Normativity in Realist Legitimacy

Introduction

Proponents of realist theories of legitimacy and their critics usually agree that political legitimacy is a normative concept. ¹ This means that the fact that an institution is or isn’t legitimate gives people some kind of reason to do or not do something. Typically, a judgement that an institution is legitimate is taken to provide at least a prima-facie reason for refraining from revolutionary action (Simmons, 1999). Realists may be somewhat more cautious about automatically inferring the existence of such a reason from the fact that an institution is legitimate (Cross, 2019: 7–9). Still, they will usually grant that a legitimate institution is one which we have some reason to support or at least not overthrow.

As with many of our normative concepts, we should find it of some interest to ask why the concept of legitimacy should have such a normative status. Here, the moralist critic of realism has a relatively simple answer: we have reason to support legitimate institutions because they satisfy certain moral desiderata. For a moralist like Charles Larmore (2017: 42), a theory of legitimacy ‘appeals therefore to a moral principle describing the conditions under which coercive power may justly be exercised’ and can thus be accurately described as ‘a morality prior to politics’.

Realists reject this answer. They hold that the normative force of judgements about legitimacy does not derive from antecedent moral values in the way that Larmore suggests. Politics, they frequently say, is not applied morality. But if judgements about legitimacy are not applications of moral judgements, where do they get their normative force from? Whence came this normativity, if not from morality?

My impression of the literature is that, in spite of moralist criticisms of realist claims to a distinctive kind of normativity (Erman and Möller, 2015; Maynard and Worsnip, 2018; Thomas, 2017; Wendt, 2016), this is a question which realists have been reluctant to address, albeit with one or two notable exceptions (Jubb, 2019; Rossi, 2010). Yet, there is one feature which seems to be play a role in almost all attempts to account for the normativity of realist legitimacy. This feature focuses on Bernard Williams’s idea that the very idea of politics implies legitimacy. It holds that there is a conceptual distinction between a political situation and a situation of war or terror. Call this the ‘Williams-premise’. If the Williams-premise is correct, then realists can identify evaluative standards internal to politics without appealing to ‘a morality that is prior to politics’ (Williams, 2005: 5). To ‘practice politics’ just means to aim for legitimacy.

Moralist critics of realism complain that this is insufficient to establish the normativity of realist legitimacy. Even if we grant the truth of the Williams-premise, the fact that political practice requires aiming for legitimacy doesn’t tell us that legitimacy is desirable. As Erman and Möller (2015: 540–541) claim, ‘a conceptual distinction does not entail prohibition’. Or as Leader Maynard and Worsnip (2018: 784; see also Wendt, 2016: 241–245) put it, realist theories of legitimacy:

simply . . . push the normative question back from ‘Why, in a given situation, should we practice politics in one way or another?’ to ‘Why, in a given situation, should we practice politics, rather than something else?’

Existing realist literature suggests two answers to this objection. The first answer holds that realist theories of legitimacy are normative only in the sense that they defend the Williams-premise and explore its implications (Hall, 2015; Jubb, 2019; Sleat, 2016). I call this answer ‘concessive realism’ – a term I borrow from Thomas (2017: 307) – because it allows that the need for political practice itself may stem from ‘a starting point within the ethical, broadly conceived (Thomas, 2017: 306). The second, less common answer, supplements the Williams-premise with a further claim about the unavoidability of political practice (Horton, 2012; Rossi, 2010). I call this answer ‘naturalist realism’. Naturalist realism does not delegate the task of explaining why we should practice politics to ethics; it denies that this question requires a normative answer at all.

In this article, I want to show that both of these answers, as they have hitherto been defended in the literature, are inadequate. Concessive realism problematically overlooks the way in which Williams-type political realism is in no small part motivated by a critique of morality (Hall and Sleat, 2017; Owen, 2017). This critique, in short, shows how reflection on certain practices often has a destabilising effect, making it difficult to retain ‘confidence’ in them (Williams, 2011). This critique may not necessarily destabilise our confidence in everything which might be called ‘morality’, but Williams (2011: chapter 10) does believe it does this for what he calls ‘the morality system’. Hence, realists must reject the concessive realist claim that there is nothing problematic about answering the question ‘why should we practice politics?’ by appealing to morality. Naturalist realism, meanwhile, does not take account of the way in which alternatives to legitimate politics may remain real options for at least some people. This doesn’t necessarily refute all forms of naturalist realism, but it does show that political practice is not quite ‘natural’ in the way that certain realists have taken it to be.

I want to emphasise that both concessive and naturalist realism as I define them here are no more than my best attempts to construct positions from a set of literature that seems – to me anyway – to be somewhat unclear and evasive on this topic. It may be that certain realists, even those whom I have labelled concessive or naturalist realists, may find my reconstructed positions unconvincing and may fully agree with my criticisms of them. If my contribution here is no more than a friendly amendment or clarification to existing accounts, then so be it.

This article proceeds as follows. In ‘Williams’s Theory of Legitimacy’ section, I briefly describe Williams’ theory of legitimacy. Since most theories of realist legitimacy take their cue from Williams, this is a helpful place to begin. ‘Concessive realism’ and ‘Naturalist realism’ sections summarise and criticise concessive and naturalist realism, respectively. In ‘What should an account of the normativity of realist legitimacy look like?’ section, I conclude by suggesting some takeaway points for constructing an account of the normativity in realist legitimacy, especially with regard to how a more plausible version of naturalist realism might proceed.

Williams’s Theory of Legitimacy

Williams’s theory of legitimacy has two components to it. First, political institutions must address ‘the first political question’ by providing order and stability. Second, they must ‘make sense’ to their citizens. Williams refers to this second requirement as the Basic Legitimation Demand (BLD). It is this second component that is the main source of attention and controversy in the now extensive literature on Williams’s political writings. This is unsurprising, since Williams’s imprecise formulation of the BLD raises a number of questions. For starters, it is not clear what it means for power to ‘make sense’. As Williams (2005: 23) acknowledges, ‘it says not much more than that coercion requires legitimation and that the will of the stronger is not itself a legitimation’. It is also unclear who or how many citizens Williams (2005: 11) believes the state must be able to ‘make sense’ to in order for it to satisfy the BLD, although it is clear that he thought that the BLD is usually best assessed in scalar terms and that he did not believe it requires legitimation either to all citizens or to all affected parties (Hall, 2015: 470–474). It is not necessary to address all of these questions here. The issue I want to address in this otherwise brief explanation of Williams’s theory is how he thought (and did not think) that his theory of legitimacy was normative.

Williams (2005: 11) claims that the BLD is an ‘evaluative’, not merely ‘descriptive’ concept. It seems that legitimacy is seen as a kind of evaluative standard internal to politics, like aesthetics is to art. But is it also normative? That is, does it also give some agent reasons for action? In general, Williams says, the answer is no. The exception, however, is ‘when we get to our own case’. By this, I think he means that when we judge that the BLD has been satisfied by the political institution governing us, this judgement has normative force. If our political institution ‘makes sense’ to us, then this seems to be an inherently normative judgement. But when we ask if another political institution ‘makes sense’ to us, this normative force is no longer present. This is not because we are unable to ask whether it ‘make sense’ to us, but because doing so no longer concerns its legitimacy. Since legitimacy is solely a matter of the relationship between a political institution and those whom it regards as citizens, the legitimacy of another political institution is not affected by whether it ‘makes sense’ to me, but by whether it ‘makes sense’ to its own citizens.

So if I want to know whether my political institution is legitimate, I should ask ‘does it make sense to me?’ My answer to this question will have normative force. But If I ask whether another political institution of which I am not a citizen ‘makes sense’ to me, my answer will have no such force. If I want to know whether such an institution is legitimate, I should ask ‘does it make sense to its citizens?’ The normative status of my answer to this question is somewhat trickier. On the one hand, there seems to be no contradiction in claiming that the political institution ‘makes sense’ to most or all of its citizens, but is still a very bad institution, and should be resisted or even overthrown (Cross, 2019: 14–16). On the other hand, Williams evidently believed that it was generally a good thing for political institutions to be legitimate, as Alice Baderin (2014: 140) points out. Hence, it seems plausible to infer that he would have also thought that we should at least sometimes act to help create and support other legitimate institutions and resist or overthrow illegitimate ones, the empirical complexities of such actions notwithstanding. Of course, these complexities are by no means trivial. Consider, for example, the kind of messiness involved in a decision to overthrow a foreign state on grounds of its illegitimacy – the 2003 U.S.-led Invasion of Iraq being a case in point. Still, I take it that it is a core tenet of Williams’s theory that judgements about the legitimacy of a political institution other than our own are at least relevant for making decisions about how to act in relation to that institution. ²

We may therefore say that Williams’s theory of legitimacy has normative force in two ways. First, when we judge whether our own political institution is legitimate or illegitimate by figuring out if it ‘makes sense’ to us, this judgement will have inevitable normative force for us. Second, when we judge whether another political institution is legitimate by figuring out whether it ‘makes sense’ to its citizens, an affirmative answer will give us at least some reason to work for its preservation, while a negative answer will give us at least some reason to work for its removal or replacement.

Concessive Realism

Concessive realism, I have said, focuses its attention on the Williams-premise: that the very practice of politics entails aiming for the provision of order and satisfaction of the BLD. But when it comes to explaining how realist legitimacy acquires its normative force, concessive realism adds a further claim: that the truth of the Williams-premise is sufficient to establish that realist legitimacy can lay claim to a kind of political normativity which is distinctive from the normativity of moralist theories.

Now I want to emphasise here that this position is one that I have tried to extrapolate from existing realist literature. To the best of my knowledge, it is not a position that has been explicitly defended by anyone. Nonetheless, I think it may be possible to attribute support for concessive realism to some of the literature without engaging in uncharitable caricature. Or least, I think there are prominent realist theorists who do not sufficiently distinguish their position from concessive realism, even if they recognise that they have reasons to do so.

In light of this, it seems that criticising concessive realism may be a less controversial task than attributing it to any one realist theorist. With that in mind, in what follows I will first explain why I think concessive realism gives an inadequate account of the normative force of realist legitimacy, before turning to the literature and pointing to instances where concessive realism has either been endorsed or tacitly hinted at.

Naturalist Realism

Naturalist realism does attempt to provide a realist account of the value of politics. It appears more explicitly determined to provide an account of the normativity of realist legitimacy that is shaped by the realist critique of morality. For this reason and others, I think it offers a much more compelling account of the normativity of realist legitimacy than concessive realism. In fact, I am willing to go so far as to suggest that there may be some version of naturalist realism capable of providing a very plausible account. I will say a little about this in ‘What should an account of the normativity of realist legitimacy look like?’ section. In this section, however, I will focus on the version of naturalist realism which has been discussed or alluded to so far in realist literature. This version, I argue, remains problematic, and cannot adequately show how realist legitimacy is normative.

Naturalist realism, like concessive realism, endorses the Williams-premise. Practicing politics well means practicing legitimate politics, just as playing football well means scoring goals without conceding them. But it adds a second premise, which I shall call the naturalist premise: avoiding politics is not a real option for anyone. This is because some form of political arrangement is necessary to secure certain basic goods that we all desire for ourselves. It might be theoretically possible for us to escape politics, but this is only in the sense that it might be theoretically possible to starve ourselves to death when food is readily available. In such a situation, norms will have little or no force or relevance. As Rossi (2010: 507) puts it, ‘the very nature of the human condition, rather than rational argument, necessitates political authority’.

There may still be an important empirical claim at work in the naturalist premise, namely, that politics is actually necessary for meeting basic human needs. Yet, a cursory glance at human history makes this claim very difficult to deny. As Horton (2012: 139–140) states,

Wherever people have flourished and prospered for any length of time . . . they have lived in political communities of one sort or other, which has inevitably involved what, in its broadest meaning, would count as political relations . . . There is, therefore, a perfectly straightforward and unobjectionable sense in which we can describe living in a political society as ‘natural’; and by this I mean no more than that we know of no viable long-term way in which human beings in general can live valuable and worthwhile lives, whatever exactly a worthwhile life is taken to be, outside of a political society.

I have no objection to the naturalist premise. In general, we stand little or no chance of meeting our basic needs without some kind of collective action. In this sense, Horton is right to say that politics is ‘natural’ for human beings. ⁶ However, it is important to recognise that, even if we grant the truth of both the Williams-premise and the naturalist premise, the conclusion which naturalist realism wants us to reach – that anything other than legitimate politics is not a real option for anyone – is yet to be established. In order to arrive at this conclusion, naturalist realism requires a third premise:

If practice P has certain evaluative standards S internal to it, then those who are engaged in P have reason to pursue S.

The third premise would then allow us to conclude that anything other than legitimate politics is not a real option for anyone. If we are engaged in political practice – and the naturalist premise tells us that we must be – then we have reason to pursue the evaluative standards internal to politics. The Williams-premise tells us that legitimacy is one such standard.

At first glance, the third premise seems plausible. However, I think we should reject it, and the case of politics offers a suitable counter-example. Illegitimate institutions, which fall sufficiently short of meeting the BLD in order for the binary judgement to become applicable, have existed throughout history and continue to exist in many places today. If the third premise is correct, then it is difficult to avoid the conclusion that the existence of such institutions is the result of overwhelming human irrationality. This is because it would imply that those engaged in political practice necessarily have reasons to pursue good politics. Insofar as legitimacy constitutes good politics as per the Williams-premise, it would follow that that every act that goes into their construction and defence is one of irrationality.

I don’t think this view can be sustained, at least according to a realist picture. To see why this is so, it would be helpful to compare the case of those who profit from illegitimate institutions with that of the naturalist premise. The latter, recall, holds that avoiding all forms of politics is not a real option for anyone because all human beings require some kind of collective action (i.e. politics) in order to meet their needs. Hence, the ‘us’ of the naturalist premise purports to be a universal ‘us’. If this is right, we all have what Williams (1981: chapter 8) calls an ‘internal reason’ to practice politics. An internal reason is one which in some way appeals to an agent’s existing motivational set (i.e. our desires, needs and/or interests), as opposed to an external reason which purports to apply regardless of the agent’s motivational set. If the naturalist premise is true, then we have an internal reason for engaging in politics, since politics is necessary to meet our needs. We might even say there is a sense in which the person who wishes to avoid politics can be ‘argued’ into it by presenting them with empirical evidence and appealing to their motivational set.

The same, however, cannot be said for those who profit from illegitimate institutions. In general, institutions which rely on widespread brute coercion are less likely to provide citizens with the benefits of politics (Prinz and Rossi, 2017: 339), that is, meeting their basic needs. But illegitimate institutions still usually function to the benefit of some people (i.e. their rulers and a select minority) but not others. In Horton’s words, the beneficiaries of illegitimate institutions can still have the opportunity to lead ‘valuable and worthwhile lives’ in the sense that they can at least count on having their basic needs met. Hence, we cannot similarly appeal to the desire of these fortunate few to satisfy their own needs in order to ‘argue’ them into supporting legitimate politics instead. To say that the mere ‘practice of politics’ rationally obligates them to pursue legitimate politics seems like a retreat back into the kind of externalism which Williams believed to be debunked by reflection.

It is worth emphasising here that realists have good reasons to be especially sceptical of externalism about reasons that extend beyond ‘because-Williams-said-so’ grounds. One reason for this is that externalism seems to invite the kind of wishful thinking about our ethical life to which the morality system is prone. The morality system’s spurious efforts to ‘recruit rationality to its cause’ (Owen, 2017: 81; Williams, 1981: 110) are often nonetheless appealing to us because they offer a kind of comfort. They allow us to think that our own way of life is somehow rationally vindicated, while those who do not share our way of life are defective reasoners. Another reason is that externalism can also have an obfuscating effect on human normativity. The only sorts of norms which realists should be open to endorsing are those which clearly connect with human needs and interests. As Raymond Geuss (2013: 92–93) suggests, a norm which diverts attention away from needs and interests, as naturalist realism (so understood) appears to do, will have a ‘stultifying effect on human thought and action’. ⁷ Such norms are ideologically suspicious since they are prone to manipulation by political agents who seek to get people to act against their own interests. Any realist theory worth the name should thus be opposed to this kind of theorising. ⁸

Naturalist realists might try responding to this objection in one of two ways. First, they might argue that supporters of illegitimate institutions are simply confused. They think they are pursuing ‘good politics’, but don’t realise that legitimacy is a requirement of ‘good politics’. Or they mistakenly believe they are pursuing legitimate politics. In either case, so the response goes, supporters of illegitimate institutions seem to have a reason to reorient their political actions towards (actually) legitimate politics once they realise their mistake.

Let us grant – probably very counterfactually – that supporters of illegitimate institutions really are just mistaken in one of these ways. On the picture painted by this response, political agents who have mistaken views of the requirements of ‘good politics’ already have a commitment to pursuing a non-descript notion of ‘good politics’ as part of their motivational set. If this is accurate, then it is this commitment, and not the mere fact that they are engaged in political practice, that generates a reason for them to realign their politics. If they lack this commitment, then it isn’t clear how any arguments over the requirements of ‘good politics’ or legitimacy will generate internal reasons for them.

Second, it might be argued that supporters of illegitimate institutions are not actually engaged in politics at all. The Williams-premise entails that illegitimate institutions actually ‘pervert politics’ (Hall, 2015: 478). Hence, one might claim that the case of illegitimate institutions does not offer a counter-example to the third premise. This is because we can grant that their supporters have no reasons to pursue good politics, while continuing to hold that all those who are engaged in politics proper do have such reasons.

This response may strike us as committing a kind of ‘no true Scotsman’ fallacy by gerrymandering the definition of politics in order to suit the purposes of an argument. ⁹ In any case, even if we grant that rulers of illegitimate institutions are not practicing politics at all, this saves the third premise only at the cost of falsifying the naturalist premise. Illegitimate institutions exist, and if their supporters are not practicing politics, then avoiding politics is clearly a real option for some.

We should therefore reject the third premise of the argument for naturalist realism. There are at least some people whose needs are met by illegitimate political institutions. Hence, there is a sense in which supporting illegitimate institutions is a real option for them.

I want to finish this section by reflecting briefly on how my objection to naturalist realism relates to Williams’s position. At first glance, it may seem like a departure. The third premise seems to reflect the kind of ‘practice-dependence’ approach to political philosophy, sometimes associated with realism, defended by Andrea Sangiovanni (2008). Williams might be read as endorsing this when he compares the BLD with Habermas’s political theory. He notes that the concept of legitimacy generates normativity in his theory in much the same way as Habermas’s discourse theory does in his own (Williams, 2005: 10).

On this basis, we might think that rejecting the third premise means rejecting Williams’s view. Yet, this may be disputed. Sangiovanni (2008: 162–163) seems to anticipate something like the objection I have sketched here, but claims that ‘the objection is not to the idea of practice-dependence, but to the idea that it is worthwhile to attempt a justification of a conception of justice for a specific institution in the first place’. This suggests that, for Sangiovanni, ‘practice-dependent norms’ are not simply determined by standards internal to a practice, but by standards internal to a practice that we have some good reason to engage in. The emphasis on the norms internal to the practice of politics can thus remain ‘a form of defetishizing, demystifying critique’ because they connect this practice with the needs and interests that we want politics to serve. I don’t see why Williams cannot say something similar for the BLD. Perhaps he would say that illegitimate political institutions ‘pervert politics’. However, he could add that what is objectionable about them is not that perverting politics is intrinsically bad, but that they deprive citizens of the vital goods that they want politics to provide them with. This does not push him back into Larmore’s position, according to which political normativity stems from an antecedent moral idea. It only acknowledges that we have certain human needs and interests that we want politics to help us meet.

What Should an Account of the Normativity of Realist Legitimacy Look Like?

I have argued that proponents of realist legitimacy should reject both concessive and naturalist realism as I have defined them here. They should reject concessive realism because they must recognise the relevance of the realist critique of the morality system and of ethical life more generally – something which concessive realism fails to do. Naturalist realism, at least in its above form, should be rejected because alternatives to legitimate politics may remain real options for some people, even if the Williams-premise is true.

I want to conclude by noting how these reasons for rejecting the above accounts of normativity in realist legitimacy point us towards certain desiderata which realists will need to observe when developing alternative accounts. The lessons to be learnt from the rejection of concessive realism are relatively straightforward. First, realists cannot confine their analysis of realist legitimacy to the defence and application of the Williams-premise. They need to say something about the value of political practice itself. Second, they cannot explain the value of political practice by assuming the question, ‘why is politics worth practicing?’ has an affirmative answer and delegating this task to morality. Any realist account of the value of political practice must avoid appealing to practices which cannot retain our confidence after we reflect on them. And third, Williams’s idea of ‘the morality system’ is an example of just such a practice. Perhaps there are certain types of morality that realists can consistently appeal to, but they must not be bound to the morality system.

The lessons to be drawn from the rejection of naturalist realism are more complicated. The main point to observe here is that realists need some way of reckoning with the fact that, as things currently stand, it seems that many people do not have an internal reason to practice politics. The idea that the Williams-premise and the naturalist premise do all the necessary work in establishing the normativity of realist legitimacy is clearly mistaken. However, this does not mean that they cannot do some of the work. I will therefore finish by sketching three non-mutually exclusive routes which an alternative account of the normativity of realist legitimacy, perhaps appropriately described as ‘naturalist’ in some sense, might take.

First, realists might adopt something of a Rortyan attitude towards concern for the needs of others. That is, they recognise that their own aversion to cruelty, wish to retain this aversion, and wish to promote it among those who do not currently share it. Yet, they might not think there is anything about this disposition that can be rationally vindicated. It is just a particular disposition they happen to share; nothing more.

Second, realists might argue that having concern for the basic needs of others is somehow an apt or fitting disposition which can serve as a normative value. The obvious difficulty here is to invoke an evaluative standard which not only isn’t bound to the morality system, but also does not fall afoul of Williams’s critique of practices. Realists might argue, for example, that psychopaths lack a ‘properly functioning empathetic engagement’ (Sagar, 2016: 378). A different, arguably more realist-friendly argument might involve employing Williams’s Critical Theory Principle (2005: 6) to help identify desires which have been distorted by power. Realists might hope that the Critical Theory Principle will show that those who have no concern for the needs of others have had their desires so distorted. But as Nietzsche argues in the Genealogy of Morality, it is also possible that it will show just the opposite.

Third, realists might argue that those who currently profit from illegitimate political institutions will ultimately recognise that such institutions are unstable or likely to be overthrown. Hence, their desire to meet their own needs will lead them to support legitimate institutions instead. Perhaps the clearest example of this is Marx and Engels’ (1998 (1848): 23) predication/observation that as the ruling class becomes able to foresee its impending demise at the hands of another class, its members start defecting to the soon-to-be hegemonic class. This approach might allow realists to account for the value of politics in a fully ‘naturalistic’, non-normative way. However, it might also require a longer view of history than those who want political transformation now and around here are willing to accept. Realists may want to have something to say to those who profit from illegitimate politics beyond that it will be in their interests to realign their politics at some unknown point in the future. Yet, perhaps this concession is itself appropriately realistic: legitimate political institutions might not be attainable until those who currently profit from illegitimate institutions either find it in their interest to cooperate with them or are otherwise subdued.