• Bernard Williams;
  • basic legitimation demand;
  • political realism;
  • political moralism


  1. Top of page
  2. Abstract
  3. The First Political Question
  4. Defending the BLD
  5. Conclusion
  6. References
  7. Biography

In this article I discuss Bernard Williams' realist conception of legitimacy. According to his critics Williams tacitly incorporates various moral claims, endorses a philosophically suspect ‘consensus’ view of politics, and employs an unrealistic and moralised conception of political rule. I argue that these criticisms mischaracterise the nature of the basic legitimation demand and the judgements about the acceptability of the state at its core and conclude that political theorists who object to the direction and style of much contemporary political theory should take seriously the possibility of developing an appropriately ‘political’ political theory on Williamsian lines.

A number of thinkers concerned by the style of much contemporary political theory have recently attacked the suggestion that political theory is a branch of applied ethics and gestured towards the possibility of developing a more self-consciously realist approach. Chief among these is Bernard Williams, whose posthumous political essays have been influential in setting the agenda of the burgeoning realist movement in political theory.1 The basic legitimation demand (BLD) is central to Williams' realism because he claims it offers a distinctively political standard of evaluation that avoids the moralism of much contemporary political theory. However, commentators have recently disputed this claim by arguing that Williams tacitly incorporates moral claims, which precludes him from offering a realist alternative to political moralism, commits to a philosophically suspect consensus view of politics and endorses an unrealistic and moralised conception of politics itself. Hence, one critic concludes that Williams' realism ‘shares significant features with liberal theory … which means that it is vulnerable to the same critique that other realists have made of liberalism’ (Sleat, 2010, p. 486) while another insists that it offers an ‘unfeasible return to an ideal-type expectation’ (Freeden, 2012, p. 6).

In this article I focus on four influential criticisms because by engaging with Williams' critics we can better appreciate the attractions of thinking about politics in the terms he encourages. I argue that Williams' critics mischaracterise the nature of the BLD and the judgements about the acceptability of the state at its core, and conclude that political theorists who are uneasy with the moralism of much contemporary political theory should re-evaluate the attractions of thinking about politics in Williams' terms.

The First Political Question

  1. Top of page
  2. Abstract
  3. The First Political Question
  4. Defending the BLD
  5. Conclusion
  6. References
  7. Biography

Williams claims that two forms of political moralism (PM) dominate political philosophy. Enactment models, like utilitarianism, formulate ‘principles, concepts, ideals, and values’ and seek to ‘express these in political action’, while structural models, like those of Rawls, spell out the ‘moral conditions of co-existence under power’. Despite their differences both models see political philosophy as a form of applied morality and accordingly represent the ‘priority of the moral over the political’ (Williams, 2005, p. 2). In contrast Williams (2005, p. 3) pursues a political realism that gives ‘a greater autonomy to distinctively political thought’. This is not Realpolitik. Rather, Williams argues that the normative standards employed in politics must be sensitive to the relationship between political practice and moral principle and that political philosophy should consequently ‘use distinctively political concepts, such as power, and its normative relative, legitimation’ (Williams, 2005, p. 77).

He identifies the first political question – first because solving it ‘is the condition of solving, indeed posing, any others’ – with the securing of ‘order, protection, safety, trust, and the conditions of cooperation’ (Williams, 2005, p. 3).2 It is a necessary condition of legitimacy (LEG) that the state solve this question, but unlike Hobbes, Williams insists that it is not sufficient because the point of politics is to save people from terror which he recognises states can also inflict. Any purported solution is subject to a basic legitimation demand (BLD) which is equated with the idea that it is an acceptable answer because:

if the power of one lot of people over another is to represent a solution to the first political question, and not itself be part of the problem, something has to be said to explain … what the difference is between the solution and the problem, and that cannot simply be an account of successful domination. It has to be a mode of justifying explanation or legitimation (Williams, 2005, p. 5, emphasis in original).

Hence, the need for a justification of political power arises when ‘A coerces B and claims that B would be wrong to fight back: resents it, forbids it, rallies others to oppose it as wrong’, because by doing this, ‘A claims that his actions transcend the conditions of warfare, and this gives rise to a demand of justification of what A does’ (Williams, 2005, p. 6).

Williams insists that the BLD does not represent a morality that is prior to politics as ‘it is a claim that is inherent in there being such a thing as politics … because it is inherent in there being a first political question’ (Williams, 2005, p. 5). For this reason, legitimacy and other political virtues are ‘different ideas’ and there ‘manifestly have been, and perhaps are, LEG non-liberal states’ (Williams, 2005, p. 4), because ‘a given historical structure can be … an example of the human capacity to live under an intelligible order of authority. It makes sense to us as such a structure’ (Williams, 2005, p. 10, emphasis in original). The crux of Williams' view is that political rule claims authority while brute force is mere coercion, with the important proviso that for political authority to be experienced as such the legitimation story must make sense (MS) to those citizens who are subjected to the coercive power of the state. This is why he insists that ‘one thing can be taken as an axiom, that might does not imply right, that … the power of coercion offered simply as the power of coercion cannot justify its own use’ (Williams, 2005, pp. 5–6). He asks us to consider a radically disadvantaged group of subjects who ‘are no better off than enemies of the state’ (he uses the historical example of the Helot population of Sparta who were openly treated as enemies by their rulers). In such scenarios we do not have ‘per se a political situation’ because ‘the mere circumstance of some subjects being de facto in the power of others is no legitimation of their being radically disadvantaged’, as there ‘is nothing to be said to this group to explain why they shouldn't revolt’ (Williams, 2005, p. 5).

Williams notes that we cannot precisely pinpoint when a genuine need for justification arises; it is not a sufficient condition that someone demands one ‘because anyone who feels he has grievance can raise a demand, and there is always some place for grievance’, but ‘it is also not a necessary condition, because people can be drilled by coercive power into accepting its exercise’ (Williams, 2005, p. 6). I return to the former point when discussing the ‘consensus’ critique of the BLD. The latter point introduces Williams' ‘critical theory principle’ (CTP), which holds that if the disadvantaged in a society accept a justification of power purely as a result of the exercise of power itself, the fact that they accept the story does not legitimate it.3

By focusing on the primacy of securing order Williams rejects ‘the basic relation of morality to politics as being that represented either by the enactment model or by the structural model’ (Williams, 2005, p. 8, emphasis in original). Political moralism pays insufficient attention to the centrality of answering the first question in realistic terms and, more often than not, forgets the contextual and historically conditioned nature of judgements about what makes sense. Williams insists that:

inasmuch as liberalism has foundations, it has foundations in its capacity to answer the ‘first question’ in what is now seen, granted these answers to the BLD, as an acceptable way … but this is not the foundation of the liberal state, because it is a product of those same forces that lead to a situation in which the BLD is satisfied only by a liberal state (Williams, 2005, p. 8).

He puts this most schematically when he writes that LEG + Modernity = Liberalism. ‘Now and around here’ we only permit liberal solutions because ‘other supposed legitimations are now seen to be false and in particular ideological’ (Williams, 2005, p. 8). This is markedly different from claiming that liberalism is the political expression of a set of timeless moral truths or that all previous legitimation stories were false. Williams accuses PM of forgetting this because it has an implausible understanding of ethics as a ‘mere moral normativity’, the result of the exercise of ahistorical reasoning. He holds that such views lack a theory of error that can explain ‘why what it takes to be the true moral solution to the questions of politics, liberalism, should for the first time (roughly) become evident in European culture from the seventeenth century onward, and why these truths have been concealed from other people’ (Williams, 2005, p. 9).4 Any fleshing out of what makes sense must be inherently local as it must satisfy the subjects to whom it applies in ways that will, in actuality and not merely in philosophical theory, ring true in the historical context in which they are made.5 When we ask what makes sense we consequently ‘need concepts and explanations which are rooted in our more local practices, our culture, and our history’ (Williams, 2006, pp. 186–7).

As answering the first political question is a matter of practical concern, legitimacy is not an abstract moral ideal but a live contextual possibility or, to put it another way, a reachable threshold. In contrast, political moralism tends to make very demanding claims of legitimacy. To employ a famous example, A. J. Simmons argues that a legitimate state must actually be consented to in quasi-Lockean terms, because of the ‘voluntarism’ that derives from his rights-based view, and infers that no existent states are legitimate.6 From the perspective Williams advocates this gets things the wrong way round. Rather than beginning with an antecedent moral view that is applied to politics, he begins by looking at the existent character of political rule and asks if we can extract an internal ethic from it, hence his suggestion that the BLD ‘is implicit in the very idea of a legitimate state, and so is inherent in any politics’ (Williams, 2005, p. 8). The basic idea is that politics contains its own internal legitimatising ‘ethic’ because the state demands a particular kind of allegiance from those agents it claims to incorporate as political subjects. Political moralism mistakes the contextual judgements about what makes sense with the basic conditions of satisfying the BLD. When the BLD is answered – which is to say, when the legitimation story is accepted in the relevant way (I discuss what this involves shortly) – rulers exercise political authority because they do not merely coerce or subordinate their subjects but act in a politically intelligible manner, and accordingly can be said to stand in a political relationship with their subjects in the way that, for example, a warlord does not.7

Mark Philp captures the underlying conception of politics at work when he writes that politics ‘involves at least some claim to authority … [while] brute force determines outcomes but it does so coercively, not authoritatively’, and it is therefore ‘integral to political rule to invoke at least some claim to authority and thereby to legitimacy … which implies some recognition of this on the part of citizens’ (Philp, 2007, pp. 55–6). This is an avowedly internalist conception of legitimacy, in the spirit of Hume and Weber, which holds that the conditions of legitimacy do not, in the first place, lie in the securing/respect of various moral principles, but in the opinion of the citizens over whom political power is exercised. Legitimacy is not achieved by enacting or respecting a set of external moral principles but is conferred by subjects. This is why Williams does not require an error theory, unlike the moralists who claim to enumerate the (allegedly) universal moral principles that ground the state's moral right to rule. Hence, the primary purpose of Williams' account is not to provide an alternative (albeit minimal) set of principles that ground a state's right to rule, but to enable us to understand the nature of politics itself. Of course, when subjects ask if a state satisfies the BLD, this is for them a normative question; as Williams puts it, when we ask what makes sense to us we think normatively ‘because what (most) makes sense to us is a structure of authority which we think we should accept’ (Williams, 2005, p. 11), but these judgements do not characterise the timeless conditions of legitimacy. We enjoy a period of stability that enables us to hold the state to a higher standard of acceptability than has been the case in the past. Accordingly, judgements about what makes sense will reflect a whole host of variables that cannot be captured by any specific set of moral ideals, which is why Williams' account is purposefully abstract and indeterminate.

Moreover, judgements about the acceptability of the state are not the same as pondering what a perfectly just society may look like. For Williams, justice, rights and liberty (and so on) are secondary political issues.8 When we ponder which normative goals we should pursue ‘now and around here’, we must remember that in many cases this is a question ‘that belongs to the level of fact, practice and politics, not one that lies beyond these in the very conditions of legitimacy’ (Williams, 2005, p. 17). This does not preclude us from claiming that certain features of existent states are objectionable – I may think that the United Kingdom's basic structure should be reformed so that it ameliorates the plight of the disadvantaged, or hold that certain decisions my government makes are morally indefensible (like the decision to go to war in Iraq) – but such judgements are not always judgements about legitimacy, even though on occasion they clearly are, because we can and do make such claims without thinking that, all things considered, the state is an unacceptable solution to the first question.9 To this extent, Williams is sensitive to the Hobbesian insight that we cannot judge the legitimacy of the state solely in light of our own (often deeply idiosyncratic) optimal moral judgements, because no functioning political state could exist if claims about its legitimacy had to cohere with all citizens' judgements about the optimal set of moral principles that should govern political life. Judgements about legitimacy are judgements to the tune that the state is, or is not, a realistically acceptable order of coercion that can secure the conditions of cooperation among groups of people who have disparate moral beliefs and conflicting interests.

Defending the BLD

  1. Top of page
  2. Abstract
  3. The First Political Question
  4. Defending the BLD
  5. Conclusion
  6. References
  7. Biography

Now that Williams' account has been outlined we are in a position to assess it. In the remainder of this article I focus on four lines of critique. The first alleges that the claim that the state must offer a justification of its power to each subject rests on a moral claim about the equal worth of persons which precludes Williams from offering a realist theory. The second criticism contends that he endorses an erroneous ‘consensus’ view of politics. The third asserts that our judgements about what passes the BLD must reference a set of normative claims held to be antecedently true to any claim of authority (which prohibits us from providing a categorical distinction between realism and moralism). The fourth criticism sees Williams employing an idealistic conception of politics itself. I argue that all four arguments fail because they mischaracterise the nature of the BLD and the view of political rule that motivates it.

The Scope of the Basic Legitimation Demand

Matt Sleat argues that Williams' insistence that the state has to offer a justification of its power to each subject ‘fall[s] back … upon some foundational moral premise that all persons matter’ and that this disqualifies him from articulating a cogent realism, because if we are to keep ‘political realism and political moralism distinct it has to be the case that it is possible to fully explicate politics and the necessary conditions of legitimacy without recourse to external moral conditions’ (Sleat, 2010, p. 495). At certain points this interpretation appears compelling. Williams (2005, p. 4) writes that a subject is ‘anyone who is in its [the state's] power, whom by its own lights it can rightfully coerce’, and that the state must offer a justification to each subject because if it does not ‘there will be people whom they are treating merely as enemies in the midst of their citizens, as the ancient Spartiates, consistently, treated the helots whom they had subjugated’ (Williams, 2005, p. 135). To this end, he claims that, ‘at least ideally’, the state must have something to say ‘to each person whom they constrain’ (Williams, 2005, p. 135).

Yet Sleat misreads Williams' intention as being to present a rival normative conception of the content of any successful legitimation, a central aspect of which is the belief that the state can only be legitimate if it offers a justification to each person for moral reasons. However, the most internally consistent reading reveals that the scope of the state's legitimation story is best understood as being offered to those persons considered to be citizens or political subjects, rather than simply those who are subjected to the state's power on other grounds (like the Helots), the point being that unless the state offers a justification to each person they constrain, the relationship between the rulers and those persons approximates internal warfare, not political authority, because the powerful exercise unlegitimated coercion. It does not follow that these regimes should thereby be classed illegitimate (in a binary sense) in the way many moralistic accounts suggest.

If Sleat is right, Williams would have to hold that Sparta was illegitimate because its treatment of the Helots did not respect the premise of basic moral equality that allegedly lies at the heart of the BLD. However, Williams explicitly does not suggest that we can judge Sparta illegitimate for this reason. He explicitly observes that the Helots were considered alien people, not political subjects, and states that it is only when there ‘is an attempt to incorporate’ the radically disadvantaged as political subjects that we can conclude that BLD ‘has not been met’ (Williams, 2005, p. 5). This idea of ‘incorporation’, and the subsequent claim to authority, is centrally important. He writes that:

There can be a case of internal warfare, of the kind invoked in the case of the Helots. There is no general answer to what are the boundaries of the state, and I suppose that there can in principle be a spongiform state. While there are no doubt reasons for stopping warfare, these are not the same reasons, or related to politics in the same way, as reasons given by a claim to authority. In terms of rights the situation is this: first anyone over whom the state claims authority has a right to treatment justified by the claim of LEG; second, there is no right to be a member of a state, if one is not a member … third, there is no claim of authority over enemies, including those in the situation of the Helots. In virtue of this last point, such people do not have a right of the kind mentioned in the first point … the significant cases for present problems are those in which the radically disadvantaged are said to be subjects and the state claims authority over them (Williams, 2005, p. 6).

Once we see the idea of ‘incorporation’ as framing the scope of the BLD we can distinguish between the citizens (or subjects) to whom a justification of power is offered (Spartans) and those people who may simply be subjected to coercion (like the Helots) to whom it is not. Williams is especially clear about this in ‘From Freedom to Liberty: The Construction of a Political Value’ when he writes that the legitimation story ‘is supposed to legitimate the arrangement to each citizen, that is to say, to each person from whom the state expects allegiance; though there may be other people within the state, slaves or captives, who are nakedly objects of coercion and for whom there is no such legitimation story’ (Williams, 2005, p. 95, emphasis in original).

With this in mind, there is little reason to hold that Williams' account is premised on the acceptance of some external moral principle of basic equality, because, as I have intimated, his basic aim is to delimit the nature of political authority, and the state need not stand in political relations with all of those persons whom it coerces. For this reason, he need not (and does not) invoke any necessary claims about the scope of the legitimation story because, as the quotation from his liberty paper shows, whom the state must justify itself to depends on those from whom it demands allegiance by incorporating them as political subjects. There is consequently no pre-political moral standard that determines who the BLD must be directed to and it is possible that certain groups will be coerced for reasons that only make sense to the constituency of persons to whom the state seeks to legitimate itself (i.e. the Spartan citizens rather than the Helots). In such a situation Williams helps us to see that political relations exist between the rulers and the subjects to whom their legitimation story makes sense even if there are other people to whom no justification is offered.

Williams thus uses a more restrictive conception of whom the state must legitimate itself to than is alleged by Sleat when he claims that Williams holds it as a necessary requirement of legitimacy that the BLD apply to all persons due to some tacit moral principle of equality. When Williams claims that, now and around here, all persons must be treated as political subjects (in a way the Helots were not), this is for reasons linked to the modern notion of the state as a legal-rational order. Hence, this is a historical development that relates to the disenchanted nature of modernity that Williams (following Weber) highlighted when he noted that traditional hierarchical justifications of inequitable treatment no longer makes sense.10 He is not claiming that all political orders throughout history should be judged legitimate or illegitimate according to this standard.

Sleat is motivated in part by Williams' claim that ‘slavery is imperfectly legitimated relative to a claim of authority over the slaves: it is a form of internalized warfare, as in the case of the Helots’ (Williams, 2005, p. 5), but he wrongly infers that Williams holds that all states containing slaves were for that reason illegitimate. This is the sort of fantastically unhistorical judgement that Williams seeks to avoid. The point of the slavery example is that slaves are not party to a political relationship, as a form of legitimation is an identifying category of politics (and they are not offered a legitimation that could make sense to them), not that all slave states were illegitimate. A state can be legitimate and contain people whom it does not treat in political terms provided its legitimation story makes sense to ‘a substantial number of the people’. In certain historical periods treating slaves, like the Helots, in this way may have made sense to the relevant constituency of citizens, even though ‘now and around here’ we do not think this.11

We can consequently see that Williams' account is not parasitic upon an external (moral) principle of basic equality once we recognise that the legitimation story is offered to each citizen, because it is the act of claiming authority that generates the need for justification. The idea is that not doing so would be a performative contradiction of some sort as the claim to authority – which is intrinsic to politics – implies some account of justification or explanation. If the state makes no such claim on certain people, by not incorporating them as citizens, the scope of the BLD is limited to those from whom it demands allegiance.12

The Consensus Critique

Critics also claim that Williams endorses an erroneous consensus view of politics. For instance, when he claims that LEG + Modernity = Liberalism, his critics allege that he is blind to the pervasive plurality of judgements about what makes sense now and around here. Jonathan Floyd (2011, p. 44) argues that there is no ‘pre-existing arrangement to be found about which forms of political authority are acceptable’, and contends that Williams' account fails because ‘the attempt to identify and then reconcile us to some putatively latent set of ideals in our local way of life is on a hiding to nothing on account of the very plurality of ideals which surrounds us’ (Floyd, 2011, p. 47). Michael Freeden makes the same point when he insists that the idea that a state could make sense to its subjects in the way Williams suggests ensures that the BLD offers an ‘unfeasible return to an ideal-type expectation’ (Freeden, 2012, p. 6). Likewise, Sleat insists that by ‘grounding his theory in the hope or actuality of agreement on … the conditions of legitimacy … Williams's theory becomes vulnerable to exactly the same challenge that other realists have posed to liberalism as a consensus-based theory’ (Sleat, 2010, p. 500).

However, these complaints trade on various misrepresentations. Sleat insists that Williams holds that judgements about the acceptability of the state will be unanimous when he writes that, for Williams, ‘universal acceptance is a necessary condition of legitimacy’ (Sleat, 2010, p. 496), and this also appears to be Freeden's complaint. Yet Williams states explicitly that having ‘something to say to each person’ does not imply that ‘this is something that this person or group will necessarily accept’ because there may be ‘anarchists, or utterly unreasonable people, or bandits, or merely enemies’ (Williams, 2005, pp. 135–6). In consequence, satisfying the BLD ‘does not coincide with this insatiable ideal of many a political theoretician: universal consent’ (Williams, 2005, p. 136, n. 8).

Who has to be satisfied by the BLD is a good question, and it depends on the circumstances. Moreover, it is a political question, which depends on political circumstances. Obviously, the people to be satisfied should include a substantial number of the people: beyond that, they may include other powers, groups … young people who need to understand what is happening, influential critics who need to be persuaded and so forth. If this position seems alarmingly relativist, it is important, indeed essential … to reflect that in the end no theorist has any way of advancing beyond it. He or she may invoke absolute or universal conditions of legitimacy, which any ‘reasonable’ person should accept; but in doing this, he or she speaks to an audience in a given situation, who share these conceptions of reasonableness (Williams, 2005, p. 136, emphasis in original).

Williams is therefore categorically not endorsing a ‘universal acceptance’ view according to which the legitimacy of the state depends on the unanimous acceptance of the legitimation story. Rather, if the state makes sense to a sufficient number of people (we cannot be more precise than that), the situation transcends the conditions of unmediated coercion in which politics is impossible. To this end, just as there is no timeless pre-political moral standard that determines to whom the state must try to legitimate itself (as I claimed), there is also no pre-political moral standard that determines how wide the acceptance of the legitimation story must be. It is possible and altogether likely that certain people will insist that the legitimation does not make sense and therefore refuse to recognise their relation to their rulers as being political in kind. Yet Williams recognises, and his critics forget, that it is often crude to make a binary distinction about legitimacy, which is why he emphasises that ‘the considerations that support LEG are scalar’ (Williams, 2005, p. 10).13 In some states it may be impossible to legitimate power to all and we may have to accept that some people are simply being subordinated (paramilitary Catholics in the recent past in Northern Ireland spring to mind). However, Williams helps us to see that politics is still occurring with respect to those to whom the order can be legitimated.14 Despite the inherently contextualist nature of judgements about who must be satisfied by the legitimation story, the difference between situations in which a story is offered and generally accepted, and those in which the powerful either fail to offer a justification at all or offer one that fails to make sense to their subjects, should in principle be clear.15 Hence John Horton is right to note that Williams does not hold that the claim to political authority is always ‘settled or uncontested’ but rather that ‘all parties that claim a relationship of political authority, rather than one of mere domination, must recognize the basic legitimation demand as something that needs to be addressed’ (Horton, 2012, p. 131).

Having disposed of this ‘universal-acceptance’ misreading, we are in a better position to assess the more charitable variant of the consensus critique. Floyd suggests that although modernity may rule out some legitimation stories it does not help us choose which ones we ought to accept; even if ‘modern populations do find the notion of the divine rights of kings unacceptable [this] does nothing to guide our choices between say, social democracy or neo-liberalism, luck-egalitarians and libertarians … and so on and so forth’ (Floyd, 2011, p. 46, emphasis in original). However, by repeating the moralistic mistake of conflating our judgements about optimal theories of justice with the idea that a state may be an acceptable solution, this misses the crux of what judgements about legitimacy concern. The appropriate criterion of making sense is not ‘Does this capture my favoured conception of justice?’, but (the inherently political question) ‘Is there an acceptable order of authority given that I must coexist with other citizens who have conflicting interests and different moral beliefs?’

With this in mind we can ask if is it fair to rebuke Williams for holding that ‘now and around here’ liberalism makes sense. It is worth bearing in mind two things when assessing this accusation: first, the importance of securing order and the conditions of cooperation; and second, Williams' capacious description of liberal societies as those that aim ‘to combine the rule of law with a liberty more extensive than in most earlier societies, a disposition to toleration, and a commitment to some kinds of equality’ (Williams, 2002, p. 264). Once we do so we ought to be less disposed to object to his claim. For one thing, throughout the Western world regimes that are loosely designated by the term ‘liberalism’ are alone capable of securing various economic goods, minimising military turmoil and ensuring that we can hold our political representatives to account. This is not a judgement about the ideal moral optimality of actually existing liberalism but a claim about the merits of realistically achievable competing ways of ordering our political institutions. Moreover, even though some people will deny that liberalism in Williamsian terms makes sense, if these complaints are to be politically convincing they must offer some reasons for thinking that viable alternatives exist that will be as good at ensuring order and the conditions of cooperation here and now. Although we cannot simply declare that this is impossible, Williams' view is buttressed by the fact that twentieth-century history shows us that alternative political creeds have on this score failed spectacularly.16

Once we grasp the force of this point we are in a better position to ask whether social democrats, neo-liberals and luck-egalitarians – or better, the citizens who endorse the commitments these theories rationalise – will disagree about the acceptability of the liberal state because they disagree about its justice optimality. This is clearly much less obvious than Floyd implies because even if some political philosophers, who link justice and legitimacy closely, may disagree about the acceptability of actually existing liberal states, it is likely that this shows us more about the esoteric and other-worldly nature of their disputes than about the opinion of most citizens. Most of the time most citizens in liberal states seemingly consider winning a democratic election, respecting the capacious liberal values that Williams focuses on and not violating some basic human rights, to be enough to confer legitimacy regardless of the content of their optimal moral theories of justice, to which their acquiescence attests.

Therefore, once we accept that Williams is not committed to thinking that every person (a) has to endorse liberalism as the ideal way to organise political society, or (b) must even consider it acceptable in the more minimal terms I have been focusing on, we can conclude that he does not endorse the unrealistic ‘consensus’ view his critics object to as he is merely committed to the view that liberalism (understood capaciously) will be considered acceptable by a sufficient number of people.17 Certain groups may think that ‘there is no legitimate government outside of their own creed, and that the liberal state makes no legitimate demand on them’ (Williams, 2005, p. 136), but Williams remarks that we often misrepresent the outlooks of such groups and depict them as consisting entirely of fanatics (he claims ‘this is a standard move, at the present time, in the demonization of Islam’). In these situations, we need to utilise a ‘realistic social understanding, a desire for cooperation if possible, and political intelligence’ (Williams, 2005, p. 137).

One way to avoid unrealistic understandings of what forms of political society people will find acceptable may be to move in a quasi-Humean direction and hold that judgements about acceptability are conditioned by our psychological propensities, so that the BLD can be satisfied if authority is exercised by rulers and acquiesced with by the population. (This makes sense of the remark that ‘it is obvious that in many states most of the time the question of legitimate authority can be sufficiently taken for granted for people to get on with other kinds of political agenda’ – Williams, 2005, p. 62.) If so, the purported problem about consensus may dissipate. In the Treatise Hume writes that he seeks:

Interest more immediately connected with government, and which may be at once the original motive to its institution, and the source of our obedience to it. This interest I find to consist in the security and protection, which we enjoy in political society, and which we can never attain, when perfectly free and independent. As interest, therefore, is the immediate sanction of government, the one can have no longer being than the other; and whenever the civil magistrate carries his oppression so far as to render his authority perfectly intolerable, we are no longer bound to submit to it. The cause ceases; the effect must cease also (Hume, 1985 [1739], pp. 601–2).

While I do not wish to deny that a thoroughgoing Humean account differs from Williams' on various fronts, the thrust of the above is compatible with Williams' view. For Williams, the state offers a solution to the first question, which enables us to pursue our interests. Williams claims (and Hume avers) that if the state ceases to secure the goods associated with it being a solution to the first question, it will no longer make sense to obey it. In contrast to what the consensus critics suggest, ‘now and around here’ there is reason to think that most subjects find the thin sort of liberalism Williams denotes acceptable in the sense he requires even if they disagree about its optimality. As Philp (2012, p. 634) notes, ‘while that opinion is not necessarily enthusiastic and positive, and for some groups is nothing more than a modus vivendi, nonetheless, the centre can and does hold in many orders, and does so with our collective concurrence (albeit motivated very differently for different groups)’.

The BLD and Its Normative Content

Charles Larmore contends that Williams fails to offer an alternative to political moralism. ‘It is not so much the BLD as rather the justification of state power, whatever it may be, in which satisfying the BLD is meant to consist’, Larmore maintains,

that must express a ‘morality prior to politics’: it has to embody an idea of what constitutes the just exercise of political order – specifically, an idea of what constitutes the just exercise of coercive power – and that is not only a moral conception but one whose validity must be understood as antecedent to the state's own authority by virtue of serving to ground it (Larmore, 2013, p. 291).

It is true that judgements about what makes sense are normative, but given that Williams acknowledges this – as I noted, ‘when we get to our own case, the notion “MS” does become normative, because what (most) MS to us is a structure of authority which we think we should accept’ (Williams, 2005, p. 11) – Larmore's suggestion that this somehow problematises Williams' conception of the BLD is unwarranted: this does not preclude Williams from reminding us of (a) the primacy of solving the first question, and (b) the unique character of life under political authority. Thus even though what makes sense to us is a normative question, this does not impugn Williams' attempt to offer an account of political rule and the ethic internal to it. To this end, Larmore muddles Williams' attempt to explain what must be in place for politics, as opposed to war, to be occurring, and the various normative judgements that we make about what makes sense to us.

This invites the possible second rejoinder to his complaint: namely, that speaking about the necessity of an antecedent conception of the just exercise of political power does not capture the nature of the sorts of judgements that actually play a role in judging purported answers to the first political question. As I have argued, judgements about what makes sense are far more nuanced than Larmore implies because they must be sensitive to the exigencies of real-world politics. To this end, the BLD fulfils the relevant criteria of realist as opposed to moralist political thought by giving ‘greater autonomy to distinctively political thought’ (Williams, 2005, p. 3) and by not conceiving of the ‘basic relation of morality to politics as being represented by either the enactment or structural model’ (Williams, 2005, p. 8, emphasis in original). Therefore, even though Williams' account references various moral/normative judgements this does not make the BLD a species of the ‘applied moral philosophy’ approach to political theory that he seeks to confute.

Moreover, the thrust of Larmore's line of critique neglects Williams' point that we should get beyond the idea that moral principles must not only underlie, but also in some way precede, political practice. This is particularly clear in Williams' papers on toleration where he maintains that ‘instead of trying to reach the politics of liberalism from a moral assumption that concerns toleration [namely, a view of autonomy], we should consider first the politics of liberalism, including its practices of toleration, and then ask, what, if any, kinds of moral assumption are related to that’ (Williams, 2005, p. 135). When we do so it is hard to ‘discover any one attitude that underlies liberal practice’ because toleration requires ‘social virtues such as the desire to cooperate and to get on peaceably with one's fellow citizens and a capacity for seeing how things look to them … some scepticism, the lack of fanatical conviction on religious issues [and so on]’ (Williams, 2005, p. 138). This reflects Williams' belief that when we think about legitimation in realistic terms we should recognise that people come to accept, in a less reflective manner than philosophers like Larmore acknowledge, certain kinds of arrangements which simply strike them as making sense for numerous reasons that may have very little to do with considerations of justice, and may indeed have little volitional quality (as they are in part given to us by history). In this sense Larmore fails to grasp the centrality of Williams' contention that in politics, as in much of life, we must remember the truth of Goethe's dictum: in the beginning was the deed.

The Unrealism of Williamsian Politics

I now move to the final line of complaint, the suggestion that Williams idealistically misunderstands politics. Michael Freeden remarks that Williams should not only be maligned for retaining the (unrealistic) liberal view that political rule should be understood in ‘terms of trust and cooperation, a theme quite central to Locke's and Rawls's versions of liberalism’, but also for failing to recognise that ‘legitimacy is not necessarily an attribute of all political arrangements, even if sought after by a large number of political actors and thinkers’ (Freeden, 2012, pp. 6–7). The worry here is that Williams' belief that there is something unique about the nature of political rule, given the peculiarity of its claim to authority and the concomitant demand for justification, is baseless because countless examples exist where rulers have routinely disregarded the ‘might is not right’ axiom at the core of the BLD and have failed to offer a justification of their power that makes sense to their subjects without failing to be ‘political’ in some sense.

Yet we can recognise this and resist the conclusion that Williams' account is accordingly flawed. Mark Philp's discussion of Nazi Germany – a regime that clearly falls into the category Freeden invokes – in Political Conduct is demonstrative in this respect. Philp claims that although it is absurd to deny that the Nazi regime acted politically in so far as it ‘sought to adapt and extend existing forms of political authority [and] relied on an array of traditional political institutions and mechanisms to achieve [its] ends’, it is equally hard to insist that it exercised political authority. For one thing, its actions led ‘to an increasingly distorted set of political ambitions and an increasingly coercive political regime’ which ensured that ‘the order retained … a political form but it was less concerned with securing its authority as opposed to establishing its domination’ (Philp, 2007, p. 71). Furthermore, the fact that the most abhorrent policies were carried out in secret shows that ‘the state could not legitimate its activities and would have forfeited its claim to a right to rule had its activities been made public’ (Philp, 2007, p. 72). Philp concludes that Nazi Germany was in effect ‘being run by a cabal within the state – an inner state that had no publicly legitimated … right to rule’, and that when ‘things move in this direction … it becomes increasingly incoherent to describe the relationship between the political order and its victims as political in character’ (Philp, 2007, pp. 72–3).

This reminder about the peculiar nature of regimes that fail to recognise something akin to the demands explicated by the BLD help us to see that although, qua Freeden, numerous malevolent regimes exist that appear to be ‘political’ in some sense – they use state institutions and so on – the relationship that they have to some of the people whom they coerce may not be political in kind. Once we accept this there is no reason to think that the existence of such regimes impugns Williams' attempt to delineate some central features of properly political relationships, as such examples can simply be seen to be deficient in this respect. Hence it is not problematic that some regimes clearly did (and do) not exercise legitimacy in Williams' sense, because when they act in this manner it is hard to describe the relationship that they have with their subjects as being political in kind.

To this end, Williams helps us to make sense of the ways in which regimes like Nazi Germany are not only morally abhorrent but can be said to pervert politics. The point is that even though political power is coercive not all coercion is political, in much the same way that even though war might be diplomacy by other means, war is not politics by other means. Hence the appropriate response to the reminder that some states violate the ‘might is not right’ axiom is not to conclude that Williams' account fails because ‘anything goes’ as far as political rule is concerned – which is something Freeden comes dangerously close to suggesting – but to recognise that such regimes do not exercise political authority. The very existence of such regimes does not therefore refute Williams' conception of legitimacy precisely because legitimacy is an evaluative standard which, as such, some coercive orders will fail to meet.


  1. Top of page
  2. Abstract
  3. The First Political Question
  4. Defending the BLD
  5. Conclusion
  6. References
  7. Biography

Throughout this article I have argued that Williams' critics fail to grasp the force of his account because they misunderstand the claims about the acceptability of the state at its heart, as well as the resulting conception of politics, which allows multiple but not infinite variety about which purported solutions can conceivably make sense as answers to the first political question. This enables Williams to avoid the dilemma of either over-specifying the political so as to impose haphazardly a set of liberal norms, or removing all content so that politics becomes compatible with any kind of coercion. At certain points his critics come alarmingly close to one or the other and mistakenly suppose that because the BLD has some normative content this implies some vestige of political moralism. In so doing, they implicitly affirm the categories of ‘politics as applied morality’ or ‘politics as bare force and coercion’ that he wants to transcend. I have defended the suggestion that the BLD is a standard of evaluation derived from the practice of politics, rather than an external moral standpoint, precisely because it is sensitive to the distinctively political demand for allegiance that characterises the relationship between the state and its subjects. Once we accept this, we should acknowledge that political theory cannot simply be an exercise in applying a set of principles, derived from an external moral standpoint, to politics. This is a compelling ‘realism’ because Williams does not deny that normative considerations have a place in politics but challenges the (typically unquestioned) authority of moral philosophy by reminding us that if political theory is to contribute to the ongoing task of making sense of the political situation in which we find ourselves, we must focus on the unique nature of political allegiance, and address the perennial first question in light of the unique historical circumstances in which the demand for legitimation arises.


I would like to thank Carlo Argenton, James Gledhill, Kajal Patel, Mark Philp and three anonymous reviewers for Political Studies for their insightful comments on this article. Countless discussions with Pietro Maffettone and Paul Sagar have greatly improved the final product (although the usual disclaimers apply) and I am very appreciative of their support. I am especially grateful to Geoff Hawthorn and Paul Kelly for their continued guidance and encouragement. Finally, I would like to acknowledge the Arts and Humanities Research Council for funding the doctorate during which this article was written.

  1. 1

    See Williams, 2005. Raymond Geuss has also been influential. He discusses Williams in Geuss (2005, pp. 219–33) but see Geuss (2008) for the most general statement of his realism. Stears (2007) and Galston (2010) introduce the realist/moralist debate while Flathman (2010), Forrester (2012) and Frazer (2010) examine Williams' contribution. Other important discussions include Cohen (2003), Estlund (2008, pp. 263–75), the essays collected in Floyd and Stears (2011) and the edition of the European Journal of Political Theory dedicated to ‘realism’ (North, 2010).

  2. 2

    Williams thus endorses a ‘normative’ conception of politics in so far as answering the first question enables us to enjoy the political goods he associates with so doing. This does not mean that Williams is tacitly a political moralist, because he does not prioritise a set of pre-political moral norms in the way the enactment and structural models do. The problem with political moralism is not that it sees politics as a normative enterprise (per se) but that it does not give autonomy to distinctively political thought.

  3. 3

    Williams does not maintain that citizens must accept a legitimation for the ‘right’ moral reasons, but endorses the idea of a modus vivendi solution (Williams, 2005, p. 2, n. 1). Regrettably for reasons of space I cannot discuss the CTP further but see Williams (2002, pp. 225–32) for his fullest account.

  4. 4

    See Tsai (2013) for a discussion of the theory of error argument. Realists typically overlook the fact that Williams does not consider Habermas to be a political moralist because he rejects the derivation of legitimacy ‘from the formal properties of the moral law, or from a Kantian account of the moral person’ by situating his account between facts and norms (Williams, 2005, pp. 9–10). However, see Williams (2005, pp. 14–7) for a discussion of some of the problems Habermasian views nonetheless face. Regrettably, a full discussion of the differences between their accounts is beyond the remit of this article.

  5. 5

    On this it is worth noting the relation with his work on internal reasons (Williams, 1981).

  6. 6

    Simmons subscribes ‘to political voluntarism as the correct account of these transactional grounds for legitimacy’ and holds that because ‘no actual states satisfy the requirements of this voluntarism … no existing states are legitimate’ (Simmons, 2005, p. 769). Although his distinction between justification and legitimacy makes his view more complex, his account remains unconvincing because legitimacy ceases to be a meaningful standard of evaluation as no political society has been, or will be, legitimate (it is like claiming that tallness is morally significant but that all men are short because they are not giants.) This does not accord with our considered use of the term: after all, it makes sense to hold that, for example, Assad's Syria is less legitimate than David Cameron's Great Britain.

  7. 7

    This kind of authority is an identifying criterion of politics through history (Williams, 2005, p. 69).

  8. 8

    Although for an engaging sceptical discussion of the relationship between legitimacy and justice in realism, see Runciman, 2012, pp. 66–8.

  9. 9

    For further discussion of this point see Horton, 2012, p. 135.

  10. 10

    As Freyenhagen notes (2011, p. 13), Williams holds that ‘in the modern world the questions of legitimacy and justification need to be answered in a way which addresses each citizen's reason and judgment’.

  11. 11

    This does not commit Williams to the view that slavery is morally acceptable: ‘crimes against stateless persons are surely crimes, and Helot-like slavery surely violates rights’ (Williams, 2005, p. 6).

  12. 12

    It might be thought that my ‘restrictive’ interpretation emphasises the more marginal papers in In the Beginning, at the cost of ‘Realism and Moralism in Political Theory’. The latter however was initially presented on 23 May 1996 in Cambridge under the title ‘Moralism and Realism in Liberal Politics’ (see, accessed 11 July 2012) and remained unpublished in Williams' lifetime. ‘From Freedom to Liberty’ was published in Philosophy and Public Affairs in 2001 (and is the most polished paper in the collection). Williams may have emphasised the more restrictive interpretation to dispel these interpretative problems.

  13. 13

    There is no reason to think that it is easy to make blanket judgements about legitimacy, and doing so will often be a matter of judgement that may depend on the perspective of the people doing the judging. Consider apartheid South Africa. Most bluntly we might say that if some white South Africans sincerely endorsed the legitimation story, there is reason to hold that they were in a political relationship with their rulers. But in regard to the non-white population, for whom the legitimation story did not make sense, political relations did not hold as they felt that they were merely subordinated. If, as I presume was the case, such people were not avowedly seen by the rulers to be enemies (like the Helots in Sparta) or slaves (as in ancient Athens) we can cogently hold that the state was illegitimate because the legitimation story failed to make sense to the vast majority of subjects. Moreover, as we – unlike the Spartans vis-à-vis the Helots – think (for complicated historical reasons to do with the disenchanted nature of modernity) that excluding groups for racist reasons does not make sense, as external observers we can conclude that politics is not occurring between the vast majority of persons and the state because they did not recognise the political authority of their rulers.

  14. 14

    We can develop the thrust of this if we consider how the coercion of those who deny that the legitimation story makes sense could be part of a strategy that seeks to secure a more wide-ranging legitimacy in the long run. Such coercion is different to the acts of a state that thinks that it can solve such problems by war or genocide. I am grateful to Mark Philp for stressing the importance of these implications of Williams' account. The question of how such people should be treated by the state prior to this remains and, as a reviewer has observed, this issue is worthy of further discussion in the wider realist literature.

  15. 15

    Thanks are due to Paul Sagar for discussing this point.

  16. 16

    Here I am indebted to Paul Sagar's illuminating discussion in his paper ‘From Scepticism to Liberalism’, which was originally delivered at the 2012 MANCEPT workshops in political theory. The point, as Williams notes, is that in many cases various utopian alternatives to liberalism ‘do not even reach the threshold of offering a serious political consideration’ (Williams, 2005, p. 92). For a contrasting view of the role of such historical considerations in political argument which misses the importance of such judgements, see Floyd 2009.

  17. 17

    Floyd also claims that Williams' view only makes sense if liberalism is ‘so broad as to radically under-determine our political options’ (Floyd, 2011, p. 45). But there is no reason to suppose that Williams holds that there will be consensus on what is required by the concrete instantiation of these thin liberal principles, because he never suggests that we can avoid debate and compromise and insists that we must reflectively consider how much philosophy can determine politically.


  1. Top of page
  2. Abstract
  3. The First Political Question
  4. Defending the BLD
  5. Conclusion
  6. References
  7. Biography
  • Cohen, G. A. (2003) ‘Facts and Principles’, Philosophy and Public Affairs, 31 (3), 211245.
  • Estlund, D. (2008) Democratic Authority: A Philosophical Framework. Princeton NJ: Princeton University Press.
  • Flathman, R. (2010) ‘In and Out of the Ethical: The Realist Liberalism of Bernard Williams’, Contemporary Political Theory, 9 (1), 7798.
  • Floyd, J. (2009) ‘Is Political Philosophy Too Ahistorical?’, Critical Review of International Social and Political Philosophy, 12 (4), 513533.
  • Floyd, J. (2011) ‘From Historical Contextualism, to Mentalism, to Behaviourism’, in J. Floyd and M. Stears (eds), Political Philosophy versus History? Contextualism and Real Politics in Contemporary Political Thought. Cambridge: Cambridge University Press, pp. 3865.
  • Floyd, J. and Stears, M. (eds) (2011) Political Theory versus History? Cambridge: Cambridge University Press.
  • Forrester, K. (2012) ‘Judith Shklar, Bernard Williams and Political Realism’, European Journal of Political Theory, 11 (3), 247272.
  • Frazer, E. (2010) ‘What's Real in Political Philosophy?’, Contemporary Political Theory, 9 (4), 490507.
  • Freeden, M. (2012) ‘Interpretative Realism and Prescriptive Realism’, Journal of Political Ideologies, 17 (1), 111.
  • Freyenhagen, F. (2011) ‘Taking Reasonable Pluralism Seriously: An Internal Critique of Political Liberalism’, Politics, Philosophy and Economics, 10 (3), 323342.
  • Galston, W. (2010) ‘Realism in Political Theory’, European Journal of Political Theory, 9 (4), 385411.
  • Geuss, R. (2005) Outside Ethics. Oxford: Princeton University Press.
  • Geuss, R. (2008) Philosophy and Real Politics. Oxford: Princeton University Press.
  • Horton, J. (2012) ‘Political Legitimacy, Justice, and Consent’, Critical Review of International Social and Political Philosophy, 15 (2), 129148.
  • Hume, D. (1985 [1739]) A Treatise of Human Nature. London: Penguin.
  • Larmore, C. (2013) ‘What is Political Philosophy?’, Journal of Moral Philosophy, 10 (3), 276306.
  • North, R. (ed.) (2010) European Journal of Political Theory, 9 (4), Special Issue, 379512.
  • Philp, M. (2007) Political Conduct. London: Harvard University Press.
  • Philp, M. (2012) ‘Realism without Illusions’, Political Theory, 40 (5), 629649.
  • Runciman, D. (2012) ‘What is Realistic Political Philosophy?’, Metaphilosophy, 43 (1), 5870.
  • Simmons, A. J. (1999) ‘Justification and Legitimacy’, Ethics, 109 (4), 739771.
  • Sleat, M. (2010) ‘Bernard Williams and the Possibility of a Realist Political Theory’, European Journal of Political Theory, 9 (4), 485503.
  • Stears, M. (2007) ‘Review Article: Liberalism and the Politics of Compulsion’, British Journal of Political Science, 37 (3), 533553.
  • Tsai, G. (2013) ‘An Error Theory for Liberal Universalism’, Journal of Political Philosophy, 21 (3), 305325.
  • Williams, B. (1981) ‘Internal and External Reasons’, in Moral Luck: Philosophical Papers 1973–1980. Cambridge: Cambridge University Press, pp. 101114.
  • Williams, B. (2002) Truth and Truthfulness. Oxford: Princeton University Press.
  • Williams, B. (2005) In the Beginning Was the Deed: Realism and Moralism in Political Argument, ed. G. Hawthorn. Oxford: Princeton University Press.
  • Williams, B. (2006) Philosophy as a Humanistic Discipline, ed. A. Moore. Oxford: Princeton University Press.


  1. Top of page
  2. Abstract
  3. The First Political Question
  4. Defending the BLD
  5. Conclusion
  6. References
  7. Biography
  • Edward Hall is a Fellow in Government at the London School of Economics and Political Science. His research interests lie in contemporary political theory and British political thought since the Second World War. He has published in Ethical Perspectives, Politics and Res Publica: A Journal of Moral, Legal and Social Philosophy. Edward Hall, Department of Government, London School of Economics and Political Science, Houghton Street, London WC2A 2AE, UK; email: