The challenge posed to liberal political theory by those who can usefully be described as political realists is not new; those who fall into this category, including John Dunn, Bernard Williams, Raymond Geuss, Bonnie Honig, Glen Newey, and Chantal Mouffe, to name a few, have been developing realist themes and critiques for over a decade.Footnote ¹ But liberal political theorists have only recently begun to recognize it as the serious challenge to liberal theory that it is. Three factors, I think, explain this recent surge of interest. Firstly, Williams's posthumously published In the Beginning Was the Deed represented the first time that a figure who had previously operated within, and made significant contributions to, liberal theory argued for the need to develop a form of political realism in response to its perceived deficiencies.Footnote ² Secondly, review papers by William Galston and Marc Stears have done an excellent job of bringing together and organizing the various arguments made by the realists, or those advocating what Stears calls a “politics of compulsion,” and, in doing so, have clarified the various strands of realism and the challenges that it poses for liberal theory.Footnote ³ Finally, there is a growing suspicion that liberal theory is ill placed to address the political realities of radical insecurity and conflict in which we now live and that a form of political realism, which puts greater emphasis on the disagreements and discord of politics, might fare better in this regard.

The general aim of this article is to make a first attempt at responding to the realist challenge to liberal theory in the hope that this will encourage similar endeavors.Footnote ⁴ Its more specific aim is to explore the possibility of developing a form of liberal political theorizing that responds to the realist challenge while remaining distinctively liberal; hence, liberal realism. With these general and specific aims in mind the article necessarily paints in broad brushstrokes, drawing on a diverse range of realists and liberals alike, such as Judith N. Shklar, Richard Rorty, John Gray, and John Rawls, in addition to those mentioned above, and moving quickly over areas in which there remains much more to be said. The article begins by setting out in more detail the exact nature of the realist challenge to liberal legitimacy. It then proceeds to map out the general features of three different liberal responses to that challenge, indicating what I take to be their strengths and weaknesses vis-à-vis the realist critique and drawing attention to those avenues that might warrant further development.

The Realist Challenge

A certain amount of construction is required to enable us to talk of “realists,” a “realist tradition,” or the “realist challenge,” given that those theorists who are being referred to when such terms are used do not understand themselves to be part of a unified theoretical movement, let alone one called realism.Footnote ⁵ Nevertheless, and as both Galston and Stears have demonstrated, there are several “family resemblances,” common concerns and arguments, that can be identified which allow us to speak meaningfully of there being a realist countermovement to liberal theory. These include an appreciation of the distinctiveness and autonomy of the political from other spheres of human life (in particular morality) and a questioning of the liberal account of agents' motivations and of the relative absence of interest within liberal theory of the role that power, hierarchy, leadership, judgment, loyalty, passions, contingency, and domination play in politics. However, like Stears, I believe that the most significant challenge that realism poses for liberal theory is its criticism of the prevailing liberal account of legitimacy. This challenge derives from the fact that realism presents a different view of the political than liberalism does, one that makes political conflict and disagreement a necessary and ineradicable feature of politics itself, and in doing so excludes the possibility of achieving a consensual basis for the use of political power, which is a necessary feature of liberal legitimacy.

Liberal political theory seeks to do the seemingly impossible: it attempts to show how people who disagree, fundamentally and seemingly intractably, about moral and religious issues can nevertheless live together under a common political authority. However, this authority must be one that is consistent with and respects the freedom of all individuals subject to it, a demand that liberals have usually taken to mean that it must be one that enjoys the endorsement or consent of each and every citizen. Jeremy Waldron has gone so far as to say that what distinguishes liberalism from all other political creeds is the idea that “a social and political order is illegitimate unless it is rooted in the consent of all those who have to live under it; the consent or agreement of these people is a condition of its being morally permissible to enforce that order upon them.”Footnote ⁶ This connection between universal consent and legitimacy can be found replicated in several other prominent liberal theorists. John Rawls wrote that “a legitimate regime is such that its political and social institutions are justifiable to all citizens—to each and every one—by addressing their reason, theoretical and practical. Again: a justification of the institutions of the social world must be, in principle, available to everyone, and so justifiable to all who live under them. The legitimacy of a liberal regime depends on such a justification.”Footnote ⁷ And Thomas Nagel likewise claims that “the task of discovering the conditions of legitimacy is traditionally conceived as that of finding a way to justify a political system to everyone who is required to live under it … the search for legitimacy is a search for unanimity.”Footnote ⁸ Identifying political principles that all persons could reasonably be expected to accept therefore becomes a, if not the, central task of liberal political theory. If such commonly endorsed (or endorsable) principles can be found, then they would enable the state to employ its power in ways consistent with the freedom of all persons, and therefore legitimately. The hope is that despite individuals' moral and religious disagreements and conflicts they can nevertheless converge on principles of social cooperation; indeed, the legitimacy of the liberal state depends on it. So the liberal aim is consensus and agreement at the level of politics in conditions of disharmony and disagreement at the level of morality.

The realist view of the political is very different. Whereas liberalism is a vision of politics grounded in agreement on shared political principles, realists are keen to emphasize disagreement and conflict about the political as a feature of politics itself. Some realists, especially those, such as Mouffe, influenced by the work of Carl Schmitt, go as far as to make conflict an essential characteristic of the political.Footnote ⁹ As Williams put it, “the idea of the political is to an important degree focused on the idea of political disagreement … political difference is the essence of politics.”Footnote ¹⁰ The sort of political disagreement that the realists think liberals have neglected is that which is internal to or the very subject matter of politics itself. Glen Newey, for example, has argued that politics “is characterized by endemic disagreement over issues which are by common consent a matter of public concern” but also by “endemic disagreement over what counts as a political question … what ought to be a matter of public concern” in the first place, the latter of which contemporary political theory has, he believes, largely ignored.Footnote ¹¹ Beyond this, other fundamental questions of politics, such as the purposes to which government power should be put, how the benefits and burdens of social cooperation should be distributed, the goods and values on which claims to legitimacy should be based, and the appropriate institutional design for political society, are all further issues that realists urge us to recognize as inevitably and maybe essentially contested matters. So, instead of seeing the political as an arena of consensus and agreement about fundamental principles, for realists, “politics takes place in the face of inevitable disagreement, and … is best understood as the functional response to that disagreement.”Footnote ¹²

The challenge that the realist emphasis on political disagreement and conflict, what is often called “deep pluralism,” poses strikes at the heart of the liberal conceptualization of politics as a human activity. Liberalism, according to realists, understands politics to be an activity that takes place against the backdrop of consensus about the fundamentals of political life. Even those liberals who consider political agreement to ideally be the result of actual deliberation between all persons regulated by the demands of public reason, and so might deny this realist charge, nevertheless assume that a consensus about the necessity to deliberate exists prior to the start of the dialogue. In other words it assumes that there must be significant agreement before the political conversation and political activity can begin. Consensus is the precondition of the political, a condition that must be present for there to be such a thing as politics in the first place. This means that agreement about political fundamentals cannot, on the liberal view, be an objective that politics hopes one day to achieve via rational dialogue but must actually exist here and now (unless liberals want to deny that there is, properly speaking, currently such a thing as politics).Footnote ¹³ Hence, rather that viewing fundamental questions of politics (rightly) as political questions, to be addressed through the processes and procedures of politics itself, realists accuse liberals of seeing them (wrongly) as philosophical matters that are to be settled in some sense prior to politics. Liberals assume, Bonnie Honig states, “that the task of political theory is to resolve institutional questions, to get politics right, over, and done with, to free modern subjects and their sets of arrangements [from] political conflict and instability.”Footnote ¹⁴ Or as John Gray has interpreted the dominant mode of Rawlsian liberalism, “all important questions about liberty and distribution are decided prepolitically, by theoretical reasonings whose results are entrenched in constitutional law, and they are not subsequently politically alterable.”Footnote ¹⁵ So by seeking to provide philosophical justifications for political institutions, practices, or principles of justice that all persons could or should accept, liberal political theorists are trying to settle political disputes once and for all and, in doing so, miss the fact that conflict and disagreement about those very things, the most basic and fundamental issues of politics, are a precondition of the political itself.

By locating conflict and disagreement at the heart of the political, several realists have been led to claim that liberalism, with its emphasis on political consensus and accord, is not a theory of politics at all; rather, by trying to settle fundamental political matters via philosophical argument, liberals are effectively attempting to evade or escape from politics, describing a world that is “after” politics, in which politics is “displaced” or “abolished.”Footnote ¹⁶ On the face of it, it might seem strange to claim that liberal political theory does not take conflict seriously enough; after all, the moral and religious disagreement that characterizes modern democratic societies is the very problem that liberal theory is intended to address. But what is lacking in such accounts, realists think, is any recognition that conflict about politics is a feature of politics itself, or that political conflict is an inevitable and ineradicable characteristic of political life. Some realists, such as Waldron, believe that the inevitability of conflict makes universal consent a theoretical chimera, an unattainable and unfeasible objective that masks the fact that disagreement about (for instance) which principles of justice should regulate the basic structure of society is a permanent and inevitable feature of the political world.Footnote ¹⁷ Disagreements about political fundamentals are not going to disappear, nor should we expect them to. Any theory that proceeds on the assumption that these conflicts can be resolved is therefore a utopian and misleading guide about the prospects for political change and inevitably leads to disappointment and disillusionment. For other realists, the aim of consensus is achievable, but wherever it exists it is contingent and fleeting, the product not the precondition of politics, and often attained via the sort of hegemonic moves and methods of exclusion and consent manufacturing that violate basic liberal moral principles.Footnote ¹⁸ On either account, liberalism seeks to solve the problem of politics in modernity by doing away with politics altogether. Realists, by contrast, believe that “there is no substitute for politics—if by politics we mean the various ways in which we arrive at collective, authoritative decisions in a world in which people legitimately hold different views about the purposes of government and the manner in which it should be carried on.”Footnote ¹⁹

It might be objected that this account of liberalism pays insufficient attention to the already present practical and political aspects of liberal theory, especially in its “political” form as developed by Rawls, Charles Larmore, and Bruce Ackerman, among others.Footnote ²⁰ Central to political liberalism is the recognition that the vast majority of people living in liberal democratic societies are, as a matter of fact, committed to liberal principles, though it is also the case that there is, and maybe will always inevitably be, a minority of individuals who reject liberalism. In such conditions, unanimous agreement on liberal political principles would be an inappropriate objective of liberal political theory. Rawls's aim, for example, was not to achieve an overlapping consensus of all persons on a conception of justice, but to achieve an overlapping consensus of all those who hold reasonable comprehensive doctrines. Far from aiming at a unanimous consensus, contemporary liberal theory recognizes the fact that modern democratic societies contain persons who reject liberalism (though it does not take this to undermine its account of legitimacy) and seeks agreement about constitutional essentials only among the reasonable. If this is right, then it would seem to weaken significantly the realists' challenge insofar as unanimity is not, despite their claims to the contrary, the ambition of liberal politics or theory.

It is important to note that the nub of the realist challenge is not the empirical question whether there is actually deeper divergence on liberal principles than liberals tend to assume (though this may or may not be true). It is not on this empirical point that realists rest their argument, and so they could gladly accept, and indeed welcome, the fact that a very high percentage of citizens of modern societies endorse liberalism. What is central to the realist challenge to liberal legitimacy is realism's vision of the political, which takes conflict, including political conflict, to be an essential and ineliminable feature of politics. In seeking agreement among only reasonable persons on constitutional essentials, political liberalism does still seek unanimity, albeit among a smaller justificatory constituency. It is still, the realist would contend, a theory of politics without conflict. There is no political dispute between reasonable persons as they all agree on the constitutional essentials of the state. So while realists might welcome the recognition that modern democratic societies do contain people who reject liberalism, the political liberal response of excluding such persons from the justificatory constituency in order to maintain the possibility of unanimity is simply to relocate the site of agreement in order to maintain the consensus-based vision of the political. There are, of course, what Honig calls “remainders” in the form of the unreasonable persons who are excluded from the justificatory constituency, and though there is a crucial political question about the basis and form of the relationship between them and the wider society, in particular about how the conflict between the two is to be managed, it is not one that contemporary liberals have spent too much time worrying about.Footnote ²¹ Despite its seeming abandonment of the aspiration of unanimity, political liberalism actually replicates the liberal consensus-based vision of the political through a redescription and reduction of the justificatory constituency relevant to the issue of legitimacy.Footnote ²²

The realist vision of the political represents a powerful challenge to liberalism. Where liberalism envisages politics as an activity between people who share common political principles, realism sees politics as generated by and responsive to the fact of deep and inevitable disagreement about fundamental political questions. As we saw earlier, the liberal requirement of political consensus is a product of the moral concern that political power be used in ways that are consistent with the freedom and equality of those over whom it is employed. This is why the liberal principle of legitimacy is so central to liberal political theory. Yet it also explains why the most forceful aspect of the realist challenge to liberalism is its criticism of liberal legitimacy. If it is indeed the case that the liberal principle of legitimacy is flawed in the way that the realists imply, this would have considerable ramifications for the prospects of liberal theory, and possibly liberal practice. Stears has even gone so far as to suggest that “the notion that individuals should lead their own lives and be subject only to laws that they can be said, in some way, to have made themselves, has always been the basis of the liberal tradition and will continue to be so.”Footnote ²³ Accordingly he is led to believe that liberals cannot abandon their account of legitimacy without ceasing to be liberals at all.

This sets the scene for the challenge that liberal realism takes up: Is it possible to reconceptualize liberalism's account of legitimacy such that it recognizes the realist insight into the essentially conflictual nature of the political while still retaining features of the theory, and the politics that it justifies, that are recognizably and distinctively liberal? This will be quite a task, for there is a significant risk of responding to the realist challenge in a way that results in throwing the liberal baby out with the legitimacy bathwater. That is to say, there is a real danger that by altering the account of legitimacy so as to address the realist challenge we abandon those aspects of liberal political theory and indeed political practice that make it distinct and desirable in the first place. If we discard the liberal account of legitimacy we may undercut the justifications for the liberal practices and institutions we want to save. The important issue for liberal realism is, therefore, whether the liberal understanding of legitimacy can be maintained in light of the realist critique or whether we can reconceptualize it in a way that addresses the challenge but still generates distinctively liberal politics.

In the sections that follow I want to set out, in the broadest of brushstrokes, three different strategies for approaching this problem. These different forms of liberal realism, which I shall call negative, minimalist, and partisan, draw inspirationFootnote ²⁴ from three different strands of the liberal tradition: the liberalism of fear, modus vivendi, and (though this may sound strange now) political liberalism, and I shall deal with them in that order.

Negative Liberal Realism

One general strategy for developing a theory of liberal realism would be to retain the requirement of universal consent as a necessary condition of legitimacy while moving the subject matter of the consensus from principles of justice or institutional structures (which will, realists contend, always be a matter of dispute and disagreement) to less controversial but nevertheless sufficient normative principles. This is the strategy underlying a negative theory of liberal realism. Rather than seeking principles of justice that all persons can reasonably be expected to accept, negative liberal realism attempts to uncover a consensus about a set of common human evils that can then function as the shared normative principles necessary for liberal legitimacy. The “negativity” of this approach derives from the fact that the consensus is about the evils that liberal politics are designed to avoid rather than the goods that they are intended to achieve, the hope being that a negative moral consensus on common human evils will not be subject to the sort of conflict and disagreement that realists take principles of justice to be subject to. It is this prospect that makes a negative moral consensus a possible basis for salvaging liberal legitimacy and hence an adequate response to the realist challenge.

The preeminent model for this negative approach to liberalism can be found in the liberalism of fear espoused by Judith N. Shklar and developed in different ways by Richard Rorty and Bernard Williams. “Liberalism,” Shklar wrote, “has only one overriding aim: to secure the political conditions that are necessary for the exercise of personal freedom.”Footnote ²⁵ For Shklar, a necessary condition for freedom is the absence of fear. The liberalism of fear, Shklar wrote, “does not … offer a summum bonum toward which all political agents should strive, but it certainly does begin with a summum malum, which all of us know and would avoid if only we could. That evil is cruelty and the fear it inspires, and the very fear of fear itself. To that extent the liberalism of fear makes a universal and especially a cosmopolitan claim.”Footnote ²⁶ This emphasis on cruelty and the fear that it causes is echoed by Rorty, who defined liberals as “people who think that cruelty is the worst thing we do,”Footnote ²⁷ and by Williams, who, defending the universalism of the liberalism of fear, stated that it “speaks to humanity. And it has a right to do this, a unique right … because its materials are the only certainly universal materials of politics: power, powerlessness, fear, cruelty, a universalism of negative capacities.”Footnote ²⁸ But if fear, and the fear of fear, are to provide a universal foundation for liberalism, we need to be clear about both what we should fear and who we should fear.

The first question requires us to give more content to the cruelty that causes the fear that politics ought to prevent. Shklar defined cruelty as the “deliberate infliction of physical and secondarily emotional pain upon a weaker person or group by stronger ones in order to achieve some end, tangible or intangible, of the latter.”Footnote ²⁹ The fear to be avoided is that “which is created by arbitrary, unexpected, unnecessary and unlicensed acts of force and by habitual and pervasive acts of cruelty and torture.”Footnote ³⁰ She went on to say that “of fear it can be said without qualification that it is universal as it is psychological. … It is simply a first principle, an act of moral intuition based on ample observation, on which liberalism can be built, especially at present. Because the fear of systematic cruelty is so universal, moral claims based on its prohibition have an immediate appeal and can gain recognition without much argument.”Footnote ³¹ If this is right, and the fear of cruelty is a universal human moral evil, then it might provide the sort of normative consensus that liberal legitimacy requires.

The cruelty that the liberalism of fear seeks to prevent is “overwhelmingly generated by governments. … And while the sources of social oppression are indeed numerous, none has the deadly effect of those who, as the agents of the modern state, have unique resources of physical might and persuasion at their disposal.”Footnote ³² It is a feature of the historical awareness that the liberalism of fear seeks to foster that the appropriate object of human fear is given by our actual lived experience and sense of the past, and for Shklar and Williams the modern state, with the huge inequality of power that exists between those whose hands are on the levers of political power and the ordinary citizen, has proved itself time and time again to be the main perpetrator of such cruelty, especially when its overwhelming force is directed toward realizing ideological or religious ends. Interestingly, the liberalism of fear is often interpreted as a form of political theory similar to Hobbes's insofar as fear plays a central role in generating the need for politics as well as giving particular form to that politics (though, of course, for Hobbes the fear that the inequality of a sovereign's power instills in his subjects was the solution to the problem rather than a problem that politics needs to solve). It also makes the Hobbesian move of responding to the problem of reaching political agreement in conditions of moral pluralism by moving to supposedly firmer psychological universals of human motivation. However, fear of the state's vastly unequal power and the concern to protect the individual in such circumstances is a recognizable theme of liberal thought that moves it much closer to Locke's political theory than Hobbes's. But while the liberalism of fear can adopt a strong defense of human rights and their legal protection, these rights can be understood not as somehow prepolitical, fundamental, and given, as they are in Locke's thought, but as “licenses and empowerments that citizens must have in order to preserve their freedom and to protect themselves against abuse.”Footnote ³³

Put all these features of Shklar's theory together and what emerges is a relatively minimal idea of the liberal state (at least in relation to many contemporary political theories concerned with distributive justice or multiculturalism), one in which rights are constitutionally protected and legally enforced via well-defined procedures and opportunities for counsel and appeal. In this case, there are multiple centers of power to act as checks and balances on government activity, and a vibrant democracy ensures equality of power as much as possible. However, this has proved to be the source of one of the most persistent criticisms of the liberalism of fear—essentially, that it generates too thin a set of institutions and practices for it to truly count as a theory of liberal politics.Footnote ³⁴ If the liberalism of fear is to provide a basis for liberal realism, a theory that responds to the realist challenge yet is still distinctively liberal, this is something that needs to be addressed.

Rorty developed Shklar's thinking in an interesting direction, one that opens up potential avenues that might help us address this issue. Though Shklar's essay “The Liberalism of Fear” focuses largely on the potential physical cruelty that government can inflict on its people, much of her other work focused more on the cruelty that individuals inflict on one another via more “ordinary vices” such as hypocrisy and snobbery.Footnote ³⁵ Rorty developed this strand of Shklar's work, combining it with some neopragmatist insights, to create a liberalism of fear that both expanded the content of the cruelty to be avoided to include humiliation and shifted the focus to the cruelty that individuals rather than governments can cause to others via the hegemony of particular vocabularies, something that the liberal ironist, with their never-ending desire to engage in redescription, seeks to avoid.Footnote ³⁶ This is important for our purposes because the sort of politics that can be generated by a concern for avoiding humiliation is potentially thicker and more substantial than that generated solely by the primary concern about avoiding physical cruelty. For Rorty, a liberal society is one populated by liberal ironists, people who accept the contingency of their own beliefs and desires, including their hope that the suffering and humiliation of others might cease, and seek via their imaginative ability to increase human solidarity. But you could equally get to many different political positions from an initial desire to avoid humiliation, including a complex account of identity politics (“My culture is humiliated by not being properly recognized in public political and social spheres”) or very egalitarian distributive principles of justice (“My status as an equal citizen is undermined by vast inequalities in wealth”). In short, the politics of a liberalism of fear can be cashed out differently depending on what is taken to count as the cruelty, physical or psychological, that it is the goal of liberal politics to avoid.

Unlike Shklar, Rorty sought to keep the content of cruelty and humiliation indeterminate; they act merely as place holders for the things that we want to avoid doing to others which can change as we become sensitive to and aware of forms of cruelty and humiliation we were hitherto ignorant of or stubbornly unwilling to recognize. Rorty would also have been disinclined to agree with Shklar that the wish to avoid fear and humiliation, or more significantly, the wish to avoid causing it in others, can be taken as a universal given. Our sensitivity, Rorty thought, to the suffering of others is dependent on our viewing them as beings of equal worth. This is not—and here maybe Rorty had the benefit (if you can call it that) of speaking after the fall of the USSR and the nationalist atrocities that followed—something that we can take for granted but an aspiration we should work toward via sentimental education.Footnote ³⁷ This hope that we can become more sensitive to the pain and humiliation of others and in doing so set our sights on something grander than simply the avoidance of physical cruelty gives Rorty's work an explicitly utopian character as opposed to Shklar's version of the liberalism of fear which she described as “entirely nonutopian … a party of memory rather than a party of hope.”Footnote ³⁸

Clearly, if it is true that the desire to avoid cruelty, fear, and humiliation is not a common human trait, then it weakens the potential for the liberalism of fear, or a version of it, to provide the basis for a response to the realist challenge. Nevertheless, Rorty's work is still potentially valuable for our purposes insofar as it shows us how, by expanding the content of the cruelty and fear that liberalism should protect us from, we can develop a vision of liberal politics that is not necessarily relatively minimal. This might allow the liberalism of fear to respond to the common criticism that it is too minimal a politics, possibly too minimal to even count as a version of liberal politics at all. Another potentially fruitful avenue might be to follow Bernard Williams's lead in thinking that though the removal of the conditions in which one group of people can terrorize another should count as the “first political question,” this does not mean that it is the only political question, that all politics needs to do is secure order, protection, safety, trust, and the conditions of cooperation.Footnote ³⁹ The conditions of legitimacy are always likely to be much more substantial than that depending on our historical and social location, but answering the first political question is always a necessary though not sufficient condition of legitimacy—that is to say, a solution to the first political question is required all the time; “it is not a matter of arriving at a solution to the first question at the level of state-of-nature theory and then going on to the rest of the agenda.”Footnote ⁴⁰

The trouble with these responses, however, is that they seem to undermine that feature of the liberalism of fear which made it attractive as a basis for responding to the realist challenge in the first place. As soon as you start requiring necessary conditions of legitimacy other than the avoidance of common human evils, you are likely to hit upon the sort of disagreements about the conditions (moral or not) of legitimate politics that the realist critique requires us to take seriously.Footnote ⁴¹ The advantage of the liberalism of fear as a basis of liberal realism was that in focusing on the common psychological propensity to avoid human evils we have principles that all can affirm and that can therefore provide the universal consensus necessary for liberal legitimacy. If that becomes only one necessary condition of legitimacy, then the fact of political disagreement, in particular conflict surrounding the appropriate grounds of legitimacy, threatens to disrupt the liberal account once again.

And, likewise, maybe by extending the content of the cruelty that we are to avoid to include humiliation, though this might generate a thicker politics, we create another opportunity for disagreement that undermines the potential for the liberalism of fear to create a universally affirmed basis from which we can work. It might well be, then, that if we do draw on the liberalism of fear for our account of liberal realism, that we must reconcile ourselves to the fact that only a very minimal set of political institutions and goals are legitimate, those dedicated to the avoidance and reduction of physical cruelty.

Finally, the potential of negative liberal realism to respond to the realist critique lies in its ability to ground liberalism in a shared desire to avoid common human evils, as well as a consensus about what these evils are. We are entitled to ask, even if we focus only on physical cruelty rather than emotional or psychological suffering, whether there is any more agreement on this than there is on the question of which moral goods humans should pursue or which principles of distribution are just. There is something intuitively right about the thought that there are simply things that all humans want to avoid, such as torture and arbitrary incarceration. But, again, we should remember that history teaches us two sides of the same story: that some people have sought to avoid causing cruelty and suffering while others have actively pursued it, possibly even seeing it as a morally worthy pursuit. Rorty was right to remind us that recent history is littered with stories of people who fully recognized, if not reveled in, the abhorrent and terrifying cruelty enacted on those not considered moral equals.

Minimal Liberal Realism

The strategy underlying the minimal approach to liberal realism is to create a consensus on political institutions that are justified not by their ability to promote or avoid particular moral goods or evils, but by their ability to reconcile, though not perennially eradicate, conflict, and in doing so provide a means for the peaceful coexistence of people with different and conflicting moral, political, and religious views. The hope is that this desire to achieve peace, stability, and order is a minimal commitment (either prudential or normative) that can be the subject of the sort of universal agreement necessary for liberal legitimacy. Politics, and in particular liberal politics, is on this view justified insofar as it provides a strategy for reconciling the sort of moral, religious, and political conflicts that characterize modern societies. So whereas the negative approach to liberal realism grounds legitimacy in the common human propensity to avoid universal human evils, the minimal view locates the relevant agreement in the aspiration to achieve the preconditions of political life, order, security, and peace, by living according to principles all can accept.Footnote ⁴² Though there are several theories that might provide resources to develop this minimal form of liberal realism, including Stuart Hampshire's procedural account of justice,Footnote ⁴³ Chantal Mouffe's theory of liberal agonism,Footnote ⁴⁴ and Richard Bellamy's politics as compromise,Footnote ⁴⁵ the concept of a modus vivendi offers a more familiar (though not necessarily for the right reasons) access point for many liberals. The work of John Gray, the contemporary political theorist who has done most to rehabilitate and defend the notion of a modus vivendi against its liberal detractors, is therefore an appropriate focus for our attention in this section.Footnote ⁴⁶

John Gray writes that “liberalism has always had two faces. From the one side, toleration is the pursuit of an ideal form of life. From the other, it is the search for peace among different ways of life. In the former view, liberal institutions are seen as applications of universal principles. In the latter, they are a means to peaceful coexistence. In the first, liberalism is a prescription for a universal regime. In the second, it is a project of coexistence that can be pursued in many regimes.”Footnote ⁴⁷ For Gray, only the second project, which he calls the pursuit of a modus vivendi, is viable in modernity. A theory of modus vivendi, according to Gray, searches not for an ideal regime but rather (and not “merely”) terms on which different ways of life can coexist peacefully in the same society.Footnote ⁴⁸ It accepts the familiar point that modern societies contain a plurality of different forms of life but instead of aiming to eliminate conflicts between them, which Gray believes the first face of liberalism aims at by showing how we can reach a rational consensus about universal principles of politics (such as principles of justice), modus vivendi seeks agreement or compromise on “common institutions in which many forms of life can coexist.”Footnote ⁴⁹ This makes it a much less ambitious form of political theory insofar as it is intended to be compatible with many different normative principles of political organization while at the same time not prescribing any particular form of political association.

Gray places the theory of modus vivendi in contrast to that of the liberal project of rational consensus, which seeks to show how universal principles of morality can generate substantial conclusions with regard to the universal form of legitimate political regimes, as exemplified in the liberalism of John Rawls and Ronald Dworkin.Footnote ⁵⁰ The problem with this approach, Gray believes, is that it overlooks the truth of value pluralism.Footnote ⁵¹ Drawing heavily on the work of Isaiah Berlin, though deviating from it in several important ways, Gray has argued on several occasions for “the proposition that there are many kinds of good life, some of which cannot be compared in worth. Some varieties of the good life are neither better nor worse than each other, nor the same in value, but incommensurable; they are differently valuable.”Footnote ⁵² The claim of value incommensurability, that some human goods cannot be compared or definitively ranked, is the most important here. The consequence of this truth is that there is no one right way of living a moral life for there are a variety of ways in which these goods can be ranked and combined, and there is no rational way of assessing which moral life is more worthwhile or valuable than any other. And, in some instances, values might be mutually exclusive in the sense that the achievement or pursuit of one logically or empirically precludes that of another, with no rational way of judging which good should be pursued and which abandoned. A way of life, or indeed any particular society, cannot combine or realize all moral values.

Unlike Berlin, who believed that value pluralism justified a liberal politics that puts high premium on the freedom of people to choose their ends, for Gray the upshot of value pluralism for political philosophy is that the aim of politics must be a modus vivendi that allows the many different and conflicting forms of human life to exist side by side. That Gray was led to a different conclusion from Berlin regarding the ramifications of value pluralism is explained by the fact that he recognizes both that value pluralism about the good will necessarily generate pluralism about the right also, and that value pluralism can lead persons to endorse normative frameworks that are not generally conducive to liberalism. Accordingly, Gray thought that the political problem value pluralism gives rise to is how we can live together with those whose normative frameworks lead them to recommend illiberal models of political association, ones that might, for example, seek to prohibit persons from pursuing particular ways of life or give the state a more paternalistic role in the shaping and fostering of beliefs and practices than liberals would generally tolerate. The aim of political philosophy cannot be to abolish such political differences and conflicts of value because pluralism, moral and political, is a fact of human life. Rather, the objective must be “to reconcile individuals and ways of life honouring conflicting values to a life in common.”Footnote ⁵³ Or, as David McCabe puts it, modus vivendi “grounds political life not via its consonance with citizens' deepest moral ideals but instead as something diverse citizens can agree to as an acceptable compromise.”Footnote ⁵⁴

But this compromise could take a variety of forms and a liberal regime is but one possible form of modus vivendi. Other, less liberal alternatives might also provide the necessary reconciliation of values equally well or possibly better, depending on the circumstances.Footnote ⁵⁵ For Gray, “the test of legitimacy for any regime is its success in mediating conflicts of values.”Footnote ⁵⁶ However, he is adamant that modus vivendi should not be mistaken for a theory that says that “anything goes.”Footnote ⁵⁷ There are limits to what counts as a modus vivendi, which are determined by a universal minimum of human goods and evils that are necessary to have and avoid if any kind of human flourishing is to be possible.Footnote ⁵⁸ Gray lists courage, prudence, and sympathy for the suffering and happiness of others as generically and hence universal human values that are necessary for proper human well-being.Footnote ⁵⁹ Equally, there are universal human evils such as genocide and enslavement which if pursued by a state would rule out any reasonable claim to its legitimacy.Footnote ⁶⁰ So only a political regime that reconciled conflicts between values within the limits of providing the space for the pursuit of universal human goods and avoiding common human evils can be legitimate, though Gray is clear that many political regimes, not only liberal ones, will meet the test of the universal minimum.

Gray's theory of modus vivendi seeks a consensus of all individuals on institutions whose purpose is to achieve reconciliation or compromise among conflicting values and ways of life. The normative foundation of his theory is minimal in the sense that it requires only that individuals, divided by their substantive moral, religious, and political disagreements, all desire peaceful coexistence and see their common political life as a way of achieving this rather than as realizing or promoting grander moral values. However, what can potentially make his theory, and others like it, a rich source of inspiration for liberal realism is the fact that it aims to provide a normative foundation for political regimes that seek not to overcome conflicts once and for all, including political conflicts, but a way of reaching temporary reconciliations between people who hold different moral, religious, and political views. Accepting that political conflicts are perennial is to concede one of the realist's central claims regarding the nature of politics. But equally, in attempting to provide a universally acceptable framework for reconciling these conflicts, it could potentially provide a way of responding to the realist challenge to liberal legitimacy.

In assessing the viability of developing a form of liberal realism along these lines, the first thing to note is that justifying modus vivendi on the basis of the truth of value pluralism has the potential to alienate those who hold “monistic” conceptions of the good that reject several, if not all, of the central tenets of Gray's meta-ethical theory. Value pluralism, Gray believes, commits us to saying that all universal religions, which claim that there is but one single right way of living a moral life, are false insofar as they take conflicts of value to be indicative of human error.Footnote ⁶¹ Though the most obvious example of monistic ways of life will be religious in content, Gray also accepts that it is not only certain religious believers who will find value pluralism contentious. He freely admits that his value pluralism also puts him at odds with most Enlightenment political and moral philosophies, by which I assume he means Kantianism, utilitarianism, and Marxism, to name but the most prominent. Such ways of life and moral theories are, according to Gray, illusions that deny a fundamental truth about the plural nature of human morality. Henry Hardy has gone further, stating that “if [value] pluralism is true, all monisms are false, and it is not honest to pretend otherwise. To be consistent, the pluralist must look forward to the time—even though it may well never come—when the holding of monist beliefs is regarded as just as strange as the belief in the propriety of slavery, or in the divine right of kings.”Footnote ⁶² He goes on to say, even more strikingly, “One cannot pick and choose: benign or otherwise, monism is the enemy of pluralism; which is to say, if pluralism is true, the enemy of truth.”Footnote ⁶³ If Gray and Hardy are right in thinking that value pluralism necessarily commits us to rejecting monistic ways of life, religious and secular, then there is good reason for thinking that the politics of modus vivendi on which it is based will be deemed unacceptable by many.Footnote ⁶⁴

Whether value pluralism need necessarily be so hostile to monism, especially religious forms of monism, is an interesting question. William Galston and Michael Jinkins have separately argued that religious beliefs are not as obviously inimical to the truths of value pluralism as Gray and Harding have made out.Footnote ⁶⁵ And if this is true, then it might be the case that religious believers could be more favorably inclined toward modus vivendi (and value pluralists to religious belief) than previously thought. But while there are clear connections that can be made between value pluralism and modus vivendi, the controversial nature of the meta-ethical theory is likely to be a persistent source of disagreement for people who hold monistic conceptions of the good, even if they are generally tolerant of a plurality of varying and conflicting beliefs.

We would be better advised to try to side-step this issue entirely by seeking reasons independent of the fact of value pluralism for advocating modus vivendi. The fact of the matter is that there is a plurality of reasons why people highly value peaceful coexistence; some of these will be specific moral reasons internal to particular ways of life, others might be more instrumental or strategic considerations, while meta-ethical factors might play a role too. There is no reason why we should claim that only certain reasons are appropriate for endorsing a modus vivendi—be they an appreciation of the positive benefits accrued by a modus vivendi or, as John Rawls claimed, simple exhaustion of trying to battle with those with whom you differ. And, indeed, by restricting the category of appropriate reasons we make the justification for a politics of modus vivendi controversial and hence exclusionary of those who cannot accept it. In this sense, and to adopt Rawlsian language for a moment, we need to make certain that modus vivendi is “political not metaphysical” by ensuring that it is not grounded in any particular comprehensive moral or meta-ethical view in order that we can achieve an overlapping consensus of different ways of life about the relevant institutions, all of whom might have very different reasons for supporting it. So, if we were to pursue modus vivendi politics as a way of developing liberal realism, we would do well to disconnect it from any controversial justification that would only serve to replicate the disagreements about the normative foundations of politics that the realist challenge highlights.

However, there are still two significant concerns about using the politics of modus vivendi as the basis for developing a theory of liberal realism. The first is that it is questionable whether the desire to reach reconciliation with those with whom we disagree, which is the minimal normative foundation at work in Gray's theory, is really as universal or pervasive as he assumes. The easy point to make here is that while it is undoubtedly true that all ways of life do indeed value peace very highly (I doubt if there are any that do not value it at all), there are some, for example those that place high worth on honor or military duty, that do not. The more important point is that while most ways of life do value peace very highly, sometimes conflict with particular individuals, or more likely groups such as “infidels” or the “bourgeoisie,” is positively encouraged by particular moral or religious doctrines either as a means to achieving other goods (such as the favor of God) or as something valuable or required in and of itself. In such more familiar cases, the worry is not that these ways of life do not value peace at all, but that they do not desire peace with particular groups, and at the extreme may prefer policies of repression. On the one hand this claim can be an empirical one that highlights the presence of such ways of life, even if held only by a minority of people, within modern liberal democratic societies. But it can also be a theoretical claim pressed by the realists that serves to remind us that the widespread agreement that all human beings are moral agents deserving of equal concern and respect is a contingent and fragile achievement, and that even when reflected in the dominant human rights discourse we should nevertheless recognize that this commitment has only very recently achieved anything near universal acceptance. Indeed, the claim of universal human worth and dignity has proven to be one of the most contestable moral beliefs in human history, both in theory and in practice, and, the realist would add, so we should expect it to be, given the inevitable conflict and disagreement that will surround such fundamental normative commitments. In this sense, it is not clear why we should even assume that the desire or moral demand to treat others as equals is a normative universal, one that all persons should share and that therefore provides the basis for a hypothetical consensus, even if they actually do not.

This is a particular problem for our purposes because, like the negative approach discussed above, what is potentially attractive about the politics of modus vivendi is that it allows us to overcome the realist challenge by finding common ground on minimal normative commitments, in particular peaceful coexistence. To be fair to Gray, he does accept the validity of this problem, though he claims that this is not a peculiar disability of modus vivendi but, because all political ideals are contingent goods, rather a limitation of all forms of ethical reasoning.Footnote ⁶⁶ If we detach the politics of modus vivendi from value pluralism, as I have suggested we need to do in favor of an overlapping consensus of different reasons for endorsing institutions that engender reconciliation, then we cannot have recourse to this defense. For our purposes, then, the fact that some ways of life do not value peace very highly or, more likely, would not seek resolution of their conflict through reconciliation with particular groups, means that it is unlikely that a commitment to peace can provide a minimal but universal normative basis necessary to overcome the realist challenge.

Even if this problem could be overcome, another possible difficulty remains. Gray is clearly right in saying that the sort of political framework that can deliver the resolution of conflicts need not take a liberal form; in some instances a liberal regime may indeed be the best way of achieving peaceful reconciliation between different ways of life, but in others nonliberal regimes might do just as well, if not better. There is an indeterminacy in the theory of modus vivendi about what particular political institutions and practices it justifies. While there might be several reasons for thinking that this is a significant advantage to the theory, especially in developing a realist political philosophy, it is a problem for our purposes insofar as the aim of liberal realism, as I set out in the first section, is to develop a compelling realist defense of a model of political association that is distinctively liberal. Insofar as the theory of modus vivendi is politically indeterminate, it is uncertain whether it can meet this particular objective.

Partisan Liberal Realism

The two accounts of liberal realism explored above both seek to address the realist challenge to liberal legitimacy by restricting the required field of agreement to uncontroversial normative premises that could (hopefully) be endorsed by all persons. The strategy of the two approaches is therefore to retain the notion that legitimacy requires consensus on shared principles but to rely on beliefs or convictions that are not the subject of the sort of inevitable and ineradicable disagreement that, the realists charge, disrupts the more familiar accounts such as that offered by Rawls. However, in both cases I have suggested that even these attempts to find less contentious principles that can be universally affirmed are still likely to be vulnerable to the realist challenge.

An alternative strategy to employ might be to accept the realist contention that disagreement and conflict will hamper any endeavor to uncover a universal consensus about principles necessary for liberal legitimacy, but to try to internalize this insight into liberal theory itself. This will inevitably require us to alter the conditions of liberal legitimacy, at the very least to abandon universal consent as a necessary condition of legitimacy. But because our aim is to explore the possibility of developing a response to the realist challenge that is still distinctively liberal, we cannot do away with the connection between consent and legitimacy altogether. Instead it might be possible to restrict the relevant justificatory constituency of individuals to those who do share common principles, which could then act as the criteria for the legitimate employment of state power. This would retain a connection between consent and legitimacy, though on this model only the consent of those who share common liberal principles is relevant from the point of view of legitimating political power; the fact of disagreement with those principles does not disrupt the legitimacy of using state power in accordance with them.

This response to the realist challenge would therefore openly endorse the necessarily partisan nature of the shared principles underpinning liberal legitimacy. It accepts that the consensus about these principles will not be universal and that there will be others in society who will indeed disagree with them. Ultimately, and maybe at first surprisingly, this recognition of the controversial and contestable nature of liberal principles makes partisan liberalism structurally similar to forms of political liberalism, at least that of Charles Larmore,Footnote ⁶⁷ who is more willing to embrace the fact that some will reject foundational liberal principles (in his case equal respect) than Rawls is, though neither see this disagreement as particularly troublesome for liberal theory.Footnote ⁶⁸ Indeed, on the one hand, such explicit partisanship has some benefits, such as relieving liberal theory from the thorny philosophical problem of trying to explain why those who do not endorse the principles employed by liberal legitimacy should do so. What is required of liberal theory is considerably reduced. On the other hand, a consequence of openly endorsing the partisan nature of liberal legitimacy is, in effect, to abandon the universal aspirations of liberalism. This would require us to accept that liberalism has many legitimate enemies, that is to say, adversaries whose rejection of liberalism and endorsement of non- or antiliberal values should not be seen as the basis of error or irrationality but as an inevitable consequence of the deep and intractable moral and political disagreement that realists have urged us to recognize.

Though partisan liberalism will likely share several features with Rortian antifoundationalist liberalism in virtue of their common attempt to avoid foundational claims that are assumed to have universal normative force, it is important to be clear that they are not the same theory. Importantly, partisan liberalism does not demand of its citizens that they hold their most fundamental moral and political beliefs with the ironic detachment that Rorty advocated and that proved to be one of the most controversial aspects of his theory.Footnote ⁶⁹ For Rorty, irony, that is to say the wholehearted commitment to one's beliefs while accepting their inherent contingency, is both the appropriate position to take in a disenchanted world devoid of objective truth or meaning and the most pragmatic route to achieving the central liberal aim of reducing the pain and suffering of others. But partisan liberalism need not accept the ontological thesis regarding the absence of value nor need it endorse that particular (Shklarian) view of the aims of liberalism. Accordingly, irony is not a necessary feature of the partisan theory. Indeed, it would undoubtedly make partisan liberalism too demanding, too partisan, if it addressed only those who held the right beliefs in the right way. The best strategy to employ, in a way that emphasizes the similarities between the partisan approach and political liberalism, would be silence on the question of what ontological or epistemological positions are most consistent with or demanded by liberal politics or morality. While this realist approach does require that citizens be partisan in relation to particular liberal commitments, it need not and, strategically speaking, should not take a stand on how citizens should understand the ontology or epistemology underpinning those beliefs. Partisanship does not, therefore, require irony.

However, there are parallels that it would be disingenuous to deny. Calling his political theory antifoundationalist was always somewhat misleading on Rorty's part; he did believe that good justificatory reasons could be offered in defense of liberal politics but wanted to avoid any sense in which those reasons (in Rortian parlance) were True, corresponded to Reality, or were consistent with the requirements of Human Nature. These reasons, in particular that liberalism reduces the pain and suffering of more people than any other political order, would be considered compelling for liberals who are committed to that political goal and the normative moral beliefs that underpin it but will probably not, nor do they need to have, similar normative force for those who do not share this commitment. But—and this is where the description was misleading—foundations need not be universal to still be foundations. For Rorty, liberalism is justified, but not according to the criterion of correspondence that he thought was demanded by prevailing accounts of truth and justification implicit in many features of Western culture, including liberal political theory. Partisan liberalism is equally happy to rely on justificatory reasons that make no claim to having compelling universal normative force and, like Rorty, does not think that these reasons are in any meaningful sense inferior or inadequate on that basis. Once we accept the realist thought that political disagreement is an inevitable and necessary feature of politics, the partisan appeal of such reasons need not be viewed as a deficiency. However, unlike in Rorty's theory, there is no reason why the partisan liberal need deny the further claim that these foundational reasons (be they moral, religious, or political) are true according to the various religious or metaphysical frameworks through which citizens understand and affirm those reasons.

But the fact that partisan liberalism shares this feature of Rorty's liberal theory means that it will be equally vulnerable to a familiar criticism: limiting the justificatory constituency to persons who share a common set of principles, and accepting that other persons will inevitably and ineradicably disagree with those principles, amount to little more than “liberalism for liberals,” with the consequence that “jihadism for jihadis,” “autocracy for autocrats,” “Marxism for Marxists,” and “anarchism for anarchists” can be justified on similar partisan grounds. The argument underpinning partisan liberalism can therefore be employed in the service of any and all political causes. Worse than this, partisan liberalism, like Rortian ironism, leaves the liberal theorist without the resources to criticize other political creeds and orders according to a common set of values and beliefs, which, if this is not relativism, as Rorty insisted, is very close to it.Footnote ⁷⁰

According to partisan liberal realism, this picture of radical political disagreement and the need to negotiate and live together with people with whom we fundamentally disagree without recourse to a common philosophical or moral framework should be seen not as a criticism but as a description of a problem, maybe even the problem, of politics. The inevitability and permanent nature of political disagreement means that the task of a theory of liberal realism cannot be to find ways to eradicate or theorize away such conflicts. Partisanship recognizes, though it does not celebrate, this fact and in doing so is better placed to respond to this criticism than political theories such as Rorty's. To criticize a political theory on the basis that it does not provide us with a philosophically justified design of (liberal) political practices and institutions that all persons can or should affirm is to replicate the mistakes that realists have charged liberals with continuously making. Partisan liberalism tells us how to best understand the theoretical foundations of the liberal-consensus account of politics and legitimacy in circumstances of radical political disagreement. There is, of course, still the question of how we are to live together with those who do not share fundamental liberal, moral, or political values in a way that is still consistent with those beliefs, but a politically responsible answer cannot be that we are to do so by attempting to eradicate those conflicts via philosophical argument, the objective that underpins this criticism of partisan liberal realism.

The fact that partisan liberalism accepts the ineradicable nature of political disagreement as a necessary feature of politics means that it can offer a more persuasive and forceful response to this criticism than is available to similar political theories. I think this represents another significant advantage of the partisan approach to developing a theory of liberal realism. Nevertheless, I want to draw attention to but one potential problematic implication that builds on some of the concerns that Stears has discussed regarding the use of coercive force and that may also have profound practical consequences, a problem generated by the restriction of the justificatory constituency that the partisan approach recommends.

According to much liberal theory, especially in its dominant Kantian vein, the answer to the question “Who must the state be justified to in order to be legitimate?” also provides an answer to the question “Over whom can the state legitimately employ its political power?” The constituency of persons for whom any state must be justified in order to count as legitimate is the same constituency of persons whom the state can legitimately coerce. The standard model of liberal justification requires that reasons be provided for the liberal state that can be accepted by all those over whom that state employs its political power. This is a necessary rather than a contingent feature of liberal theory because it is generated by two central commitments in liberal morality: firstly, that all individuals are of equal moral worth and are therefore due equal respect and, secondly, that a violation of persons' freedom occurs if they are forced to obey laws that they did not or could not have endorsed. It is by combining these two principles that we are led to the thought that the political regime must be justified to all persons subject to it in order to count as legitimate. So the constituency of persons over whom the liberal state can legitimately employ its power is the same as the constituency of persons to whom liberalism must be justified in order to count as legitimate. And necessarily so, given the moral basis on which the liberal account of legitimacy is grounded.

It follows from this that if the justificatory constituency of persons to whom liberalism must be justified for a liberal regime to be legitimate is restricted, then the legitimate use of political power will likewise be restricted to those persons also. Here is where the problem I want to point to arises because restricting the justificatory constituency necessarily opens up a theoretical gap in which we can account for how it is legitimate to use state power over those who endorse liberalism but not over those who reject it. Any use of political power over those who reject liberalism will necessarily represent a restriction of their freedom insofar as they have not consented to the liberal regime. Nor would the state be treating them with equal respect if it did so. The familiar liberal response is to claim that those who reject liberalism are being irrational or are guilty of some moral or cognitive error, in which case the use of state power over them is legitimated by the fact that they would support liberalism if they were only rational or reasonable (as they should be). We respect such persons' freedom and treat them as moral equals by taking account of what they should will rather than what they do will. But such responses are not open to us if we take seriously the realist insight that such conflicts and disagreements about politics are a necessary and inevitable feature of the political itself, and hence of human coexistence, rather than a consequence of error. As has often been feared, liberal appeals to the “universal real will” of individuals only serve to mask an attempt to impose partisan views on those who, with no unreasonable ground, reject them. According to the liberal account of legitimacy, therefore, by restricting the justificatory constituency we necessarily beg a question about the legitimacy of using state power over those who fall outside that constituency.

Further than this—and here we enter the realm of liberal practice—there is a concern that those individuals who are excluded from the justificatory constituency might not enjoy the same rights and toleration enjoyed by those included in it. Where the justificatory constituency is taken to be all individuals, the duties and obligations are owed to all individuals. Reduce the justificatory constituency and you reduce that group of persons to whom such duties and obligations hold. This concern has been voiced in a couple of places in particular with reference to the rights of unreasonable persons in Rawlsian political liberalism.Footnote ⁷¹ Jonathan Quong has defended political liberalism on this point, but his whole account ultimately still rests on the assumption that no one can reasonably reject liberalism.Footnote ⁷² Such concerns are particularly pertinent, I suggest, at a time when there is general concern about how the civil liberties of those who are suspected—not tried, let alone found guilty via due process—of attempting to undermine liberalism are being overruled by their governments in favor of security or order. It might indeed follow that those who reject liberalism (leaving aside the issue, which is obviously of incredible importance, of whether such people are guilty) are not owed the same rights as those who fall within the justificatory constituency, but if this is true then it would have profound consequences for many contemporary debates.

These are most certainly significant worries that any full account of a partisan approach to liberal realism would need to address. A story is required that explains to us, once we accept the partisan nature of liberalism, why it is legitimate to coerce those who reject it. And this need is especially pressing when, as Stears has observed, political groups as diverse as animal rights protestors, Islamic fundamentalists, and antiabortion activists are turning away from the democratic process and embracing the threat or actual use of force as their preferred means of furthering their cause.Footnote ⁷³ But this new account of legitimacy must be one that is still distinctively liberal in character. This means that prudential or pragmatic appeals to the need to maintain law and order or Schmittian accounts of the existential threat that the other poses are not likely to be satisfactory. Neither, even if true, will easily satisfy the liberal demand for respecting the moral equality and freedom of those who disagree with us. Every society, maybe especially liberal ones, will always include some people who do not accept the political regime. We need an account of what legitimates the coercion of such individuals, especially in cases when they seek to undermine liberal institutions, possibly via the use of violence. And, in order that the account remain distinctively liberal, it will have to somehow retain a sense of how such persons are being treated with respect as moral equals when state power is used over them—even though a restriction of their freedom is, on liberals' own terms, taking place.

What would be needed, then, is an account of liberal legitimacy that explains why it is legitimate to use political power over those who reject liberalism that is consistent with the respect owed to them as free and equal moral agents. Let me highlight just two possible options. The first would be for partisan liberal realism to try to develop a two-tier account of legitimacy, one that maintains the standard liberal link between justification and legitimacy for the liberal state and those who endorse liberalism but has a different account of what legitimates the use of state power over those who reject liberalism. Once we accept that the foundations of liberalism are necessarily contested and partisan, there is no need to think that the basis of legitimacy need be the same for all individuals. The second possible route would be to separate two questions that liberal theorists tend to run together: “What are the conditions for a legitimate state?” and “What are the conditions for the legitimate employment of political power?” A standard liberal answer to this latter question is that political power can only be legitimately employed by a legitimate state or in accordance with legitimate constitutional essentials. Hence the two questions are conflated. But it may be that if we separate the conditions for a legitimate state and the legitimate use of coercive power, then we could carve out a position in which a state can be illegitimate yet use its power legitimately, or more relevantly for our purposes, in which a liberal regime could be rejected by an individual, and hence illegitimate from his perspective, yet can nevertheless legitimately employ power over that person (and others like him). In such a scenario it would be a condition of state legitimacy that it receive the consent of those it rules, but this would not be a condition of the legitimate use of state power. This would mean that the conditions for the legitimate use of state power would not include consent and would more likely take the form of conforming with pre- or nonpolitical criteria, with basic (by which I mean so-called first generation) human rights being the most obvious candidate. According to this “mixed mode” account of liberal realist legitimacy, while the legitimacy of the state would turn on whether it enjoys the consent of those over whom it rules, the legitimacy of the use of its power would depend on the purposes to which it is put.Footnote ⁷⁴

Final Remarks

This article has sought to make a first move to responding to the realist challenge to liberal legitimacy by setting out, in the broadest of terms, three possible frameworks for the development of a theory of liberal realism. It has also discussed the potential virtues and problems of each. I have argued that any developments of the negative or minimal approaches to liberal realism are highly unlikely to ever fully overcome the realist challenge because even though they attempt to uncover a consensus about less contentious principles, this pursuit nevertheless represents a continuation of the standard liberal strategy. Accordingly, they are likely to be vulnerable to exactly the same criticism from the realists regarding those principles' contestability. Making the foundations thinner does not automatically make them universally acceptable, and realists are likely to push the point that universal assent is both highly unlikely and an unreasonable hypothetical assumption to make in the face of the inevitability of political disagreement. I am therefore led to believe that the partisan approach to liberal realism is likely to be the most fruitful line of inquiry, largely because it is discontinuous with the liberal strategy of seeking universal consensus while retaining the relationship between consent and legitimacy. I also think it is superior to other nonuniversalist liberal theories, such as Rorty's, on the basis that it can view familiar criticisms as grounded in mistaken views of the political which the realist view of the ineradicable nature of political disagreement rectifies. Nevertheless, there are still substantial concerns that this approach needs to address, and in discussing how liberal accounts of legitimacy necessarily take the legitimacy of employing state power over individuals to be conditional on their consent to it I have highlighted what I take to be one of the most significant. This problem, and no doubt many more, would need to be addressed in a full account of partisan liberal realism.