II. Justification and Legitimacy

I shall assume that a government’s legitimacy entails that other agents—citizens and foreign third parties alike—have a moral duty not to overthrow it.Footnote ² I take this to be a thoroughly conventional assumption. Although we shall ultimately have occasion to revise this assumption in certain respects, it provides a good place to start. Ordinarily, the duty that citizens have not to overthrow a legitimate government makes it morally impermissible, all things considered, for them to overthrow theirs. Ordinarily, that is, their duty prevails. There may be exceptions to this generalization; it can certainly be debated how much scope there is for such exceptions. That debate need not detain us, though, since there is more and clearer room for slippage in the other direction and all I need to locate the ambiguity in “justification” is slippage somewhere.

Suppose that a government has lost its legitimacy. As a result, its subjects no longer have a duty not to overthrow it. To put this point precisely, there is a specific duty not to overthrow the government that its subjects now lack, namely, the duty entailed by a government’s legitimacy. However, it does not follow that these subjects are permitted, all things considered, to overthrow their illegitimate government. Various other moral considerations still count against a revolution and those may be decisive. For example, a revolution may be likely to cost a great many lives and, morally, this severe cost may not be worth it. In that case, citizens are forbidden to overthrow their government, all things considered, notwithstanding its illegitimacy. (Below, we shall encounter other examples of additional considerations bearing on the permissibility of revolution.)

We can therefore distinguish two senses in which a revolution—the violent overthrow of a government—may be “justified.” On the one hand, a revolution can be justified in a narrow sense when the citizens lack a moral duty not to overthrow a legitimate government due to their government’s not being legitimate.Footnote ³ For this leaves the moral permissibility of their revolution unopposed by any such duty, that is, unopposed by the very moral factor that ordinarily makes revolution impermissible. A government’s being illegitimate may be sufficient to make a revolution against it justified in this narrow sense. To describe this case, I shall say that citizens are “at liberty” to revolt or that they have “a liberty” to revolt. A liberty to revolt is nothing other than the absence of a duty not to revolt.

On the other hand, a revolution can be justified in the full-blooded sense of being morally permissible, all things considered. Footnote ⁴ To describe this case, I shall say that citizens are “all things considered permitted” to revolt. A government’s being illegitimate cannot suffice to make a revolution against it all things considered permitted, since all things considered permission also depends on a number of additional considerations. In the interest of clarity, let me emphasize that the important contrast here does not run between the locutions “liberty” and “permission,” but rather, between a justification for revolution that takes all applicable moral considerations into account and one that does not, since it only focuses on the presence or absence of the duty entailed by a government’s legitimacy.

I can now restate the central claim of my essay with more precision. I shall argue that a minority of citizens can have a liberty to revolt against an illegitimate government, even when a majority of their fellow citizens does not agree that their government is illegitimate. That is what I shall mean by a liberty to “go it alone.” I do not claim that this minority is always permitted to revolt, all things considered. That depends, as I have said, on various other considerations, which have to be examined separately.

The position I defend is opposed by classical and contemporary writers on revolution alike. On the classical side, I shall take John Locke as my illustrative opponent and, on the contemporary side, Massimo Renzo.Footnote ⁵ In Locke’s theory of revolution, the conditions of illegitimate government are gathered together under the banner of “tyranny.” Locke offers a detailed account of what these conditions are in the final chapter of his Second Treatise of Government. However, he denies that when a government satisfies one of his conditions—and so, qualifies as a tyranny—its citizens are thereby released from their duty not to revolt. Rather, Locke imposes two additional necessary conditions for citizens to acquire a liberty to revolt. First, they must have exhausted all lesser remedies: revolution is a last resort, when there is “no Appeal [left] on Earth.” Second, and crucially, the majority of citizens must agree that the conditions of tyranny are satisfied. Thus, on Locke’s view, as long as the majority in some society judges that the government is not a tyranny, any minority in that society remains duty-bound not to revolt, whatever their own judgment may be.Footnote ⁶

My argument will focus on Locke’s second additional necessary condition,Footnote ⁷ which I reject. I shall conduct the entire discussion under the assumption that there is a right to revolt against tyranny, that is, a liberty to revolt.Footnote ⁸ To isolate what is in dispute, let me stipulate throughout that the relevant majority and minority of citizens agree on how tyranny or illegitimate government is defined, an agreement that represents their common contribution to the conversation among humankind.Footnote ⁹ All they disagree about is whether their government satisfies those agreed conditions. Let me further stipulate that not only does the minority genuinely and sincerely believe that its government satisfies the conditions for tyranny, but the minority is also correct about that.Footnote ¹⁰

Against this background, the issue becomes whether the mere failure of the majority to agree with the minority on this score keeps the minority duty-bound not to revolt, that is, prevents them from acquiring a liberty to revolt. In my view, it does not. At least in a subset of the cases just described, a minority of citizens can have a liberty to “go it alone,” to revolt without the agreement of the majority.

I make no attempt to close the list of conditions under which a minority of citizens can have this liberty. In that sense, my aims are modest. My fundamental contention will be that the moral force of majority rule is subject to certain presuppositions that inevitably limit its scope. Nothing that falls outside of this scope can be justifiably vetoed by the majority and, as we shall see, this includes judgments of tyranny and their consequences. Since I only provide a partial enumeration of the presuppositions of majority rule, the ramifications of a more complete enumeration for the liberty to revolt against tyranny remain open.

The rest of my essay proceeds as follows. In Section III, I further describe the position held by Locke and Renzo. Apart from his pedigree, Locke has the considerable advantage of advancing an explicit argument for his position, which is a useful point of reference. My discussion of Renzo serves, in the first instance, to exhibit the enduring and nonhistorical interest of this position. In addition, it enables us to integrate the terms of our analysis with those of just war theory. Unlike Locke, contemporary writers on the ethics of revolution treat the two questions as continuous.Footnote ¹¹ As we shall see, the wider lens on revolution they employ is closely related to the distinction between a liberty to revolt and all things considered permission to revolt.

In Section IV, I distinguish two versions of majority rule—“conditional” and “unconditional”—and show that Locke’s position requires the unconditional version of majority rule. I then argue, in Section V, that the moral force of majority rule presupposes at least five specific background conditions. Majority rule is therefore conditional rather than unconditional. Whenever one of its presuppositions fails, minorities are immune from having their judgments of tyranny vetoed by the majority and hence can have a liberty of going it alone.

In Section VI, I first reflect on what we learn about legitimacy from the fact that a minority of citizens can have a liberty to go it alone, as distinct from all things considered permission to revolt. Here again our discussion will be enriched by further engagement with Renzo. Finally, I return in Section VII to the theme of global crises, where I locate a more defensible role for a majority veto in relation to the justification of revolution.

III. Locke and Renzo on Revolution

Locke’s position, as I have said, is that any minority of citizens remains duty-bound not to revolt, as long as a majority of their fellow citizens judges that their government is not a tyranny. No minority has a liberty to go it alone. Locke’s rationale for this position, so far as it goes, can be conveniently divided into three propositions. First, membership in political society is voluntary, where consenting to join involves an individual’s surrendering her executive power of the law of nature to the political society. Second, consent to join political society is irrevocable. Third, decisions in political society are made by a simple majority of its members.Footnote ¹²

It follows that, for members of a political society (that is, citizens), any judgment that originally fell within the scope of their power to execute the law of nature was irrevocably surrendered to the majority’s decision on their entrance into that society. From this point, the short route to Locke’s assignment to the majority of the authority to judge whether the conditions for tyranny have been satisfied involves reporting that, within his political theory, judgments about when these conditions are satisfied clearly fall within the scope of the executive power of the law of nature.Footnote ¹³

In a recent article, Massimo Renzo affirms a position functionally equivalent to Locke’s. His affirmation comes in the course of defending a version of the asymmetry view, which holds that citizens in a given state are sometimes permitted to revolt against their own government in cases where foreign third parties (for example, other states) are not permitted to intervene militarily in that same state.Footnote ¹⁴ In Section VI, I shall return to pursue an important implication of this asymmetry. Here, all we need to observe is that Renzo only defends the asymmetry view in cases of what he calls “ordinary oppression.” By definition, ordinary oppression excludes “supreme humanitarian emergencies,” where foreign military intervention is permissible even without the consent of the domestic political community.Footnote ¹⁵ He offers Abdel Fattah el-Sisi’s regime in Egypt following the military coup in 2013 as an example of ordinary oppression.

Renzo goes on to subdivide revolutions under ordinary oppression. In most real-life cases, he remarks, the revolution is actually initiated by a domestic vanguard, with the rest of the population only being mobilized later (if at all). However, he points out that—in principle, though not very commonly—a revolution can also occur when the “entire population decides to come together as one and rise in arms against the regime.”Footnote ¹⁶ Let us call these two categories of revolution “vanguard cases” and “idealized cases,” respectively. Renzo’s defense of the asymmetry view turns out to be further restricted to idealized cases of ordinary oppression.

To begin with, however, we shall be more interested in his vanguard cases. This is not because they are more realistic, but because Renzo argues that the

In imposing this requirement on the domestic vanguard to obtain the consent of the rest of the political community as a condition of being permitted to revolt, Renzo is effectively affirming Locke’s position on the liberty to revolt. He, too, denies that a minority is at liberty to go it alone.

The equivalence I claim to find between Locke and Renzo here depends on a pair of assumptions that may be helpful to spell out. On the one hand, I assume that the revolutionary “vanguard” constitutes a minority of the domestic citizenry. On the other hand, I assume that “obtaining the consent” of the domestic political community means obtaining it from the majority of citizens. Renzo does not engage the question of where the threshold of approval lies. However, that does not matter, as long as he does not mean to require unanimity, as I assume he does not. For simplicity, then, I shall stick with majority consent.

Like other recent philosophers, Renzo treats the ethics of revolution as a subject continuous with just war theory. Just war theory can be divided into two parts, which are traditionally labeled jus ad bellum and jus in bello. The principles in the first part govern the initiation of war, while the principles in the second part govern the conduct of combatants during the war, so we can ignore jus in bello. Furthermore, jus ad bellum can be perspicuously divided, in turn, into a just cause principle and a number of other principles, such as a reasonable prospect of success principle, a proportionality principle, and a last resort principle. However, we shall not be concerned with either the nature or the number of those other principles. To remind us that there is more to jus ad bellum than a just cause principle, we can let the reasonable prospect of success principle serve as our representative illustration of its other principles.

By contrast to recent accounts of revolution, Locke’s theory is basically restricted to the just cause principle.Footnote ¹⁸ His detailed analysis of tyranny can be read as specifying in some depth what counts as a “just cause” in a particular kind of war (namely, an internal war). Nevertheless, precisely because it does not address the question of whether, say, the revolutionaries have a reasonable prospect of success, Locke’s theory is in no position to conclude from any well-established verdict of tyranny that even a majority of citizens has permission to revolt, all things considered. This point is reminiscent of the basis on which we originally distinguished a liberty to revolt from all things considered permission to revolt. Indeed, the other ad bellum principles provide a useful inventory of the principal additional considerations that bear on the permissibility of a revolutionary war. They must also be satisfied before revolutionaries with a just cause can have all things considered permission to revolt.

When the justification of revolution is analyzed through this wider lens, it turns out that not every minority of citizens is on a par. If a minority of citizens can have a liberty to go it alone, as I shall argue it can, then any minority—at the limit, even a single individual—can have one. However, while some minorities of citizens can plausibly satisfy the reasonable prospect of success principle all by themselves, others cannot. For example, to take the easiest case, a vanguard that constitutes a majority of the fighting age citizenry, even though it is only a minority of the voting age citizenry, may have a reasonable prospect of success all by itself.Footnote ¹⁹ By contrast, a scattering of individuals, let alone a solitary one, plainly has none. All things considered, minorities with no such reasonable prospect of success are therefore forbidden to revolt all by themselves, even if they remain at liberty to go it alone. For convenience, then, we can reserve the expression “[revolutionary] vanguard” to designate a minority of citizens that has a reasonable prospect of succeeding to overthrow the government all by itself.

IV. Majority Rule

On the face of it, the pivotal premise in Locke’s rationale for his majority consent requirement is the proposition that consent to join political society is irrevocable. However, this premise suffers from two significant weaknesses. To begin with, Locke offers no defense of it whatever. To that extent, his rationale is obviously incomplete, but I shall not dwell on this point.

Furthermore, the relevance of this premise itself is heavily mediated by Locke’s more fundamental contention that membership in political society is voluntary. It clearly follows from his first two premises that if one is a member of some political society, one’s continued membership in that society is compulsory. Yet outside a voluntarist framework, consent is treated as irrelevant to membership in political society. On some accounts, for example, membership in political society is itself compulsory.Footnote ²⁰ For these nonvoluntarists, continued membership in a given political society might also be compulsory, but if it is, that has nothing to do with consent’s being irrevocable.Footnote ²¹

A more perspicuous formulation of Locke’s rationale should therefore treat the proposition that “continued membership in the political society at issue is compulsory” as a basic premise. This formulation has the significant advantage of transcending his voluntarism. In the interest of efficiency, let us stipulate that any revolutionary vanguard of interest to us remain members of their political society. This covers both the case in which continued membership is compulsory (for either voluntarist or nonvoluntarist reasons) and the case in which it is not compulsory, but the vanguard chooses not to repudiate their membership (although they always could).

I am not granting all of this out of generosity. Rather, I contend that a further, unacknowledged basic premise is still required to secure Locke’s position. If I am right and this further premise is false, my concessions cost me nothing. To see what is missing, we should distinguish two ways of understanding majority rule as a political decision procedure. I shall call them “conditional” majority rule and “unconditional” majority rule, respectively. As I shall go on to show, the rationale for Locke’s position has to be augmented to specify that political decisions are settled by unconditional majority rule.

To grasp the difference between these two ways of understanding majority rule, we should ask whether the moral force of a majority decision presupposes any background conditions. Since the answer presumably varies with the level of moral force ascribed to majority rule, let us begin by setting this force at a minimal level. Let us say that a justification for majority rule needs to provide every member of political society with a good reason to abide by what the majority decides. On one view, majority decisions have this minimal level of moral force inherently, without any background conditions having to be satisfied at all. In that sense, majority rule settles questions “unconditionally.” On the opposing view, majority decisions do presuppose some background conditions; they only provide good reasons to abide by the result when certain conditions are satisfied.Footnote ²² When these conditions are not in place, majority rule has no moral force. In that sense, majority rule settles questions “conditionally.”

This initial difference between the conditions presupposed by majority rule under these respective understandings yields a consequent difference in its scope. Whereas unconditional majority rule applies to any and every decision, there are some decisions to which conditional majority rule simply does not apply. At a minimum, conditional majority rule does not apply to decisions that negate the presuppositions of its own moral force.

To illustrate this idea, suppose that the absence of slavery is one of the background conditions presupposed by the moral force of majority rule. In that case, a majority decision to institute slavery would not have any moral force, since implementing this decision would negate a presupposition of majority rule’s having moral force in the first place. Any decision to institute slavery therefore falls outside the scope of conditional majority rule.

It is an open question how best to interpret this scope effect. Suppose, for example, that some political society permits slavery. I take it that under conditional majority rule, enslaved members of this society have no good reason to abide by anything the majority decides. As applied to an enslaved member, any decision the majority makes negates a presupposition of majority rule’s having moral force. However, it may be debated whether decisions made by the majority on subjects having nothing to do with slavery provide the rest of this society—every free member—with a good reason to abide by the result. A case can certainly be made that such decisions do not themselves negate the relevant presupposition of majority rule’s having moral force. At the same time, it cannot be denied that the presupposed condition—absence of slavery—fails to obtain in this society. I shall not try to evaluate how these other kinds of decision should be treated, since the basic idea behind conditional majority rule is clearly workable without a resolution of this point.

Suppose that the revolutionary vanguard is composed of enslaved members of this political society. Under conditional majority rule as we have illustrated it, nothing the free majority decides has any moral force in relation to an enslaved member. A fortiori, any decision by the majority that the government is not tyrannical has no such moral force either. It is therefore difficult to see how the vanguard’s liberty to revolt could depend upon the majority’s concurrence in their judgment of tyranny or, equivalently, how the majority’s refusal to concur could serve to maintain the vanguard’s duty not to revolt. To avoid this particular result, Locke has either to embrace unconditional majority rule or at least to deny that absence of slavery is one of the presuppositions of majority rule’s having moral force. However, to preserve his position in its full generality, Locke has specifically to embrace unconditional majority rule.

V. Why Some Conditions on Majority Rule Are Necessary

We have been examining the thesis that any minority group’s liberty to revolt against tyranny depends upon some decision by the majority in their political society. On this thesis, a revolutionary vanguard’s liberty to revolt depends, for example, on the majority’s concurrence in the vanguard’s judgment of tyranny (Locke’s view), on the majority’s consent to overthrow the regime (Renzo’s view), or perhaps on some other decision of theirs. While I have been calling this thesis “Locke’s position,” Locke is not the only one whose theory requires unconditional majority rule to be correct. Anyone who affirms the thesis does, for the counterexamples to it that ensue when the moral force of majority rule is merely conditional do not depend for their effectiveness on how one arrives at the companion premise that continued membership in this political society is compulsory. In particular, they do not depend on political voluntarism. Indeed, as we have seen, they go through even when continued membership on the vanguard’s part is merely stipulated. I shall now argue that conditional majority rule is the correct view and, on that basis, I affirm a minority’s liberty to go it alone.

Various strategies can be employed to establish that majority rule is conditional. Some of them are familiar from arguments about the merits of judicial review of ordinary domestic legislation. I shall begin with an ultra-minimalist version of a proceduralist strategy,Footnote ²³ which will serve as the thin end of my wedge. As we shall see, it will not be necessary to drive the wedge very far. How far this wedge can plausibly be driven is a separate question, one I am happy to leave open.

Before the majority of the people can decide anything, certain procedures have to be in place defining who can vote, how votes are to be cast and counted, how questions or candidacies are to be presented, whether there will be prior discussion and under what conditions, and so on. A certain minimum apparatus is thus presupposed by the very existence of “the people” as an actual decision-making body, and so too by the ability of any majority of them to arrive at a decision. It is simply incoherent for the majority to decide, for example, not to count all of the validly cast votes on a given question, while nevertheless representing the outcome as the “decision of the majority.” Hence, there is at least one background condition presupposed by majority rule, namely, counting all of the valid votes.

What about universal suffrage? Does this likewise belong to the very constitution of majority rule as a decision-making procedure? Despite the obvious moral attractions of universal suffrage, it cannot plausibly be claimed that the operation of majority rule without universal suffrage is incoherent. To give but one familiar example, male-only majority rule is all too coherent as a decision-making procedure. However, it is undeniable that male-only majority rule does not give women a good reason to abide by what the majority of men have decided. In other words, it is not difficult to see that the decisions it yields have no moral force in relation to women. This makes universal suffrage another background condition presupposed by majority rule, even though establishing this conclusion requires us to go somewhat beyond the most austere proceduralism.

Moreover, unlike the previous presupposed condition, universal suffrage licenses counterexamples to Locke’s position on precisely the model employed in the discussion above about the “absence of slavery.” In a society where women are disenfranchised, a revolutionary vanguard could be composed entirely of women. It may be easy to imagine a majority of men in this society objecting that their common government is not tyrannical, but this objection could not serve to deprive a vanguard so composed of its liberty to revolt, since no decision by this majority has moral force in relation to women. In that case, the vanguard’s liberty to revolt holds independently of the majority. While we are on this point, I should add that “absence of slavery” is not a random illustration. It also belongs among the background conditions genuinely presupposed by majority rule.Footnote ²⁴

Let us see whether our wedge can be driven one increment further. To this end, consider freedom of speech, freedom of the press, and freedom of association as background conditions for collective decision-making.Footnote ²⁵ The significance of these freedoms being presupposed by majority rule rests, in part, on the fact that the counterexamples to Locke’s position they consequently license work somewhat differently from those licensed by either universal suffrage or the absence of slavery. In the first instance, those two earlier conditions concern participation in the collective decision-making governed by majority rule. Where certain members are excluded (or worse), decisions made by the majority are drained of moral force in their application to the excluded members.

By contrast, the freedoms of speech, press, and association describe some possible conditions under which collective decisions may be reached, regardless of who participates in that process. If these freedoms are genuinely presupposed by the moral force of majority rule, then where they fail to obtain, any and every decision the majority makes in their absence is drained of moral force, no matter to whom the decision is applied. Among other things, this means that the composition of the revolutionary vanguard is then irrelevant to the conclusion that the vanguard’s liberty to revolt holds independently of the majority. Where the majority’s dissent from the verdict that the government is tyrannical is reached in the absence of the freedoms of speech, press, and association, a vanguard remains at liberty to revolt, no matter which members of this political society belong to that vanguard.

Some philosophers unequivocally maintain that the freedoms of speech, press, and association are strictly necessary conditions of majority rule’s having any moral force. Most famously, perhaps, Ronald Dworkin asserts that all three freedoms must be in place before majority rule carries “any automatic moral advantage over other procedures for collective decision.”Footnote ²⁶ Indeed, in Dworkin’s view, our wedge actually has to be driven much further than this, insofar as he insists on a considerable number of additional conditions as equally necessary for majority rule to have any moral force. For example, he places freedom of conscience and religion, as well as the right to marry (along with other rights associated with sexual orientation) and possibly an even wider right to privacy, on a par with the first three freedoms as presuppositions of majority rule.Footnote ²⁷

However, let us leave Dworkin’s more ambitious claims to one side. We do not need to settle on an exhaustive or definitive list of the conditions that are presupposed by the moral force of majority rule. That issue can remain open. It will be enough to decide whether the freedoms of speech, press, and association should be added to the background conditions we have already accepted as minimal presuppositions of majority rule. I take it that John Hart Ely agrees with Dworkin that these three freedoms are necessary conditions, even if that way of putting it gets the order of influence backwards.Footnote ²⁸ While the proposition on which they agree is certainly attractive, I am not convinced that denying it puts one beyond the moral pale either.

Nevertheless, there is an important sense in which we do not have to settle that issue. I have already suggested that whether majority rule presupposes this or that background condition will vary with the level of moral force ascribed to majority rule. So far, we have only contemplated it having the bare minimum of moral force, enough to generate “a good reason” for members of the political society to abide by what the majority decides. We do not have to decide whether majority rule presupposes the freedoms of speech, press, and association at this minimal level of moral force—which is what Dworkin and Ely contend—because Locke’s position will be vindicated only if majority rule has considerably more moral force than this bare minimum.

More specifically, Locke’s position requires majority rule to have enough moral force to defeat the reason the vanguard has to resist tyranny (or to annul this reason in some other way). Of course, different accounts can be given of what constitutes true “tyranny.” That is to say, different accounts can be given of what the conditions are under which members of political society are no longer bound, as they ordinarily are bound, by a duty not to overthrow the government. For present purposes, it does not matter what the correct analysis of tyranny is; readers are free to fill in the details of their own preferred analysis. All I need to assume is that some such analysis is correct. In other words, I am only committed to the assumption that there is a liberty to revolt, at least under some conditions. This much is plainly common ground in the debate in which I am intervening.

As we have stipulated in our cases of interest, the government is in fact a tyranny (according to the preferred analysis of tyranny) and a revolutionary vanguard affirms the same.Footnote ²⁹ In that case, the vanguard has a weighty reason to overthrow the government. Now add that the majority in this political society disagrees with the vanguard, denying that their common government is a tyranny. On Locke’s position, the majority’s dissent here has the consequence that the vanguard remains duty-bound not to overthrow the government, and so not at liberty to act on its weighty reason. For this consequence to be morally justified, the moral force of majority rule would have to be strong enough to overcome the vanguard’s weighty reason to revolt.

Granted, the question of whether the government is a tyranny will typically be in dispute and not a matter that anyone is in a position to stipulate. Still, the stipulation usefully serves to highlight what is at stake for a revolutionary vanguard, certainly in cases where the vanguard’s position in the dispute is correct. Thus, unless majority rule has considerable moral force, invoking it to block a vanguard’s liberty to revolt against tyranny runs a clear risk of a morally unjustified effect. While it may not be necessary to eliminate this risk altogether, even to attenuate it adequately requires majority rule to have a good deal of moral force. However, having thereby raised the level of moral force ascribed to majority rule appreciably higher than the bare minimum, I cannot see how to avoid the conclusion that majority rule does presuppose the freedoms of speech, press, and association as background conditions.

Two distinct considerations support this conclusion. Both of them are anchored in the observation that, in practice anyhow, freedom of the press, freedom of speech, and freedom of association are effectively necessary conditions of the majority’s decisions being adequately informed decisions. The first consideration turns on the proposition that unless the majority’s decisions are adequately informed, majority rule cannot have the appreciable level of moral force that Locke’s position requires it to have. I take it that this proposition is not only secure, but that it is evidently so. Together, these two premises entail that the three freedoms belong to the background conditions presupposed by (the relevant operations of) majority rule.

Unlike the first consideration, the second holds independently of the level of moral force ascribed to majority rule. It turns, instead, on the idea that part of the case for majority rule as a collective decision-making procedure is epistemic.Footnote ³⁰ On the simplest version of this idea, the position favored by the majority in disputes settled by majority rule is more likely to be correct; and that is a good reason for settling them that way. Here, the main point is that any epistemic case for majority rule must itself presuppose that the majority’s decisions are adequately informed. Otherwise, the epistemic benefits on offer will not ensue. However, as we have already observed, the three freedoms under discussion are effectively necessary conditions of the majority’s decisions being adequately informed. Hence, insofar as its justification rests partly on an epistemic basis, majority rule inevitably presupposes the same three freedoms as background conditions.

VI. Legitimacy Revisited

Let us take stock. I have argued that, correctly understood, majority rule presupposes the satisfaction of certain background conditions. Some of these are generic procedural conditions. Others have more substance and are presupposed by majority rule’s having the bare minimum of moral force. My examples were universal suffrage and the absence of slavery. Yet further conditions are presuppositions of majority rule’s having sufficient moral force to block a vanguard’s liberty to revolt against tyranny. This third category includes freedom of speech, freedom of the press, and freedom of association. (Nothing much is changed if someone believes that one or more examples I have assigned to the third category actually belong in the second category.)

On this basis, we can affirm that any minority of citizens can have a liberty to go it alone. We thereby reject Locke’s position on who may revolt, according to which a domestic vanguard’s liberty to revolt against tyranny always depends on some decision by the majority of their fellow citizens (for example, on their concurrence that the government is a tyranny). This is demonstrably false, since whenever one of the background conditions presupposed by majority rule fails to obtain, no domestic majority can justifiably block the vanguard’s liberty to revolt.

More specifically, when universal suffrage is absent or slavery is present, a vanguard drawn from the excluded members of political society is fully at liberty to revolt against tyranny, whether or not the majority concurs that their common government is tyrannical. Similarly, when freedom of speech, freedom of the press, or freedom of association substantially fails to obtain, any domestic vanguard—no matter who belongs to it—is fully at liberty to revolt against tyranny, whatever the majority’s position in the matter may be. In some of these cases, the majority’s dissent has no moral force at all, whereas in others it lacks sufficient moral force to justify blocking the vanguard’s liberty. However, in all of them, the vanguard’s liberty to revolt holds independently of the majority, contrary to Locke’s position.

In arguing that various specific background conditions are presupposed by majority rule, I left it open whether additional conditions are required (as Dworkin maintains). This means that the range of limitations on the domestic majority’s ability justifiably to block a vanguard’s liberty to revolt may be even more extensive than has already been established. Nevertheless, I do not mean to suggest that the vanguard’s liberty to revolt is wholly independent of the domestic majority. For all I have said, domestic majorities may sometimes be justified in blocking a vanguard’s liberty to revolt, such as when all the presuppositions of majority rule are satisfied.

The preceding summary alternates between “any minority of citizens” and “a domestic vanguard” as its description of who can have a liberty to revolt against tyranny, even when the majority dissents. From Section IV onward, I reserved the expression “vanguard” to designate a minority of citizens that has a reasonable prospect of success in defeating the government militarily all by itself. The point of restricting attention to revolutionary vanguards, so conceived, is to make it plausible that other things can be equal as far as the rest of jus ad bellum is concerned—that is, in relation to its principles besides a “just cause.” When other things are equal, a liberty to revolt is equivalent to all things considered permission to revolt. Hence, in principle, a “vanguard” can be all things considered permitted to revolt all by itself.

Unlike our regimented vanguards, mere scatterings of individuals have no reasonable prospect of overthrowing the government all by themselves. It follows, as we have said, that they are forbidden to revolt, all things considered. Moreover, this is a fact that will always be known in advance of further analysis. One might therefore wonder what point is served by insisting, as I do, that scattered individuals can also have a “liberty” to go it alone. In the context of a moral theory of revolution, there is admittedly none. For there, the focus is squarely on the all things considered permissibility of attempts to overthrow a regime. However, things are different in the context of a theory of legitimacy.

To make sense of the difference, we should return to the asymmetry view defended by Renzo and Michael Walzer. Recall that, according to this view, revolution under ordinary oppression is sometimes permissible, but outside military intervention in the same cases is never permissible—at least not on the assumption that the intervening state has not been invited (or otherwise authorized) to intervene by the citizens in the target state.Footnote ³¹ In the ordinary cases contemplated by the asymmetry view, the domestic citizenry’s permission to revolt entails that it also has a liberty to revolt, that is, that it no longer has the particular duty not to revolt entailed by the government’s legitimacy.

On reflection, it should be clear that accepting the asymmetry view requires us to complicate the simple assumption about legitimacy with which we began. A government’s “legitimacy,” so our simple assumption had it, entails that other agents—citizens and foreign third parties alike—have a moral duty not to overthrow it. Yet even if each of these groups has “a” moral duty not to overthrow a given regime, the duties they respectively bear cannot be the same duty (that is, they cannot be tokens of the same duty-type). For in certain cases—namely, the very cases where the asymmetry holds—the duty citizens have not to overthrow their own regime lapses, while the duty foreign third parties have not to overthrow that regime remains firmly in place.

We thus have two duties not to overthrow one and the same regime, and the types of duty these tokens instantiate differ at least insofar as their conditions of application differ or, equivalently, the conditions under which the two duties lapse differ. It therefore stands to reason that the bases on which these two duties are entailed must differ, too, as between citizens and foreign third parties. If “legitimacy” is our label for the moral basis on which a duty not to overthrow a regime so designated is entailed, then legitimacy turns out not to be one thing, but at least two different things. Notably, legitimacy in relation to citizens is one thing and legitimacy in relation to foreign third parties (for example, other states) is something else. I should like to suggest that this is a rather important lesson about political legitimacy.Footnote ³²

In contrast to Walzer, Renzo only ever offered to defend the asymmetry view in a very narrow range of cases. Specifically, as we saw, he restricts its application to idealized cases of ordinary oppression, that is, to cases in which the entire population rises up as one against the regime. Thus, while we have paid much more attention to vanguard cases, Renzo explicitly excludes them from the scope of his asymmetry view. Instead, he treats vanguard cases as symmetrical to foreign third-party interventions.

Against this background, the most obvious consequence of my argument that domestic vanguards can have a liberty to go it alone is that the range of asymmetry cases becomes significantly wider relative to what Renzo allows. Under conditions of ordinary oppression, a domestic vanguard can have a liberty to revolt against tyranny, even when a majority of their fellow citizens does not agree that the government is a tyranny. Renzo and Locke deny this, but we have seen why that is a mistake. Nevertheless, it remains the case that no foreign third party can have a liberty to intervene in these same cases. A fortiori, none can have a liberty to intervene to overthrow the government. The domestic majority’s dissent remains sufficient to block a foreign third party’s liberty to overthrow the regime, despite not always being sufficient to block a domestic vanguard’s liberty to do the same. This is precisely the asymmetry that gives the asymmetry view its name and shape. It follows that the lesson about legitimacy we extracted from reflecting on the asymmetry view is anchored in a much wider and more realistic range of cases than Renzo recognizes.

In the first instance, then, the fact that scattered individuals can also have a liberty to go it alone means that the range of asymmetry cases has to be widened yet again—not by a lot more, perhaps, but clearly some more. More importantly, however, attending to cases where the minority of citizens seeking to revolt all by themselves falls below the threshold for having a reasonable prospect of success forces one to frame the evaluation of attempted revolutions in terms of a liberty to revolt rather than an all things considered permission to revolt. One is forced (on pain of only ever getting the same negative answer) to frame the evaluation in these terms because, below that threshold, all things considered permission to revolt is unattainable.

The advantage of evaluating attempted revolutions specifically in terms of whether the revolutionaries are “at liberty” to overthrow their government is that one is then naturally led to articulate the asymmetry between citizens and foreign third parties in parallel terms. In the context of the theory of legitimacy, this is a real advantage because it is precisely the liberty to overthrow the government—as distinct from either the all things considered or even the “other things being equal” permission to overthrow it—that is the contradictory of a duty not to overthrow the government. As we have seen, there is an important lesson about legitimacy to be learned from observing the asymmetry between the duties citizens and foreign third parties respectively have not to overthrow a given government. Moreover, the cutting edge of this revealing asymmetry is to be found exactly in the differential conditions under which these two duties lapse. It is therefore perfectly expressed in the language of a “liberty” to overthrow the government.

VII. Revolution and Global Crises

By juxtaposing revolutions and third-party interventions, the asymmetry on which we have just been concentrating doubles as a reminder of the variety and complexity of the pathways over which revolutions can lead to global crises. The opening discussion focused on pathways between a revolution in one place and revolutions in other places, distinguishing broadly between inspirational and causal pathways. However, a pathway from revolution to global crisis does not have to run through other revolutions. Along causal pathways, there is at least as much danger, if not a greater danger, that a global crisis will be precipitated by third-party interventions in some domestic revolution. This is particularly true when different states intervene on opposite sides of the same revolution, with the danger being further exacerbated when the intervening states belong to rival military alliances and all the more so when superpowers are involved. Something close to this dynamic was an omnipresent concern during many wars of colonial liberation in Africa (for example, in Guinea-Bissau), taking place as they did during the height of the Cold War.

The prospect of third-party intervention in a domestic revolution also brings to the surface a latent ambiguity in the expression “going it alone,” as this figures in my titular notion of a liberty to go it alone. On the one hand, “alone” can refer to the revolutionary vanguard’s dispensing with the moral support, as we might call it, of the domestic majority’s agreement that their common government is illegitimate. On the other hand, it can refer to the vanguard’s dispensing with military support from any other quarter. For the purpose of my central line of argument, which prosecutes the dispute with Locke, “going it alone” means the former. That is, a minority of citizens has the liberty to go it alone when they are released from their duty not to overthrow the government, despite the fact that a majority of their fellow citizens does not agree that this government is a tyranny.

At times, however, I have also been at pains to emphasize that some minorities of citizens have a reasonable prospect of succeeding at overthrowing the government all by themselves, reserving the expression “vanguard” to designate such minorities. Revolutionary vanguards so regimented have a reasonable prospect of success in “going it alone” in the latter sense. As we have seen, a reasonable prospect of military success is necessary for any minority of citizens to have all things considered permission to revolt, although it plays no role in their having a liberty to revolt.

Yet even in relation to all things considered permission to revolt, a reasonable prospect of success does not have to be secured without military support from any other quarter. This is where the possibility of third-party intervention becomes relevant to the justification of an attempted revolution, quite apart from its causal contribution to the risk of precipitating a global crisis. This is because a domestic minority that lacks a reasonable prospect of success when seeking to overthrow the government “all by itself” may nevertheless acquire one insofar as it can rely on effective military assistance from a foreign third party.

The catch is that, in order to contribute to a domestic minority’s all things considered permission to revolt, the foreign third party’s intervention on their behalf not only has to be effective, but also has to be itself permissible. According to the asymmetry view, though, the domestic majority always has the power to make a foreign third party’s intervention all things considered impermissible in cases of ordinary oppression. That is because a foreign party’s ordinary duty not to intervene will remain in force in the absence of the domestic majority’s authorization to intervene, where this duty not to intervene also suffices, under conditions of ordinary oppression, to make intervention all things considered impermissible. Hence, in the special subset of these cases where a domestic minority’s reasonable prospect of success in overthrowing the government depends on a foreign third party’s intervention, this minority’s all things considered permission to revolt is subject to being vetoed by the domestic majority.