a “limited practical must” amounts to, namely that it defeats other reasons, regardless of the latter's weight. Gans nonetheless holds back from a full-throated endorsement of authority here. While the duty to obey the law does enjoy “absolute precedence” when the only other conflicting reasons are amoral — such as consideration about what is convenient — nonetheless, for Gans, the duty ceases to be a “must” when it conflicts with moral reasons, such as the values of freedom, justice, and equality, and here what matters is “the intensity of the damage” the duty causes to those values (Gans, Reference Gans1992, pp. 20–21).

Of course, as discussed already, the pluralist position is that because of concerns regarding freedom, morality, and rationality, the law does not operate as a content-independent, exclusionary reason. It is, therefore, noteworthy when pluralists, admittedly in different ways and to varying degrees, nonetheless utilize these very ideas. They attempt to restrict the scope or extent of these concepts, for they do not want to end by endorsing authority. But, are these efforts successful? I do not think they are, as can be seen by looking again at Gans’ work.

He maintains that the duty to obey the law is a content-independent reason in a “morally sound” legal system. But this concedes something of great importance, namely that when a regime has legitimacy, its laws operate as content-independent reasons. There is disagreement over what considerations do justify political obligation, as we know, but let us consider the argument Gans advances, according to which a legitimate regime must be (among other things) democratic and just (Gans, Reference Gans1988, pp. 95–96). It would follow that once a polity meets these requirements of legitimacy, the law is (or can be) a content-independent reason under these circumstances. Even though a regime must satisfy certain normative requirements for it to have legitimacy, when it is legitimate, there is no need, and no rationale, to examine the content of each specific law in order to determine whether it is a reason for action. It is a reason for action because it has been issued by a legitimate polity, and it is a reason independent of the content of the action it requires. And, of course, this is at least part of what it means for a regime to have authority.

Gans also accepts that a legitimate law is a reason for action that cannot be defeated by a certain class of reasons, regardless of the latter's weight. What this amounts to is that the duty to obey defeats other reasons by excluding them (and their weight) from consideration. Gans says that the duty to obey is a “limited practical must” only with respect to amoral reasons. In this way, he hopes to retain the idea that what matters when considering obligation is the weight of the reasons to act as the law demands — here, the weight of moral values such as equality, justice, and freedom. However, this attempt to limit the scope of exclusionary reasons is highly questionable. If a moral duty does not also exclude moral values, then there is no real difference between the two when they conflict. Then, the moral duty ceases to be a higher-level reason, and instead operates only as a first-order reason; that is, it defeats the moral value in question only if it is weightier. But Gans’ original aim was different from this. As we saw, he started from the need to distinguish not just between what is convenient and what is morally obligatory, but also between what is morally desirable and what is morally obligatory. To make the latter distinction, the duty to obey the law must also exclude moral values. And if we stick consistently to that point (a consistency missing in Gans’ arguments), we arrive at a position that is perfectly compatible with the idea of authority.

I have tried to show that when pluralists each introduce the ideas of content independence and exclusionary reasons, they fail in their efforts to limit the scope of those concepts, and so stop short of endorsing authority. Of course, this alignment with authority is strongest in Gans’ work. Nonetheless, his work is an important influence on other pluralists (Klosko, Reference Klosko2004, p. 821 n. 4; Wolff, Reference Wolff1995, p. 11). And the direction in which Gans has taken his argument is one that seems to be logically required by what has gone before. Pluralists are defending the duty to obey the law, and as Gans’ work makes clear, such an obligation is conceptually distinct from both consideration of what is convenient as well as what is morally desirable. The idea of an authoritative directive would seem to provide what is needed but missing from the pluralist argument, namely a way to capture what is distinctive about that obligation.

V. Explaining Anarchist Scepticism

A further reason that pluralists give to reject the idea of authority is that it is, they believe, responsible for a surge in anarchist scepticism. Scepticism about the legitimacy of political obligation is explained by the fact that many political theorists assume that if a law is legitimate, it is a content-independent, exclusionary reason (Klosko, Reference Klosko2011, p. 498; see also Wolff, Reference Wolff1995, pp. 18–19).

Some philosophical anarchists do, it is true, conceptualize political obligation in this way, and from that basis they do arrive at a sceptical conclusion, as pluralists predict. Indeed, this is the path taken by Raz, as well as Simmons (Raz, Reference Raz1979/2009, p. 245; Simmons, Reference Simmons1987, p. 275). What they each (separately) argue is that if a law has legitimacy, it is a content-independent, exclusionary reason; but, in political regimes as they currently operate, most citizens are not bound in this way. For example, according to Raz, authority has legitimacy only if obedience to authority better ensures conformity to reason. This might be true of certain laws designed by experts — which better ensure that citizens are protected from certain kinds of harm — but, in most cases, the laws of a political community do not meet this exacting standard (Raz, Reference Raz1985, p. 147).

However, this is not the only route to scepticism. Other anarchists arrive at much the same sceptical conclusion but they do so even while rejecting the very idea of legitimate authority. This is true of a priori philosophical anarchists. Their argument is that one cannot have a duty to obey the law simply because it is the law. Approaching law in this way is not, it is argued, compatible with either freedom or rationality. Instead, both freedom and rationality require that one obey only those laws to which one has consented. Indeed, when one has consented to a law, it operates as a content-independent, exclusionary reason. It is what one can be said to “want,” even when it is no longer what one is “willing” to do (Wolff, Reference Wolff1970/1998, p. 51). There are obligations in an anarchist community but there is no authority because, the anarchist argues, authority obliges those who disagree with a rule. For a priori philosophical anarchists, only those who have consented to the rule are obligated to obey.

What this shows is that some anarchists (i.e., a priori philosophical anarchists) reject authority in much the same way that pluralists do; but, nonetheless, unlike pluralists, they arrive at the sceptical conclusion that, here and now, there are no justified political obligations. Pluralists have, it seems, misinterpreted a priori philosophical anarchism. This is clearly the case with Gans’ reading, according to which Robert Wolff's scepticism about political obligation follows directly from his critique of the idea of authority: namely, that authority entails the “logical contradiction” of “consenting to obey the law's directives while giving up the possibility of acting for moral reasons that may rule against this obedience” (Gans, Reference Gans1992, p. 11). In fact, as we have seen, Robert Wolff is saying that in an anarchist utopia, one can be compelled to do what one is no longer willing to do, and the obligation is legitimate if based on consent, whereas his scepticism arises from the observation that very few have obligated themselves in this way. This is important because pluralists claim to offer both an explanation for, and an antidote to, anarchist scepticism. They have done neither, however, as rejecting the idea of authority, as pluralists suggest, is no guarantee of avoiding anarchist scepticism.

VI. Justifying Obligation

Even if pluralists have not succeeded in explaining anarchist scepticism, they will still insist that it is necessary to jettison the idea of authority. Only by doing so, pluralists argue, will it be possible to identify the plural grounds of obligation. Klosko gives perhaps the clearest account of this argument, according to which the duty to obey the law rests on a wide variety of content-dependent reasons: hence, “careful consideration of the reasons in favor of behaving in accordance with particular laws […] will be largely successful in establishing reasons to comply with all morally defensible laws” (Klosko, Reference Klosko2011, p. 516). As we know, other pluralists differ from Klosko in accepting, where he does not, the idea of a duty to obey the law simply because it is the law. Like Klosko, however, Gans and Jonathan Wolff each make an explicit connection between plural grounds of obligation, on the one hand, and rejection of authority, on the other. For Gans, the anarchist rejection of political obligation rests on what he thinks is a mistaken premise, namely that acknowledgement of a duty to obey the law involves “refraining from actions motivated by considerations which follow from the laws’ contents” (Gans, Reference Gans1992, p. 13). In fact, Gans says, a duty to obey the law that does not require refraining from action on content-dependent considerations can be justified by bringing together plural grounds (including fairness, justice, communal belonging, and consideration of consequences) (Gans, Reference Gans1992, p. 43). Similarly, for Jonathan Wolff, it is possible to identify “the grounds of any individual's obligations in respect of different parts of the law,” and while there may be “only one reason why we should obey the law, […] there are perhaps many different reasons for having the laws we do” (Wolff, Reference Wolff1995, p. 19; see also Wolff, Reference Wolff1990–1991, pp. 155–156).

The idea here, expressed in different ways, admittedly, is that the plural grounds of legitimate obligation can be discerned only if the duty to obey is not a content-independent, exclusionary reason. Is that so? For a start, is it the case that authority is conceptually tied to singularity of ground? Singularity of ground is, as Simmons observes, “the requirement that there be one and only one ground of political obligation.” At one point, he appears to reject this criterion of legitimacy. It is “fairly obvious,” he says, “[t]hat this is a mistake” and that “the consent theorists have been more guilty of this mistake than anyone else.” Hence, “in the absence of special argument” it is, he says, “unwarranted” to assume that there is only one ground of legitimacy, namely consent (Simmons, Reference Simmons1979, p. 35). For some critics, however, Simmons’ objection to singularity of ground is simply a point about methodology. According to this interpretation, offered by Klosko, Simmons is saying that every putative ground of authority should be examined, but this is in order to determine whether any one of them does, as a single ground, justify political obligation (Klosko, Reference Klosko2004, p. 801). Indeed, as already discussed, at various points, Simmons does claim to have identified the single ground of obligation, namely consent. Simmons, as we know, assumes that a legitimate law is a content-independent, exclusionary reason, and, as already noted, he reaches the sceptical conclusion that no one is bound in this way in existing polities. And the pluralist argument is that not only is the singularity of ground approach bound to fail in an attempt to justify political obligation, a plurality of grounds approach can succeed precisely because it will jettison the idea of authority. This would be the case if the position one takes on the grounds of legitimacy has strong implications for the position one must take on the mode of operation of a legitimate law. However, what I hope to show is that this is not so.

The first thing to note is that some anarchists who reject the very idea of authority do so from a commitment to singularity of ground. As we saw, the argument of Robert Wolff is that obligation can only be based on respect for autonomy, and that autonomy is incompatible with authority. Hence, commitment to singularity of ground does not also require acceptance of authority. Second, some of those who defend authority nonetheless embrace plurality of grounds. This might even be the case with Simmons. Indeed, not every pluralist agrees that Simmons embraces singularity of ground. The alternative view, offered by Jonathan Wolff, is that Simmons is rejecting the assumption that if there is some justification for political obligation, it must come from a single ground (Wolff, Reference Wolff1995, p. 13). On this interpretation, Simmons leaves open the possibility that the upshot of our investigations will be to uncover a plurality of grounds, working together, justifying political obligation. Whether or not Simmons accepts plurality of grounds, it is definitely true that Raz does. He identifies at least four separate grounds of authority: expertise, solving coordination problems, remedying prisoner dilemma type situations, and avoiding the negative consequences of disobedience (Raz, Reference Raz1985, pp. 147–148, Reference Raz1986, p. 56, Reference Raz2006, p. 1022).

Therefore, what the literature suggests is that the position one takes on the grounds of legitimacy (for or against plurality of grounds) does not have strong implications for the position one takes on the mode of operation of a legitimate law (for or against authority). And what Raz's work in particular suggests is that authoritative directives may have plural grounds of legitimacy. Pluralists themselves have shown ways in which it might be possible for plural grounds to justify obligations (see Section II). But given that the pluralist critique of authority has not succeeded, there is nothing standing in the way of our interpreting those obligations as content-independent, exclusionary reasons.

This is the case regarding Klosko's argument bringing together a plurality of considerations in order to justify obligations to support common good policies. For a start, he says, the principle of fairness does justify one's compliance with government actions from which one benefits (including those that promote one's own security and health), but does not justify policies designed to benefit others (Klosko, Reference Klosko2004, p. 805). The latter require considerations of natural duty, although they in turn provide a stronger justification when combined with those of fairness. The principle of mutual aid on its own is weak because it is “qualified in regard to cost,” whereas the principle of fairness is not, and so can justify more extensive welfare duties (Klosko, Reference Klosko2004, p. 809). But neither fairness nor natural duties on their own could justify a whole range of other common good state activities, such as those to “keep inflation and unemployment in check,” as well as to support “public education, museums, symphonies, and national parks” (Klosko, Reference Klosko2004, p. 812). The latter can be justified only by bringing in the common good principle, which itself combines the commitment to fairness with a consequentialist concern for promoting the common good of a particular community (Klosko, Reference Klosko2004, pp. 812–813).

There is no compelling case for the argument that the law implementing a common good policy of this sort cannot be an authoritative directive. First, the appeal to a plurality of grounds does not entail or require that any resulting legitimate law must then operate as a first-order, content-dependent reason. Plural grounds can go hand-in-hand with authority, as we have seen. Second, it is not unreasonable to think of the law in question as an authoritative directive. As I have tried to show, authority is not incompatible with freedom or rationality, and does not entail that one ought, all things considered, act in deeply immoral ways. Authority, simply, is one way to deal with the reality of disagreement. Hence, if one is obliged to contribute to the cost of public education, for example, this is the case even when one disagrees with the way in which public education is being provided, say, for pedagogical reasons. One is to put to one side, exclude, the reason one has not to act in this way.

Of course, this may not be the end of the matter, depending on how weighty any conflicting reason is. It may be sufficiently weighty to defeat a directive, and to justify disobedience to an otherwise legitimate law. But this is not evidence that the defence of authority has, somehow, failed. Instead, such a conflict is to be expected. It is an instance of the moral problem of authority. Although authority is not incompatible with freedom or rationality, it does come at some cost to each. Pluralists reject the very idea of authority, and in so doing, it could be argued, they give a falsely consoling picture of political obligation, one that is free from the possibility of a certain kind of moral conflict. My defence of authority in this article is offered as a salutary reminder that this threat of moral conflict cannot be so easily dispelled. It is because authority can have legitimacy that the moral problem of authority arises.