Radicalizing realist legitimacy

1. Introduction

Critics of prominent realist theories of political legitimacy often accuse them of status quo bias. Alan Thomas argues that realist legitimacy – specifically in the form articulated by Bernard Williams – is what John Rawls calls ‘political in the wrong way’, meaning that it is too accommodating of certain anti-liberal positions. Thomas is especially troubled by Williams’ willingness to countenance the possibility that a theocratic state which coercively imposes restrictive gender roles might, under certain circumstances, possess a degree of legitimacy. ‘Williams’ hedging as to whether or not such a hierarchical society meets the basic condition of legitimacy’, he argues, ‘illustrates the hazards of his weak and widely shared conception’. ¹ In a similar vein, Fabian Wendt is especially disturbed by the possibility that Williams’ political theory could be interpreted as allowing that slave-owning states of antiquity can qualify as legitimate. He asks: ‘How far should we go down the road to pure relativism about legitimacy? Is it not cynical to say that Babylonian slavery was perfectly legitimate?’ ²

The picture of Williams’ realism painted here is of an excessively bleak and pessimistic theory, willing to accept some of the most objectionable forms of injustice in order to secure the basic political goods of stability and good order. ³ Many realists, for their part, have done little to deny this. William Galston’s influential overview of realist literature presents realists as fearing an ‘excessively ambitious’ approach to politics, in contrast to the approach of moralists, whose fear is that ‘an experience-based concept of feasibility will preserve an unjust status quo that could be changed through determined action’. ⁴ The frequency with which realists appeal to the ubiquity of conflict, ⁵ the need for compromise ⁶ and the importance of modus vivendi arrangements ⁷ seems to confirm the picture of realism as a kind of resignation to an admittedly depressing status quo.

It would be one thing to demonstrate that Williams’ theory of legitimacy need not have the kind of status quo bias that critics like Thomas and Wendt attribute to it. It would be quite another to show that it might support a kind of radically anti-status quo politics. In this article, I aim to do just this. Williams’ theory, I argue, can facilitate a way of thinking and engaging with politics that can enable and facilitate subversive criticism of the status quo. I do not attempt to show that political legitimacy is best understood in the way Williams suggests. I expect that my argument will not make those who remain unpersuaded by attempts to defend the coherence of Williams’ theory any more sympathetic to it. ⁸ Rather, my intention is simply to show that it can deal effectively with the charge of status quo bias – and, in fact, go a great deal further.

There is already a growing body of literature that resists the characterization of realism as a necessarily conservative force. ⁹ It would therefore be helpful for me to distinguish my argument from two others.

First, I am not aiming to show that realist methodology in general does not entail status quo bias, because I think this has been sufficiently established by others. Lorna Finlayson demonstrates that seemingly conservative aspects which dominate contemporary realist literature have only a contingent connection to realism, reflecting more closely the political convictions of authors than realist methodology. ¹⁰ I aim to show that Williams’ theory of legitimacy is an example of a form of realism in which this connection can and should be broken.

Second, my claim that Williams’ theory of legitimacy may have potentially radical implications is not based primarily on the claim that certain aspects of it may have considerable emancipatory potential when it comes to challenging various hierarchical arrangements like slavery and restrictive gender roles. ¹¹ I will briefly discuss the relationship between Williams’ theory and hierarchical states, but my main purpose in doing so is to help clarify what I take to be the essential features of Williams’ theory. Instead, I argue that its radical potential derives from two important features. First, it need not commit us to the view that contemporary liberal states have superior legitimation credentials when compared with alternatives. Second, even where it requires us to acknowledge that certain forms of political power are legitimate, it may still facilitate of a kind of subversive action which may challenge power in spite of its legitimacy. ¹²

In making this sort of argument, I should acknowledge that I am likely taking Williams’ theory of legitimacy in a direction that he did not intend. Much of Williams’ writing on politics arguably reflects a kind of status quo bias that at least partly resembles that of other realists. Like Horton, he seemed to support the idea of a modus vivendi political arrangement as a worthwhile political goal, to such an extent that ‘those who enjoy such a thing are already lucky’. ¹³ Williams was also a committed liberal, aligning his theory of legitimacy with Judith Shklar’s ‘liberalism of fear’. ¹⁴ If I am correct, then Williams’ realist methodology can be deployed against some of his own political positions. In particular, I suggest that the same historical considerations he invokes in support of liberalism can also be used against it.

This article is in six sections. Sections 2 and 3 summarize Williams’ theory of legitimacy, and how it does not provide support for hierarchical states. Sections 4 and 5 focus respectively on my two arguments for the radical potential of Williams’ theory of legitimacy: that it allows for critique of illegitimate political power in a way that does not compel us to accept the legitimacy of the liberal state; and that recognizing the legitimacy of certain forms of political power may go hand in hand with a kind of political activism which seeks to subvert it. Section 6 concludes.

2. Williams’ theory of legitimacy

Williams’ theory of legitimacy has two important components. The first component is largely the same as what other realist theories of legitimacy focus on, namely, the importance of order and stability. This distinctively Hobbesian point is the ‘first’ political question ‘because solving it is the condition of solving, indeed, posing, any others’. ¹⁵

Williams does not, however, believe that the provision of stable political order is sufficient to secure political legitimacy. His theory has a second component: the political order must ‘make sense’ to its citizens. ¹⁶ Williams seems deliberately vague about what it means for something to ‘make sense’ here, noting that ‘it is indeterminate about what it requires: it says not much more than that coercion requires legitimation and that the will of the stronger is not itself a legitimation’. ¹⁷ He refers to this second component as the Basic Legitimation Demand (BLD).

Importantly, Williams claims that the BLD is not necessarily satisfied whenever citizens accept the exercise of political power. There seem to be two important exceptions. The first is when citizens are in situations of ‘radical disadvantage’. This means that they live in a state of Hobbesian fear, where they are vulnerable to ‘coercion, pain, torture, humiliation, suffering, death’. Insofar as Williams believes in the existence of universal human rights, they are exclusively concerned with situations of radical disadvantage, so understood. The BLD cannot be satisfied for any person who finds herself in such circumstances, whether because of the state’s actions or because of the state’s failure to protect her. The second exception concerns what Williams calls the Critical Theory Principle. If the whole point of the BLD is to effectively distinguish political authority from brute coercion, then citizens’ acceptance of a form of power is not sufficient to satisfy the BLD if that acceptance has in some way been produced by that same power. If citizens accept the authority of the government only as a result of the government’s propaganda, then this acceptance is not enough to show that the BLD has been met.

Now perhaps the most obvious question surrounding the BLD is its scope. In order for a political institution to count as legitimate, who are the people for whom the BLD must be satisfied? Two possible answers seem to suggest themselves: (i) the BLD must be satisfied for all affected by the exercise of political power; or (ii) the BLD must be satisfied for all citizens. Both answers are mistaken.

The first answer is mistaken because, as Edward Hall points out, Williams does not believe that the legitimacy of political institutions depends on the extent to which their legitimation story is accepted by non-citizens. ¹⁸ Consider Williams’ discussion of the politics of Ancient Sparta. He does not claim that the exercise of political power in Sparta was legitimate. He does, however, insist that the fact that political power was not accepted by the Helots is not grounds for regarding it as illegitimate. This is because the Spartans considered the Helots to be internal enemies, whom Sparta made no ‘attempt to incorporate’ as subjects. ¹⁹ This is not to say that the way in which political institutions treat non-citizens will have no impact on its ability to satisfy the BLD, but only that any such impact will be contingent on the attitudes of the citizens. If the Spartan citizens became so displeased with the treatment meted out to Helots that the exercise of political power in Sparta no longer made sense to them either, then this would amount to a loss of political legitimacy. Insofar as this was not the case, however, the legitimacy of political power in Sparta was unaffected by its treatment of Helots.

Consider now the second answer. What if there are some citizens for whom the exercise of political power does not make sense? Citizens – or if you prefer, subjects – are for Williams distinguished from non-citizens by the fact that rulers have something to say to them about why the power they exercise over them is justified. ²⁰ But having ‘something to say’ is not the same as saying something that ‘this person or group will necessarily accept’. ²¹ Some citizens might not accept the justification that rulers offer to them. Even so, as Hall points out, ‘Williams helps us to see that politics is still occurring with respect to those to whom the order can be legitimated’. ²² That is, we may say that ‘legitimation’ occurs whenever a citizen does accept the rulers’ exercise of power over them, and this acceptance is not produced by the rulers’ power or situated in radical disadvantage. ‘Politics’ is occurring for such people because their relationship with the rulers is properly political in character. It can be successfully distinguished from a relationship of brute coercion.

We might then speak of not just one, but two ways in which Williams uses the concept of legitimacy. First, ‘legitimacy’ is a property of political institutions, which they may or may not possess, or may possess to some degree. Second, ‘legitimation’ is a property of the relationship between political institutions and each individual citizen, obtaining if and only if this relationship is properly political – that is, distinguishable from sheer domination. Although Williams does not explicitly make this distinction, he does use the words ‘legitimacy’ and ‘legitimation’ in ways that seem to map neatly onto this distinction. ²³

Now legitimacy and legitimation are obviously related. The legitimacy of political institutions will depend on how successful they are at achieving legitimation to each of their citizens. But as we have seen, the legitimacy of a political institution does not necessarily require legitimation to all its citizens. Can Williams’ theory offer us any further clarification than this on the relationship between legitimacy and legitimation?

The answer, I think, is some, but not much. There are two points worth observing here. First, Williams states that the legitimacy of a political institution can be assessed in scalar, non-binary terms. ²⁴ We might say that, other things being equal, the legitimacy of an institution increases with the percentage of its citizens to whom it successfully legitimates itself, although the fact that Williams singles out the importance of achieving legitimation to ‘young people’ and ‘influential critics’ may suggest that the state’s legitimacy may be more greatly affected by its ability to legitimate itself to certain select citizens than to others. Second, Williams allows that binary judgements about the legitimacy of a political institution are sometimes possible, but that they are ‘artificial’ and ‘only needed for certain purposes’. ²⁵ At what point can we plausibly make such a binary judgement? Hall says the most precise answer we can give is: when the political institution in question has legitimated itself to ‘a sufficient number of people’ so that ‘the situation transcends the conditions of unmediated coercion in which politics is impossible’. ²⁶ Of course, it is much more difficult to identify when a ‘situation’ or society as a whole is not characterized by brute coercion than it is to do this for an individual citizen. Williams’ seemingly intentional vagueness on this question, however, should not prevent us from seeing what is the important point for our purposes here: the legitimacy of a political institution is not necessarily eradicated – although it may be reduced – if it should fail to legitimate itself to some people whom it does treat as citizens.

Let me summarize what I have said here. Williams’ theory of legitimacy consists of two features: a demand for stability and order and the BLD. We have seen that the BLD can be used to assess two distinct but closely related phenomena: the legitimacy of a political institution and the legitimation or lack of legitimation that the institution achieves with respect to each individual citizen. Legitimation is achieved when an institution’s power makes sense to a citizen who does not live in a situation of radical disadvantage, and the citizen’s attitude towards the institution’s power is not itself produced by that power. The legitimacy of an institution, meanwhile, does not depend on its being acceptable to all affected by its exercise of power, or even on its achieving legitimation to all citizens. While its legitimacy may be best assessed in scalar terms, a binary judgement may be possible when we think we can plausibly whether the society in question faces a situation that is characterized by politics proper, rather than brute coercion.

3. The BLD and critique of hierarchical states

Recall that the objection to Williams’ theory of legitimacy pressed by Thomas and Wendt is that it has a problematic relationship with hierarchical states – those which institute slavery, gender hierarchies and other similar kinds of injustice. I think this objection can be formulated as follows:

According to Williams’ theory, hierarchical states can be legitimate in certain circumstances.

I grant that the first premise is true, but not for the reasons Thomas and Wendt believe. In any case, I argue that the second premise is false. I will discuss each premise in turn.

Thomas and Wendt are led to accept the first premise because they take the BLD to entail that the hierarchical state is legitimate if it is accepted by the people whom it disadvantages. For Williams, however, acceptance is not the same as legitimation. We saw that there are two circumstances in which power cannot be legitimated to a citizen regardless of whether she accepts it: (i) she is radically disadvantaged; or (ii) her acceptance is produced by the power itself (as per the critical theory principle). It seems that Williams held that slavery is a state of radical disadvantage. ²⁷ Hence, a slave-state can never be legitimated to the enslaved. Although it seems he did not believe that women living under a state which coercively imposes gender hierarchies are necessarily radically disadvantaged, it seems likely that any instances of women who accept the state’s power will depend on the exercise of state power itself. A scenario like this is an ‘easy case’ for the critical theory principle. ²⁸ Hence, there are two ways in which acceptance and legitimation can come apart. It seems plausible to suppose that most situations where hierarchical states are accepted by the disadvantaged will fail to count as legitimations for at least one of these two reasons.

However, Williams (2005, 27) also appears determined to consider what may strike us as the highly counterfactual case where, it seems to us, the hierarchical state is accepted by the disadvantaged in such a way that acceptance does produce legitimation. This would mean that the disadvantaged are not radically disadvantaged, and that their acceptance of the state is not produced by its power. ²⁹ While this might seem troubling, there are at least two reasons for thinking that this scenario is probably counterfactual. First, disadvantage often leads to a state of radical disadvantage. Consider how African Americans were vulnerable to lynching (and are still vulnerable to police violence), or how women in hierarchical states often lived under the daily threat of domestic violence or sexual assault. ³⁰ Second, it is difficult to determine with any degree of confidence whether a disadvantaged constituency actually does accept the exercise of power over them, as opposed to merely feigning acceptance or concealing their objections to it for prudential reasons. As Paul Sagar argues, James Scott’s influential study of oppressed constituencies provides substantial historical evidence suggesting they are normally fully cognizant of their situation. ³¹ Far from being ‘happily oppressed’, they acknowledge that they are treated substantially unjustly but develop ways of communicating this acknowledgement in such a way that is concealed from their oppressors. This way, they maintain a kind of dignity among their peers, while avoiding possible violent retribution from their oppressors. ³²

The first premise is not, therefore, supported by Thomas and Wendt’s concern that the BLD allows for hierarchical states to count as legitimate if they are accepted by the disadvantaged. It seems highly unlikely that the disadvantaged will accept the power of the state, but even less likely that they will accept it in a way that produces legitimation (i.e. their acceptance is not given under conditions of radical disadvantage nor manufactured by the state’s power). This does not mean, however, that the hierarchical state is entirely illegitimate; in fact, we can identify three ways in which it might retain at least some legitimacy. First, even if the disadvantaged are citizens, the state may possess some legitimacy if its power is widely accepted by other citizens. Second, if the state were to begin to treat the disadvantaged with brute coercion like the Spartan state treated Helots, it might retain more or less full legitimacy as long as this treatment of the disadvantaged is widely accepted by those whom it treats as citizens. And third, ‘unlikely’ is not the same thing as ‘impossible’: Williams’ theory does not entail that the hierarchical state is necessarily incapable of achieving legitimation, even if Scott’s empirical studies cast doubt over whether there have been any actual historical instances of this occurring. For these reasons, then, the first premise is correct.

This may seem very troubling to us. However, I think this is largely because we are accustomed to assuming the truth of the second premise – that the legitimacy of a state entails that it can rightfully coerce its citizens, and that its citizens should not attempt to overthrow it. Undoubtedly, this is how many moralist political philosophers view the concept of legitimacy. ³³ For Williams, however, things are not so simple. As Sagar points out, Williams follows Max Weber in defending an ‘internalist’ account of legitimacy, according to which ‘legitimacy is, and can only be, a function of the beliefs of those subject to power’. ³⁴ His theory is initially a description of what it means to have political authority at all, rather than brute coercion. He does not thereby commit himself to the usual set of normative assumptions about legitimacy that moralists typically make.

This does not mean, however, that Williams’ theory is purely sociological and not normative, as David Miller seems to suspect. ³⁵ The trouble for us is that Williams’ statements on the normative status of judgements about legitimacy are often unclear and sometimes contradictory. For example, at times he suggests that finding oneself being subjected to coercion without legitimation is a paradigmatic form of injustice. ³⁶ However, it is not clear that this sort of moralized judgement is consistent with the basic methodology behind Williams’ realist approach to politics. ³⁷ In what follows, I will try to reconstruct what I take to be the most charitable interpretation of the normative aspects of Williams’ theory, with reference to both legitimation and legitimacy, before explaining why I think this is the best way to interpret Williams.

First, consider legitimation. Recall that this concept is concerned with the relationship between a political institution and an individual citizen. One clear normative aspect of legitimation which Williams explicitly recognizes is ‘when we get to our own case’. If I judge that the institutions governing my society make sense to me, these institutions have successfully legitimated themselves to me. But I cannot coherently make this judgement without also ‘taking a stance’, as Thomas Fossen puts it. ³⁸ This stance entails that I believe I have some reason – though it is unclear how strong this reason is – to cooperate with them. To put it another way, when a particular state S has successfully legitimated itself to citizen C, thereby creating a genuine political relationship, C must take this relationship to have some normative implications for how she relates to S. It is important to remember, however, that legitimation can only occur between an institution and its citizens. Suppose C reflects on S₁, when she is not a citizen of S₁. There may be nothing stopping C from asking ‘does S₁ make sense to me?’ But regardless of how she answers this question, S₁ has not legitimated itself to her, because its rulers make no effort to ensure they have ‘something to say’ to her.

Next, consider legitimacy. Recall that the reason for distinguishing between legitimation and legitimacy was that it may be possible for a political institution to possess at least a measure of legitimacy even if it fails to legitimate itself to certain citizens. When we assess the legitimacy of our own political institution, we are still asking the same question, ‘does its exercise of power make sense to me?’ But when we are assessing the legitimacy of an institution of which we are not a citizen, this all changes. Williams’ internalist theory, we have noted, renders legitimacy a function only of the views of citizens. If C wishes to assess the legitimacy of S₁, she should therefore ask: ‘to what extent does S₁ make sense to its citizens? Has it managed to legitimate itself to a sufficient number of people, such that it can be plausibly described as a political institution, rather than an agent of brute coercion?’ Her answer to this question will be an evaluation of the legitimacy of S₁.

Suppose C determines that S₁ has achieved a high degree of legitimation, perhaps even to the extent that she can plausibly use the artificially binary distinction to describe it as ‘legitimate’. Does this judgement have any normative implications? Yes and no. On the one hand, there does not seem to be any contradiction if she judges that S₁ is legitimate, but is still in some sense a very bad state, and that it would be a good thing if it was overthrown by its citizens, even with outside help. ³⁹ If a hierarchical state ever really did achieve a measure of legitimation, we may want to say something like this. But this would not be a judgement about its legitimacy. ⁴⁰

At the same time, even though there is nothing inconsistent about judging another institution as legitimate but worthy of being overthrown, we can still hold that legitimate institutions are normally preferable to illegitimate ones. As Alice Baderin points out, Williams himself clearly thought it a good thing for states to be legitimate. ⁴¹ However, this is not because states have some kind of moral duty to legitimate themselves to their citizens but because of what illegitimate states normally do. There is enough historical evidence to show that the idea of a ‘benevolent tyranny’, if not a contradiction in terms, is seldom if ever realized. Those subjected to brute coercion often eventually find themselves in situations of radical disadvantage. Perhaps this normative consideration is reducible to a moral principle requiring us to prevent radical disadvantage, although Williams might well claim that this is ‘one thought too many’. ⁴² If illegitimate institutions typically fail to prevent radical disadvantage, or even exacerbate it, then we have a good pro tanto reason to want them to be legitimate. But this does not commit us to the view that a particular institution is necessarily worthy of our support in virtue of the fact that it is legitimate. On the other hand, it does seem plausible to suppose that our judgement that a severely unjust institution (e.g. a hierarchical state) is legitimate will have important strategic implications for how we relate to it. I will say more about this in section 5.

So this interpretation of Williams does not, then, commit us to the view that the BLD has no normative implications. Judgements in favour of legitimation are normative in our own case (but only in our own case), while judgements in favour of the legitimacy of institutions of which we are not citizens are normative insofar as legitimate institutions are generally better at protecting from radical disadvantage. But we might still ask why this interpretation is correct. Readers accustomed to moralist accounts of legitimacy may still be perplexed by the idea that legitimacy can be understood as a description of the nature of the relationship between citizens and their government, as opposed to some kind of moral assessment of the government itself. I find it hard to see this as anything other than an inescapable aspect of Williams’ theory, but is he right about this?

Since my primary aim here is to explore the implications of Williams’ theory for anti-status quo politics, I don’t want to attempt to defend Williams’ theory in its entirety here. But I think there is an important reason for regarding his theory of legitimacy as I have outlined it here – with its somewhat limited normative application – as essential to the realist position. A more substantively normative understanding of legitimacy is likely to draw upon some antecedent idea of the just exercise of political power or of the moral equality of all citizens. Such a theory of legitimacy would be another form of pre-political morality. While realism is not necessarily hostile to all forms of morality, it must avoid appealing to any form of morality that is pre-political in character. Proponents of realist theories of legitimacy have always insisted that the demand for legitimate politics does not arise from antecedent moral standards like justice and equality but from the practice of politics itself. ⁴³ Hence, Williams’ attempt to steer discussions of legitimacy away from pre-political moral commitments and towards examination of the nature of political authority itself should not be seen as an idiosyncratic aspect of his theory but as a necessary consequence of the realist emphasis on attention to politics as a distinctive category in its own right.

Let me summarize what I have said here. I have claimed that, while it is highly unlikely that hierarchical states will legitimate themselves to the disadvantaged, it is still possible for them to possess some legitimacy if their power is widely accepted by citizens. Hence, the first premise of the argument I have attributed to Thomas and Wendt is true, but not for the reason they believe. I have also argued that the second premise is false. Legitimacy, for Williams, is first and foremost an account of the nature of the relationship between an institution and its citizens, rather than an account of the moral permissibility of coercion or the obligations of citizens. I should repeat that I am not claiming that Williams’ theory of legitimacy is the correct theory, and that moralist theories are necessarily mistaken. I hope only to have shown that the BLD does not provide normative support for hierarchical states.

4. Critique of illegitimate power

So far, I have defended Williams’ theory of legitimacy from the charge that it is biased towards the status quo in a way that might ally it with hierarchical states. I think, however, there is more that can be said for it than the mere fact that it avoids this particular pitfall. In this section and the next, I suggest two ways in which Williams’ theory may facilitate a potentially radical critique of the status quo. I first examine how it may be used to critique illegitimate political power. As part of this, I will argue that, contrary to Williams’ own beliefs, it need not lead us to accept the legitimacy of liberal states.

Instead of critiquing political power in terms of its (lack of) conformity to moral principles, we can employ Williams’ conceptual distinction between legitimacy and illegitimacy. This would mean that brute coercion and domination replace injustice – insofar as injustice is necessarily a moralized concept – as the phenomena which activists seek to expose, denounce and eradicate. An advantage of this sort of critique is that the distinction between legitimacy and illegitimacy has a motivational force for most people which morality often cannot match. ⁴⁴ While universal moral truths – insofar as they exist at all – are seldom if ever universally recognized, let alone followed, the fear of being subject to coercion without legitimation is both widely shared and a powerful incentive for action. As Williams points out, ‘everyone needs not to be tyrannised’ and ‘everyone knows what tyranny is likely to bring with it’. ⁴⁵

Now Williams believes the legitimacy/illegitimacy distinction tells decisively in favour of liberalism. His reason for this is not derived from the moralistic liberalisms of Locke, Kant or Mill, but from Judith Shklar’s ‘liberalism of fear’. Rather than start with an extensive set of liberal values and ideals, this liberalism ‘takes the condition of life without terror as its first requirement’. ⁴⁶ It has a ‘unique right’ to speak to all humanity, because it addresses the only universal political concerns: ‘power, powerlessness, fear, cruelty, a universalism of negative capacities’. ⁴⁷ Its main goal, in other words, is the protection of citizens from radical disadvantage. This does not mean that the liberalism of fear aspires to nothing more than this, but that it seeks to attain further goods only once the first goal of protecting citizens from radical disadvantage has been secured, and with careful attention to how the pursuit of loftier political goals might threaten the security of this first goal under certain circumstances.

Why should we think that prioritizing this set of concerns provides support for the legitimacy of liberalism? The answer given by Shklar and Williams does not appeal to the capacities of an idealized liberal state but to the historical accomplishments of actually existing liberal states. According to Williams, modern liberal societies have done better than others at protecting their citizens from radical disadvantage, their other flaws notwithstanding. ⁴⁸ Since the failure to protect citizens from radical disadvantage is one way in which the legitimacy of a state can be undermined, it follows that liberal states have a better claim to legitimacy than their rivals.

This ‘historical argument’ for the superior legitimation credentials of liberal states can therefore be summarized as follows:

If a certain type of political institution P has been more successful at protecting its citizens from radical disadvantage than rival types of political institutions, then P has a better claim to political legitimacy than rival types.

The argument has a couple of significant limitations. For starters, the first premise is not straightforwardly true on Williams’ own terms. We have seen that the failure to protect citizens from radical disadvantage is but one of three ways in which states can fail to satisfy the BLD, the other two being when citizens do not accept the exercise of state power, and when their acceptance of state power is produced by the state’s power. Even if liberal states are better at protecting from radical disadvantage than others, they might still be worse in one of these other two respects. We can, however, deal with this problem by inserting a ceteris paribus clause into the first premise and conclusion. Exactly how strong this ceteris paribus case for liberal states might be depends on how severely they fail to legitimate themselves in these two other ways in comparison to non-liberal political institutions. I do not deny that liberal states do have significant failures of legitimacy in these respects. Many indigenous peoples, I take it, reject the legitimation stories offered for the liberal state’s authority, and any widespread impression that this is not the case is best explained by the way in which indigenous voices are often systematically silenced or ignored, or by the way in which states receive a boost in international credibility from instituting various arrangements which give a semblance of self-rule within the confines of a liberal regime. ⁴⁹ Nonetheless, I shall assume for argument’s sake that liberal states do not fail to legitimate themselves in these two respects to a significantly greater degree than their rivals.

Another limitation of the historical argument is that the conclusion is only that liberal states have better legitimation credentials than their rivals, not that liberal states are legitimate. It is certainly not true that liberal states have eliminated radical disadvantage among their citizens. Much could be said about the effects of police violence on the security of ethnic minorities. However, it is particularly worth noting one kind of radical disadvantage: that of women who experience domestic violence. It is noteworthy not just for its extraordinary prevalence but also because liberalism has arguably not merely failed to prevent this form of radical disadvantage, but actively contributed to it through its propagation of the public/private distinction. ⁵⁰ Insofar as those who suffer domestic violence live under conditions of radical disadvantage, they are people to whom the liberal state has failed to legitimate itself. Even if the historical argument is sound, it still allows for the use of the legitimacy/illegitimacy distinction as a basis for a critique of the failures of legitimation within liberal states. ⁵¹

This second point might give us grounds for thinking that the second premise requires some supporting historical evidence. Is it really self-evident, as Shklar and Williams seem to think, that liberal states have been more successful at protecting their citizens from radical disadvantage than others? They might, however, point out that there are various aspects in which liberal states appear to have done better than others (e.g. avoiding arbitrary imprisonment, cruel and usual punishment, etc.) while the failures of liberal states are not unique to them. Domestic violence is a serious problem in liberal states, but it is also a serious problem virtually everywhere. Again, we might want to question the extent to which liberal states really have ‘succeeded’ in these aspects, perhaps by pointing to the way in which various ethnic minorities seem to receive asymmetrical treatment from judicial systems. However, let us bracket this concern for now, noting that racism is not exclusively found in liberal states either. The historical argument still faces one small complication which Williams acknowledges only in passing:

Liberal societies are more successful in the modern world than others in helping people (at least in their own territories – their influence elsewhere has been less benign) to avoid what is universally feared: torture, violence, arbitrary power, and humiliation. ⁵²

The answer is that the historical argument starts to appear problematic if it turns out that the very reason why liberal states have done better than their rivals at protecting their citizens from radical disadvantage is that liberal states have played a significant causal role in the failures of non-liberal political institutions. This is something that defenders of Williams’ theory seem reluctant to acknowledge. Hall ⁵⁴ and Sagar ⁵⁵ both appeal to what they view as the ‘catastrophic failure’ of 20th century non-liberal regimes as evidence for the superior legitimation credentials of liberal states. Sagar also acknowledges that the United States’ actions in Latin America show that liberal states’ foreign policies can be ‘extremely unpleasant’. If, however, such ‘extremely unpleasant’ foreign policy is causally responsible for the failure of certain non-liberal regimes, as it undoubtedly was in the case of Salvador Allende’s government in Chile or Jacobo Arbenz’s government in Guatemala, then this may leave us feeling at least a little uneasy about the second premise. If we defend it by appealing to the ‘failures’ of non-liberal political institutions whose shortcomings are in no small part due to the interventions of liberal states, then its status is not unlike the claim of victory made by a gymnast who pushes opponents off the beam.

Perhaps, however, this is just part of the job of political legitimation. It may well be that certain historical non-liberal political institutions could have done a very good job of not inflicting radical disadvantage on their citizens, but legitimacy also requires that they protect their citizens from radical disadvantage inflicted by others. Perhaps non-liberal politics is poorly suited for this task in a world where powerful liberal states have a predilection for taking advantage of weaker ones not serviceable to their interests. In this sense, we might grant the truth of Williams’ equation ‘LEG + Modernity = Liberalism’, as long as we understand ‘modernity’ to refer, among other things, to the rampant predatory power exercised by certain liberal states outside their borders. Hence, the fact that the legitimation failures of many non-liberal political institutions were brought about in no small part by the intervention of liberal states does not stop them from counting as legitimation failures. Perhaps the second premise of the historical argument remains intact. ⁵⁶

I think, however, that this point about the foreign policy of liberal states might instead give us grounds for questioning the first premise of the argument. In light of the role played by liberal states in the downfall of non-liberal political institutions, we can plausibly dispute their claim to superior legitimation credentials even if we grant that they have been more successful at protecting their citizens from tyranny than their rivals. There are at least three reasons for this.

First, the current hegemonic status enjoyed by liberal states need not be treated as an immutable fact. As Finlayson points out, historical study can demonstrate variation far more easily than it can establish constants. ⁵⁷ Even if liberal states are currently the most powerful around the world, this was not always the case, and it might not be the case in the future either. Without a liberal hegemony, it might turn out that non-liberal political institutions are rather good at protecting their citizens from radical disadvantage. Of course, this is no more than speculation. The point is that we don’t yet know one way or the other, and that the interventions of liberal states have often been so devastating that the modern history of non-liberal politics gives us little if any clue regarding how such forms of politics might fare in a world no longer dominated by liberal states.

Second, even if the hegemonic status of liberal states will remain in place for the foreseeable future, it does not automatically follow from this that the foreign policy of these states must remain as aggressively interventionist and destructive as what they have exhibited throughout the 20th and early 21st centuries. Although one might suggest, as Raymond Geuss does, that there is a link between interventionist liberal foreign policy and liberal ideology, it would not follow even from this that liberal states will always retain this interventionist character. ⁵⁸ Real political institutions are often a product of different and potentially incompatible ideals and interests which have come together in a contingent way, even if it is possible to correctly identify one such ideal (i.e. liberalism) as dominant within a single political institution. Hence, just as liberal states may from time to time do illiberal things, it may also be the case that liberal states of the future will not always pursue the sort of foreign policy so characteristic of their 20th-century manifestations.

Third, even if liberal states continue to pursue interventionist foreign policies around the world, it may turn out that they can only sustain this by increasing reliance on brute coercion directed at their own citizens. Given that many of the foreign interventions of liberal states have required mass propaganda campaigns in order to navigate the difficulty of winning over an increasingly ‘casualty-averse’ public, ⁵⁹ it seems plausible to predict that any future interventions will require increasingly severe forms of state-friendly propaganda and information control in order to produce acceptance, or at least acquiescence. In other words, future liberal interventionist states are likely to find themselves violating the critical theory principle to an ever increasing degree. And if we also suppose, as I have, that coercion without legitimation usually ends up producing or exacerbating radical disadvantage, it follows that the capacity of future liberal interventionist states to protect their own citizens from radical disadvantage will be diminished. If the liberal states of the future continue to prosecute expensive wars that are highly unpopular at home as well as abroad, we might expect that this will be accompanied by arbitrary arrests, imprisonment, gutting of social services and further expansion of the national security state in ways that start to distinctively resemble Hobbesian fear. So even if we grant that liberal states have hitherto possessed the best track record of preventing radical disadvantage among their own citizens, this may no longer be the case in the future if they retain their interventionist character.

On this basis, I think we can plausibly reject the first premise of the historical argument. Even if we grant that liberal states may have better historical track records of protecting their citizens from radical disadvantage, it does not follow that they are, in fact, better at protecting their citizens from radical disadvantage, period.

I should again emphasize the modesty of the argument I am making here. I do not claim that we can clearly identify a particular non-liberal form of politics which possesses better legitimation credentials than the liberal state. My point is only that the historical argument fails. Since liberal states have played pivotal roles in crushing non-liberal political institutions throughout modernity, we are not well positioned to evaluate the capacity of the latter to protect citizens from radical disadvantage in a world without interventions from liberal states. Nor is it clear that liberal states can sustain future interventions without exacerbating radical disadvantage among their own citizens. Attempts to use history to demonstrate that liberal states possess the best legitimation credentials are therefore inconclusive at best.

I also do not claim that all instances of radical disadvantage within non-liberal political institutions are due to the machinations of liberal states, still less that all liberal states are somehow illegitimate. It would be highly questionable, to say the least, to claim that the foreign policy of liberal states was the sine qua non causal factor in the mass killings carried out by Stalin, Mao and Pol Pot. But I think my criticism of the historical argument would still hold as long as liberal foreign policy might have been the sine qua non factor in the failure of at least some non-liberal political institutions. Hence, I hope to have shown that Williams’ theory of legitimacy does not necessarily entail that liberal states are either legitimate in absolute terms or in possession of better legitimation credentials than all others.

5. Critique of legitimated power

I have argued that one way in which Williams’ theory might serve a more radical politics than Williams himself may have anticipated is that using the legitimacy/illegitimacy distinction to criticize illegitimate power involves criticizing many dominant features of existing political regimes, including liberal ones. But what if we find that the state’s power has been successfully legitimated to many or most of its citizens? In at least some such instances, let us assume, radical activists will nonetheless want to criticize the form of power in question. Although I have argued that the case of the legitimate hierarchical state is largely unrealistic, there may be other instances where we think a legitimate form of power should be resisted, or its illegitimacy is very difficult to establish. Perhaps the existence of free markets in education is an example of this. Sure, we may think the critical theory principle applies here, and that we can often show that the acceptance of the forms of power which establish for-profit educational institutions is produced by the power itself. Let us suppose, however, that this is not the case for a great many citizens. That is, exercises of state power which guarantee free markets in education make sense to them in such a way that satisfies the BLD. Let us call forms of power which are in some sense undesirable but have nonetheless attained widespread legitimation ‘L-power’. ⁶⁰

How should activists respond to L-power? It is important to distinguish between: (i) what kinds of power have/have not currently been legitimated; and (ii) what kinds of power we would like/not like to be legitimated. The fact that L-power has been legitimated does not mean that activists should not seek to resist it. However, the fact that it has been legitimated seems highly relevant for how they might go about this. I think there are two main strategies which they might adopt.

First, activists might campaign for the immediate eradication of the L-power in question. It may be that they believe that it can be replaced with an alternative kind of power which is also at least as capable of attaining widespread legitimation. In certain circumstances, however, the L-power might be the only option capable of this. In such cases, any efforts at eradicating the L-power would involve a significant measure of brute coercion, at least for a certain period of time. Consider again the case of the hierarchical state. If there ever really was a scenario (to be clear: I don’t think there is, or ever was one) where so many citizens, including most women, were steadfastly committed to the view that the state must impose restrictive gender roles on women, then any removal of these restrictions will involve brute coercion of a great many people, and the use of methods far more coercive than liberal models of persuasive civil disobedience. ⁶¹ Of course, the advocate of this path of resistance may claim this is a price worth paying, and hopefully, in the not-too-distant future, the experience of life without restrictive gender roles will cause most citizens to come to regard their removal as legitimate. The question of whether activists should seek the immediate eradication of L-power such as this is not one that can be answered a priori; it will involve a kind of political judgement which takes into account the severity of the badness of the L-power, the likelihood of its replacement gaining swift and widespread legitimation and the costs necessitated by the reliance on brute coercion and disruption of order in the meantime.

Second, activists may attempt to shift the views of the public regarding the L-power, so that it will eventually come to be seen as illegitimate itself. Sometimes this might involve attempting to persuade people to abandon certain beliefs, including moral beliefs. However, if we recall the relatively weak motivational force of morality, other methods might be more appropriate. Since legitimation narratives can be sustained or undermined by the mobilization of citizens’ passions or affective ties, activists have at least some reason to pay close attention to these as well.

I want to discuss one method which activists might use in order to change public opinion about L-power. This involves highlighting particular aspects of the behaviour of the state exercising the L-power which cannot be legitimated to citizens. Prominent examples of such ‘illegitimate actions’ include corruption, nepotism, and most instances of deception. Suppose a government decides to remove environmental regulations on fossil fuel industries, and attempts to legitimate this decision by claiming that it will help protect workers’ jobs and salaries. However, it is later revealed that the public funds previously used to enforce these regulations are being pocketed by government personnel. Additionally, the government publicizes unemployment and wage growth statistics, which are then publicly exposed as deliberate fabrications. It may be that neither of these illegitimate actions immediately causes the government’s methods to cease to make sense to the public. Nonetheless, it might be especially fruitful for activists to ensure that these illegitimate actions are widely publicized, and to suggest that they might in some way be connected to the legitimation story which the government has tried to sell in order to justify its regulatory rollback. If the main beneficiaries of the policy are government personnel, what does this say about its claim to be motivated by concern for workers? And if the government thinks it must lie in order to sell its story, what does this say about the story’s plausibility? It may be possible for activists to point to such illegitimate actions as evidence that the government’s legitimation story is not what it seems, and in so doing, may ultimately lead people to reject it. ⁶²

Much more could be said about the methods which activists might use to combat L-power by attempting to cause it to lose its capacity for legitimation. The important point to note is that activists are not faced with a binary choice between replacing an L-power in a potentially tyrannous fashion or resigning themselves to its continued existence. They may also seek to alter the conditions which determine what sorts of power can and cannot attain widespread legitimation. It may be possible and often best for activists to adopt both strategies; that is, they may campaign for the immediate replacement of L-power, all the while working to delegitimize it, and pave the way for different sorts of political arrangements which they consider preferable.

It is important to acknowledge that, while this kind of critique of legitimated power may be used to further radical anti-status quo political positions, it does not necessarily do so. The particular use to which it is put will ultimately depend on the sort of politics activists consider desirable. ⁶³ What I hope to have shown here is that activists who do pursue an anti-status quo politics may find this kind of critique of legitimated power a helpful and effective tool to have in their political armoury.

5. Conclusion

In this article, I have tried to show that Williams’ theory of legitimacy is not only free of unhealthy status quo bias but may also be a useful ally for radical politics. The legitimacy/illegitimacy distinction enables activists to expose and criticize illegitimate power while also providing helpful resources for resisting undesirable forms of power that have nonetheless managed to attain widespread legitimation.

I am mindful of the fact that I have been trying to show how the political theory of a philosopher with undeniably liberal leanings can be taken in directions that may be anything but liberal. Some may find any such attempts problematic, either because they do violence to the integrity of the philosopher’s thought, or because they are repulsed by the idea of bringing the work of another (white male) liberal to the rescue of radical politics. I am troubled by these concerns, and I don’t have any good reply to offer them. At the very least, however, I hope to have demonstrated one way in which the realist method of political reflection, involving closer attention to real-world political phenomena, is something that can be of great service to those who seek a kind of politics which poses a radical challenge to present hegemonic political arrangements.