The culture of liberalism and the virtue of ‘balance’

Introduction

In this article I wish to defend what has become a minority position is contemporary liberal theory. Liberalism, I contend, needs an account of the ‘self’ if it is to remain normatively attractive in a modern, culturally diverse society. In making this claim I am challenging a school of post-Rawlsian liberal theory that sees the ‘self’ as a liability for a liberalism designed to appeal to a broad range of cultures and ‘comprehensive’ beliefs. The view here is that a rich account of the self (an account most often associated with Kantian autonomy in the liberal tradition) relies on controversial metaphysical assumptions that seem alien to most non-western moral and political traditions. Instead we are to rely on strictly political argument that appeals to the common interests of all citizens as reasonable political agents, all capable of reaching a ‘reasonable’ consensus on the basic institutional structure of society, regardless of their more ‘comprehensive’ moral beliefs.

My argument is that the denial of a deeper account of self is in fact a crucial mistake. Indeed, it is a rich view of the self that makes liberalism attractive as a normative political doctrine appealing not just to those born into the Enlightenment tradition, but also, potentially, to immigrants from very different traditions. Liberalism, I will argue, needs to appeal as a ‘culture’ rather than just a set of rights embedded in political institutions. The basis of this argument is the claim that, properly conceived, liberalism rests on a strong account of virtue. More specifically, the argument is that the normative value of liberal institutions ultimately reposes in an account of virtue and human flourishing. In arguing for this conclusion I am, of course, making a claim about which of the many strands of liberal thought properly belong in the liberal tradition. Later in this article I shall have more to say on this score.

But the key point of the first half of this article is that we need to recover this sense of virtue in the subtext of contemporary liberal theory, and, in so doing, reclaim the resources with which to defend a thicker version of ‘liberalism’ than that currently associated with the label. The thrust of my argument here is that this process of recovery is best begun by revisiting a particular debate in the recent history of political theory: the Walzerian attack on the (early) Rawlsian conception of self. It is not my intention to retread the well-worn ‘liberal–communitarian’ debate. Rather, my primary aim is to demonstrate that within this debate there is actually a shared account of virtue and, with it, a shared account of liberal culture that offers the most promising contemporary starting point for the normative justification of liberalism in a multicultural world.

Thus, a close reading of both (early) Rawls and Walzer reveals that the normative value of both their positions rests upon a strikingly similar conception of self. This conception expresses the deepest values of an earlier liberal tradition – the liberalism best expressed by Mill, and later by Hobson and Hobhouse – that has apparently been displaced in recent years.

This exercise in recovery is necessarily descriptive and historical. But it also feeds into the normative position that is developed at the end of this article. For the point there is that liberalism can offer up an account of self around which a diversity of different intellectual and moral traditions can coalesce. The tacit consensus of Rawls and Walzer is thus normatively important. For these are two theorists ostensibly at loggerheads and their actual convergence is a demonstration of the unifying power of a virtue-based account of self. For both thinkers, the value of liberalism, as distinct from the possibility of abstract liberal thought, lies in a very particular and deeply normative conception of a ‘balanced’ self. As we shall see, this notion of ‘balance’ expresses the ability of the individual to weigh competing ethical demands, in contrast to an unbalanced self that is dominated by a single end. Even if rationally chosen, this dominance of a single end constitutes a failure of Aristotelian virtue, even if it harms no one else. Rawlsian and Walzerian balance is, in short, a species of ‘perfectionism’; more specifically, in this context, a rich account of human flourishing.

What’s more, as I shall argue in the second half of this article, it is this normative account of human flourishing that we need – over and above an account of instrumental utility and mutual advantage – in order to justify liberal principles and institutions to an increasing number of doubters; not only amongst some immigrant minorities, but also amongst those born into ostensibly liberal cultures. First, however, I want to uncover the deeper normative sources of liberalism in both Walzer and Rawls.

The context of the argument: ‘political philosophy’ and ‘social criticism’

Whilst there is no need to revisit the familiar terms of the ‘liberal–communitarian’ debate in any detail, some brief comments are in order before turning to the main thrust of my argument – that Walzer and Rawls share a perfectionist account of ‘balance’.

It should be uncontroversial to assert that Walzer objects to an apparently bifurcated account of the self in early Rawls, on the grounds that Kantian ‘political philosophy’ forces us to repudiate the thick, but parochial, values and attachments that give life meaning, thereby denuding the self of its essential (and flawed) humanity. ¹ The classically communitarian corrective to the bifurcating effects of Kantian reasoning is to embrace the ‘passions’ (in essence our parochial moral attachments) rather than leaving them behind in a heroic bid to attain impartial objectivity. This is the basis of Walzer’s ‘connected’ social criticism, and also a central component of his account of the self. I shall refer to this sense of an un-bifurcated self as the ‘whole’ self.

At times I will also refer to this sense of self as the ‘democratic’ or ‘deep democratic’ self, by which I intend to indicate the sense in which the analogy is with a rich form of reasoned, deliberative democracy, rather than with the type of democracy that simply seeks the aggregation of votes or preferences. This categorization of self is also a reflection of another aspect of Walzer’s objection to ‘political philosophy’. Because it privileges the elite view of apparently objective reason over local meanings and traditions, political philosophy is for Walzer profoundly undemocratic, in the literal sense that it limits popular access to political debate and decision-making.

But the contrast between political philosophy and democracy also has a more figurative role. Political philosophy as a way of thinking has an impact on the proper constitution of the ‘whole’ self because it denies the individual a range of moral sources, in a way that is analogous with the more literal institutional exclusion of popular opinion that Walzer says is the practical consequence of elite (‘philosophical’) dominance of the body politic.

‘Democracy’ therefore figures in this article not just as a type of political and social structure, but also as an important analogy, describing not just the proper structure of society but also the proper constitution of the self. Thus, in a discussion of the effects of a hierarchical, Kantian view of the self, Walzer tells us that ‘[t]he hierarchical view requires a thick, pluralist and democratic correction’ (my emphasis). ² This, I contend, is the basis of Walzer’s account of virtue and ‘balance’, to which I now turn.

The psychology of balance in Walzer and Rawls

I shall look at Walzer’s need for a deep (Rawlsian) self in the following section. In making the case for this need the point is to demonstrate a more general need for such a deep self in a properly normative and coherent liberal theory. My purpose now, however, is to demonstrate the striking similarity that Walzer and Rawls share when it comes to the virtue of ‘balance’. As we are about to see, the notion of psychological and moral balance is central to both Walzer and Rawls. But this balance does not imply that all competing demands are accorded the same status, as this would effectively rule out a process of rational decision-making. So we must allow for a degree of internal hierarchy if the self is not to be overdetermined by its social context, and this dual commitment to both balance and hierarchy (a necessary prerequisite of moral freedom) is precisely what we find in these two writers. The normative upshot of this comparison is that the extent to which two such apparently opposing views (Rawls’s liberalism and Walzer’s communitarianism) converge upon the same conception of balance as an individual virtue is a promising indicator of the broader appeal of this account to very different normative traditions.

Thus, turning to Walzer first, the idea of a flourishing, ‘democratic’ and creative view of the self has been present throughout his work, and emerges explicitly when he says, in Spheres of Justice, that we are all each other’s equals by virtue of being culture-creating beings. ³ We also see the notion of balance very clearly in Thick and Thin. Here he writes about precisely the issue that I have just highlighted. His concern is that the social critic comes to be dominated by the demands of just one sphere. The key examples of this internal domination are the two ‘totalizing’ views of the individuals who identify their sources of value either exclusively with the ‘market’ or the ‘political’ sphere, allowing these values to suffuse everything they do. ⁴ These individuals are oppressed by their own narrow and obsessive identification, internally dominated by an ‘autocratic’ ‘tyranny’. It is this latter failing, a pathological obsession with a single source of value, which is now doing the work in Walzer’s a priori conception of self, a self that should be diverse, thick and balanced. The contrast is with an unbalanced and ‘totalizing’ self, stunted and ‘thin’, emotionally narrow and dominated by a ‘terrible one-sidedness’. ⁵ In terms of an overtly normative political analogy, the self that is dominated by one value is ‘autocratic’ and ‘tyrannical’; but the differentiated (plural) self is ‘democratic’. ⁶ Hence, as we have seen, ‘[t]he hierarchical view requires a thick, pluralist and democratic correction’ (my emphasis). ⁷

And this, I contend, is a ‘democracy’ that requires the control of a deep self, lest it descend into an unregulated clamour and anarchy in which the loudest voice at any given time holds sway, providing neither freedom nor balance.

The most immediate and obvious comparison here is between Walzer’s deep self – the sense of self needed to adjudicate between the conflicting demands of justice and identity that spheres force upon us – and the ‘self’ that we find in A Theory of Justice; a self that is, notoriously, said to be ‘prior to the ends that are affirmed by it’. ⁸ We can see this most clearly by first looking at the psychology of the ‘normal’ Walzerian and Rawlsian individual. Take the following juxtaposition:

[Walzer:] Plutocrats and meritocrats … are tyrants as much as autocrats are, and their personalities are distorted in comparable ways. In all three cases the self is dominated by a single set of interests and qualities. ⁹

[Rawls:] Surely the preference for a certain attribute or feeling or sensation above all else is as unbalanced and inhuman as an overriding desire to maximise one’s power over others or one’s material wealth. ¹⁰

The central message of this principle is that we should, wherever possible, pursue the short-term course of action that is most compatible with the aims of our overall ‘life-plan’. Thus, following Rawls’s own example, the trip to Europe that allows us to fulfil both a love of art and a love of history will be preferable to one that only fulfils one of these desires. ¹⁴ Clearly, the wider point here is precisely that a rich life-plan should not be dominated by a single end (whether that be the pursuit of high art or football). A worthwhile and rich life for Rawls is one that is lived with reference to a variety of meaningful ends. Indeed, without this assumption there would be no pressing need to invoke the inclusiveness principle as an end in itself. It would be no more than a practical recommendation concerning means and ends, a hypothetical imperative simply telling us that if we have a variety of ends then we would be rational to adopt the inclusiveness principle. But Rawls has to mean more than this, otherwise we can’t make sense of the normative judgement that the pursuit of a single dominant end ‘disfigures’ the self. As he knows full well, this pursuit would not be ‘irrational’ in terms of any instrumental understanding of that category. It would certainly not be irrational under the terms of modern rational-choice assumptions that have sometimes been attributed to the early Rawls. What’s really doing the work here is a value-laden conception of ‘balance’.

The Aristotelian principle: balance as a human excellence

This conclusion is buttressed by the ‘Aristotelian Principle’. This principle bears special emphasis because it draws out the sense in which the early Rawls was explicitly committed to an account of virtue that runs deeper than the rational choice strand of thought in A Theory of Justice. The full discussion of this principle is in fact flagged by an earlier discussion of it, some ten pages before the section dedicated to the principle in the text. In many ways, the earlier comments are clearer than the fuller discussion, and they are worth quoting at length. Thus:

I assume that human beings have a higher order desire to follow the principle of inclusiveness. They prefer the more comprehensive long-term plan because its execution presumably involves a more complex combination of abilities. The Aristotelian Principle states that other things being equal human beings enjoy the exercise of their realized capacities … and that this enjoyment increases the more capacity is realized, or the greater its complexity. [My emphasis.] ¹⁵

So the crucial point in terms of Rawls’s commitment to ‘balance’ only lurks in the subtext here, obscured by the distraction of a cruder perfectionism. The deeper and more sophisticated perfectionist assumption comes in the earlier passage’s discussion of the inclusiveness principle, when the language slips from short-term plans to the idea of a ‘comprehensive long-term plan’. In this locution, it is clear that the inclusiveness principle embraces far more than is suggested by the model of instrumental rationality: it recommends that a life properly lived is one that embraces a variety of meaningful (and often competing) ends. And it is also clear that, when combined with the ‘principle of inclusion’, the Aristotelian Principle embraces more than the intuition that some forms of human activity are intrinsically more valuable than others. It recommends, in short, a life of ‘balance’. For balance itself is, clearly, one of the human excellences that we are to pursue.

For both Walzer and Rawls, then, it is important that our understanding of a properly constituted self should draw heavily on the notion of balance. Of course, this similarity alone cannot bridge the gulf that has come to separate communitarian and Rawlsian understandings of what actually constitutes the self. In suggesting that Walzer and Rawls share common ground in the psychological notion of ‘balance’ I may indeed run the risk of committing the familiar error of quoting Rawls too selectively. Nevertheless, there is further common ground between the two theorists, and in the following section I hope to demonstrate just how deep this commonality is. The analysis now rests on a fundamental conceptual distinction that, I argue, neither Walzer nor Rawls (including Rawls in his later turn to political liberalism) can avoid if they are to offer us coherent accounts of moral reasoning. This is the distinction that Michael Sandel claims must be inherent in Rawls’s position; an (in his view) false distinction between the ‘values I have’ and ‘who I am’. ¹⁷ In utilizing this distinction we shall also see the full extent to which Walzer must rely upon the deep self in a way that brings him far closer to Rawls’s ‘political philosophy’ than he would care to admit.

The values ‘I have’ and the person ‘I am’

In this final section of exegesis I will take the analysis of Walzer’s ‘self’ a level deeper. The purpose is twofold. The first is to cement the conclusion that Walzer –despite being a communitarian committed to the position that the self is constituted by its social and historical context – must cleave to a sense of agency that has universal implications in the same way that the Rawlsian conception does. The ‘must’ here is important. For my second aim is to suggest that the account of balance derived from Walzer and Rawls is indeed a necessary feature of all coherent moral reasoning, and potentially a common point of reference in all of the great moral traditions.

This is a controversial claim and I cannot hope to fully justify it here. But a great deal of my purpose is achieved through the less ambitious aim of establishing the deep convergence of Walzer and Rawls. That would at least establish, to a significant degree, the de facto hegemony of this view of rationality and virtue in the deepest traditions of western liberal societies. It also leaves open a less strictly philosophical and more interpretative means of arriving at the same conclusion. This process would work by immersion in different cultures and traditions with the aim of finding the same (tacit or explicit) account of balance, for example, in contemporary Islamic scholarship. I shall return to this shortly.

Now to the business at hand – Sandel’s distinction between the values ‘I have’ and the person ‘I am’. Sandel’s purpose in employing this distinction is to demonstrate that it splits the self in two and denudes the agent of any real identity. The conclusion, by implication, is that Rawls’s Kantian inspired account of moral reasoning is morally and philosophically incoherent. As we have seen, Walzer comes to a very similar conclusion. Yet it transpires that Walzer’s account of the deep self displays exactly the same structure that Sandel attributes to Rawls. So if the distinction is true of Rawls (as it is) it should be equally true of Walzer.

We can see the values ‘I have’/person ‘I am’ distinction very clearly in the structure of Walzer’s ‘democratic’ self, and also in the parallel that he draws between social criticism and self-criticism. ¹⁸ Just as the social critic must be aware of the competing voices and values of a plural moral and social world, so must the deep self listen to its ‘many internal critics’. ¹⁹ One might indeed suppose that to silence them is to commit the same error as ‘political philosophy’; and this is, in fact, what Walzer has in mind. It is a normative (and personal) mistake to silence our many critics and to impose a universal and dominant value on ourselves. To do so clearly offends Walzer’s notion of psychological and emotional balance.

But how do we hear these competing critical voices in any intelligible way? How do we order and control them so that they are not just a cacophony of competing demands? As I argued in the previous section, there must be some form of underlying self that is capable of ordering and ranking the values that these voices represent. In terms of Walzer’s central analogy, it seems very clear that the ‘democratic’ self (plural and non-hierarchical) must have some form of legislature, some mechanism of control, if it is not to descend into anarchy. What Walzer really wants is a self that allows room for a variety of competing internal voices, tolerating their diversity and giving them all a fair hearing. This is the function that is performed by the deep self, a self that encompasses and orders the competing ‘democratic’ demands that are made on it. All well and good. But it becomes increasingly difficult to see how this deep, ‘democratic’ self differs in function from the ‘philosophical’ Rawlsian self. Where the Rawlsian self must be ‘antecedently individuated’ prior to the ‘ends it affirms’, Walzer’s self must surely be prior to the voices that compete for its attention. There must, as it were, be ‘somebody’ who listens; just as there must, on the Rawlsian model, be a self that ‘chooses’. Even this small difference breaks down when we begin to ask what it is that these voices articulate. These voices surely cannot come out of the blue or arise spontaneously from the individual mind, not unless Walzer wants to sign up to the most peculiar idealism in which the self creates its own reality from scratch. So how could there be ‘competing’ voices if they did not recommend potentially conflicting courses of action or forms of life; if they did not, in other words, force upon us the necessity for ‘choice’? It is in fact very difficult to see how this competition takes place without reference to exactly the same kind of ends and life-plans that the Rawlsian self affirms or rejects. The only discernible difference at this stage is in the mediating image of Walzer’s ‘voices’.

The important point here is that Walzer’s democratic (and ‘canny’) self also fits, rather neatly, Sandel’s distinction between the values I have and the person I am. The person I am can only be the underlying self that listens to the voices and the values I have must be those very same voices; values, that is, that are mediated by those ‘voices’ that can only have any sense at all if they are construed as the articulation or reflection of ‘values’.

Somewhat ironically, Walzer, a fully paid up proponent of the social thesis, states the quandary that this creates better than Rawls: ‘We need to think critically about the parts we play and the identities we affirm. But how can we do this when these same parts and identities are constitutive of the self that does the thinking?’ (my emphasis). ²⁰ One possible answer would be to adopt a rather weak conception of ‘thinking’ or rationality. Rationality on this view need not be the function of the kind of ‘deep’ self that I have attributed ‘thinking’ to. It need be no more than a broadly Humean instrumental rationality, simply the cognitive function that pursues the best means towards pre-given ends. This is the kind of rationality that has now become familiar to social scientists in the assumptions and doctrines of rational choice theory. Whilst Hume attributes these ends to desires rather than to reason, Walzer could also conform to the general structure of the Humean account by attributing our ends to the feelings that are imbued by inherited culture and socialization. This ties into Walzer’s view of the self as ‘connected’ and ‘thick’, embedded in the values and conceptualizations of its cultural milieu. The Humean account, moreover, as many will be quick to point out, has the added virtue of theoretical economy. For a start, it doesn’t require an elaborate theory of what the good is, as this is simply reduced to desires (or the desires derived from cultural location in Walzer’s case), or to ‘preferences’ in the rational choice idiom. Furthermore, the Humean model does not require a ‘deep’ self (the self is no more than a bundle of experiences) and thus neatly avoids the nebulous issue of what exactly the deep self ‘is’.

However, although this limited conception of rationality would suit Walzer’s purposes quite well in so far as it is compatible with the social thesis (we are nothing deeper than the values and norms that constitute us), it is far more problematic for his conception of the deep democratic self. In fact, the Humean or rational choice route is not really open to Walzer once we consider the role that ‘reason’ must play within the democratic self. What Walzer envisions here is to be understood in the terms of reasoned democratic discourse. The democratic self is not to be construed as the arena of competing demands that are only regulated by the populism of plebiscite, simply giving way to the most keenly felt value or desire. So rationality in this context cannot be merely instrumental horse-trading between competing desires and values. It has to be more than this for the very good reason that he wants his ‘self’ to be able not just to order the parts that he plays, but also to ‘think critically’ (not ‘instrumentally’) about them. In other words, the ‘ordering’ of our desires and values requires more than a simple calculation of the strategy that will allow us to pursue as many of our ends as possible. This ordering is of course a central part of all practical reasoning. We must, after all, decide which impulse or desire to act on. But this does not mean that we have to regard reason as, in Hume’s famous phrase, a ‘slave to the passions’. Our rational calculations require us not just to listen to the passions before servicing the needs of the most vocal. Listening alone is not sufficient for (deliberative) democratic action. That requires more than instrumental rationality if it is to perform the role that Walzer requires of it.

This need for something more is what Walzer is really getting at when he describes the pursuit of a single sphere of value as ‘fanatical’. Likewise for Rawls: for the self to be balanced and not ‘disfigured’, the individual must actively weigh competing values and objectives, thinking critically rather than instrumentally. To do this well is a challenge – and the exercise of a virtue.

Balance as a virtue

This does not yet fully settle the question of balance as a distinctive virtue. I hope to have established that it is the central virtue within both Walzer and Rawls, but this leaves open the question of whether or not this reading adds to the desirability of either account as a political or moral doctrine. My argument in this section is that ‘balance’ does offer a distinctive and normatively attractive approach to a range of challenges faced by modern liberal theory and practice, not least the problem of justifying liberal institutions in ever more plural and ethnically diverse societies.

One obvious area for exploration is the relationship between balance and the virtue of tolerance. This is a typically liberal virtue that allows the coexistence (within certain much discussed limits) of competing conceptions of the good. When the virtue of balance is displayed through the critical thinking of the democratic self, there is clearly a close affinity between ‘balance’ and ‘toleration’. Both concepts map onto individual characteristics that help maintain liberal societies, notably respect for the beliefs and practices of others. It would be a mistake, however, to think that the two concepts are interchangeable, or that ‘balance’ is just a redescription of toleration. Certainly, the democratic psychology I have described looks, at first sight, like it might be an account of toleration. Balance requires engagement with, and respect for, a range of values, giving them all a fair hearing. But in fact this is a far more stringent and demanding process than ‘toleration’ itself requires. As Joseph Raz has observed, toleration is much less demanding. ²¹ It need not require either respect or engagement with other values or ways of life; grudging acceptance and non-interference is enough.

A return to the analogy with the democratic self illustrates another important contrast between balance and toleration. A tolerant individual’s grudging acceptance of someone else’s values points to the limited psychological demands that the requirement of toleration places on an individual. It is highly significant that there is nothing in this virtue to rule out the ‘terrible one-sidedness’ that Walzer laments and which Rawls thinks ‘disfigures’ the self. The tolerant individual could be dominated by a single value and an obsessive pursuit of one goal, and a liberalism based just on toleration would have nothing to say about this, as long as there was no actual attempt to interfere in the lives of others. Of course, some would welcome this limited liberalism, seeing it as a pragmatic ‘modus vivendi’ – the great historical achievement of ‘Reformation’ liberalism in bringing to an end generations of religious conflict. It is not, however, the liberalism of either Walzer or Rawls and, as I shall shortly argue, it only offers weak normative support for modern liberal institutions.

In contrast to the virtue of toleration, balance positively requires engagement with conflicting sources of value if we are to see it as an exercise of human excellence. This is not to say that this need be a constant exercise. At times balance will be an exercise in reasoning, when, to paraphrase Walzer, we need to ‘think critically’ about the plural values and identities we affirm. Indeed, the demonstration of this crucial aspect of the virtue of balance has been the central purpose of this article.

But balance is also a more everyday attribute and there is a more prosaic side to the virtue. It also reflects the important ability to internalize and live by values that may have a strong tendency to come into conflict, either because of a specific social context (e.g. modern pressure on women to both raise a family and pursue a successful career), or because of the inherent incommensurability of certain values. I shall have more to say about the relationship between value pluralism and balance in a moment. The immediate point is that the virtue of balance is not just a rational virtue of reflection or critical engagement. It can also be exemplified by the individual who embraces the ‘passions’ – the sentiment and parochial attachments that Rawlsian reasoning is typically seen as repudiating.

Thus, the ‘passions’ do not conflict with the idea and virtue of balance because the virtue is also displayed in the ability to cope with inevitable moral tension with sophistication and good grace. Indeed, I take this to be the fundamental point of value pluralism: thinking critically can only ever take us so far (Sartre’s example of a ‘tragic choice’ – of the resistance fighter torn between caring for his sick mother and fighting against fascism – helps to illustrate this in an extreme form) and there will always be an important role for intuition, feeling and tradition when we run up against these limits. When this happens the virtue of balance looks more like ‘character’ than an excellence of human rationality and discourse. But this distinction does not undermine the power of ‘balance’ as a normative category. There is no ‘either/or’ choice here. On the contrary, it has the merit of widening its normative scope, enabling it to embrace both reason and passion, and to escape the constraints imposed by this dichotomy.

Moreover, it is important to note that it is these two behaviours – the ability to first reason ‘democratically’ and to accept remaining moral tension with good grace – rather than the values that are to be brought into the balance that are doing the normative work here. If balance were merely a formal idea we would probably say that the ethical relevance of the account I have developed here lies not in balance itself, but in the virtues or values that are being moderated. But the point of the analogy between the ‘deep democratic’ self and a rich form of reasoned and deliberative democracy is that the process itself – over and above the outcome – is of ethical significance. It is the process of achieving balance that is the primary exercise of virtue and expression of human flourishing. Clearly, there is a crucial role for secondary virtues, by which I mean the particular virtues associated with specific practices and values; for example, the virtues of honesty and reciprocity in the use of shared resources. But the key point here is that we must understand balance as more than a formal concept – as more than just an abstract description of the form of reasoning rather than the content. ‘Balance’, in short, is working independently of the values that it moderates (just as it works independently of ‘toleration’).

Finally, I will close this section with some brief comments on the independence of balance from the theory of value pluralism. There is an obvious and important connection here: there would be no need to posit the ideal of balance if one did not believe in the truth of value pluralism. However, one could be a value pluralist and remain entirely agnostic with regard to the value of balance itself. Value pluralism can remain at the level of phenomenological description of the conditions we need to exercise the virtue of balance, whereas ‘balance’ describes the virtue itself. Balance, as a description of our reaction to the fact of value pluralism (praiseworthy or not), is the more powerful normative concept – and one that should take its rightful place in the liberal tradition.

Balance within the liberal tradition

That place lies somewhere within the broad grouping of ‘Enlightenment’ liberalism. The ‘deep democratic’ self bears a strong resemblance to ‘comprehensive’ liberalisms based on the value of rational autonomy as a human excellence. Given what I have said about the perfectionist assumptions of Walzer and Rawls, this should come as no surprise. It is worth stressing, however, that this rules out certain approaches to liberalism as a political doctrine. In particular, ‘reformation liberalism’ is ruled out by the virtue of balance. It is ruled out because the value of liberalism – and the normative justification of liberalism – reposes in a conception of the human good, and runs deeper than the justification offered by a pragmatic modus vivendi. Nor, for similar reasons, can we describe the account I have offered as a variant of political liberalism, precisely because it does not eschew a comprehensive account of the good.

But the perfectionism implied by the liberal virtue of balance is not a hard perfectionism: there is no question of imposing a vision of the good by political means, and we need not be drawn into crude assertions of inherent cultural superiority (for reasons that will become clearer in the following section). However, there is, equally, little sense in saying that this kind of liberalism is neutral between conceptions of the good. It is not neutral in the same way that J. S. Mill’s liberalism is not neutral: it encourages the diversity of ways of life, whilst at the same time celebrating the ability of the individual to rationally revise her conception of the good and to navigate plural sources of value. This celebration itself is a vision of the ‘good’. Where this perfectionism becomes less accommodating is in its principled distaste for oppressive social customs that stifle individual development and ways of life that cannot plausibly be said to harm others.

The positive corollary of this protection of individual diversity is a commitment to liberalism as a cultural doctrine, as an assertion of the intrinsic value of a free and pluralist society. In this respect, we must also say that my account of balance takes us away from the kind of political liberalism associated with the later Rawls. Rawls objected to the kind of cultural and Millian account I am endorsing precisely because it extended beyond the ‘over-lapping consensus’ of a purely political liberalism – just as my account of ‘balance’ does. As I have suggested, balance also brings with it important elements of a ‘comprehensive’ doctrine: it takes the ability to rationally revise and reframe one’s conception of the good as a human excellence – and one that is necessary for moral autonomy. Indeed, if this is correct there is good reason to think that the later Rawls cannot do without such an account. This is an argument that has been well made by others. ²² My only observation here is that ‘reasonableness’ of political liberalism very quickly takes us towards balance, as we can see from the seminal paper, ‘Justice as Fairness: Political Not Metaphysical’:

[C]itizens are to recognize that the weight of their claims is not given by the strength and psychological intensity of their wants and desires (as opposed to their needs and requirements as citizens), even when their wants and desires are rational from their own point of view. ²³

Balance and political legitimacy in a multicultural world

In increasingly multicultural societies it is unsurprising that liberal theorists have sought to strip back the normative reach of liberalism, falling back instead to a more modest position, seeking political but not moral legitimacy. This is mistaken for two reasons, over and above my argument that the value of Rawlsian liberalism ultimately resides in an account of virtue. First, it is mistaken because this account of virtue is not necessarily exclusive to the liberal tradition. As such, it could potentially be the basis for a thicker cross-cultural consensus than that offered by a political ‘overlapping consensus’. Second, we need this deeper appeal if political stability and liberal solidarity are to be based on more than an individual’s instrumental advantage.

Let me simply sketch the first argument. It is important to stress that this not a deductive argument based on the validity of my analysis of the need for a deep self in moral reasoning. Rather, this kind of argument would proceed by interpreting non-liberal traditions in a way that draws out their latent similarity, in particular the virtue of being able to negotiate the demands of different ‘spheres’. We can see this in one of the most pressing cultural tensions of the day: the apparent clash of values between liberalism and Islam. At first sight, Islamic doctrine, with its principled denial of the separation of church and state, looks highly problematic. My argument has been that the value of balance is forced upon us by the necessity of rationally adjudicating between many different ‘spheres’. It is far less easy to make this claim for Islamic doctrine: is there not one source of value, the teachings of the Koran? I am not well placed to make pronouncements here, so I shall make the point tentatively and with the assistance of Tariq Ramadan, who elucidates the sense in which a virtuous Islamic life is one that requires the reasoned negotiation of competing and complexly overlapping spheres. Thus, the Koran and Prophetic Traditions ‘touch upon every area of life in ways both general and diverse and summon human intelligence to discern the difference between the categories … and to try and bring the whole of the message into harmony and make its guidance more accessible’. ²⁴

Hence, the corrective to the first mistake (thinking that balance is an exclusive value) is immersion in non-liberal cultures, with the expectation that there will be common normative assumptions. This clearly takes us away from the dominant assumption that political stability requires cultural neutrality. But it does not amount to a crude assertion of cultural superiority either. This is very much a dialogical process, one of mutual exchange and learning. ‘Our’ side in this process will be to display the values of liberalism more openly – to show that it is indeed a cultural doctrine, rather than simply a political settlement.

The underlying belief here is that some forms of society allow the ‘democratic’ self of this article to flourish more easily than others. Liberal society properly conceived – a society that reflects and fosters the virtue of balance – is the best vehicle for this human flourishing. But here is the crucial twist: to the extent that Muslims and other minority groups are alienated from liberal society and values, it is likely that part of the alienation is because we are not a liberal society in the sense that I am describing. In other words, we are not obviously a society conducive to human flourishing via the virtue of balance. This, I suggest, shifts the terms of the multicultural debate: we should be worried about deeper allegiance to our political institutions, but a large part of that worry must be that they do not command such attachment precisely because they do not reflect or encourage anything other than a very narrow conception of individual flourishing, with a very weak account of the common bonds of citizenship. Crucially, this means that we are faced not just with a problem of multicultural solidarity and common belonging, but also with a wider lack of liberal solidarity.

Let us return, one final time, to the democratic analogy. I have used the analogy of aggregative democracy to highlight the limited appeal of a thin, instrumental view of rationality. In the analogy the virtue of balance has no meaningful role to play in judging the value of ‘democratic’ outcomes. The mostly keenly felt desire takes precedence, even if this leads to a ‘terrible one-sidedness’. Taken as a literal description of political institutions, this kind of democracy does not offer a promising starting point for those seeking to encourage the virtue of balance. Institutional background does much to shape the identity of citizens and the way that they relate to one another, a point that is well made by the literature on ‘welfare regimes’. ²⁵ Yet it is the ‘thin’ model of democracy that dominates institutional processes in many ostensibly liberal cultures. Indeed, many would think that the modern British state is the exemplar of this. The democratic state looks very much like the mere aggregation of individual preferences without substantive discourse to accompany it. Indeed, there is no shortage (and a long history) of writers proud to declare this: ²⁶ democracy is just another marketplace, with parties competing for our votes. This, we should note, is all grist to the mill for rational choice enthusiasts.

The extension of the sphere of the market has of course gone even further in the last two decades. The welfare state has become increasingly driven by consumer demand (in principle if not in practice) and has come to be modelled on the market as far as possible (not very far – hence the ubiquity of ‘quasi-markets’). This model of interaction with the welfare state is designed to bring out the rational choice calculator in us all, with our revealed market preferences leading to the most efficient allocation of goods. What’s of most interest about this account, however, is the way in which it also places the individual in a vertical relationship with the state as a simple service provider. It reduces the individual to the category of ‘consumer-citizen’, ²⁷ and it threatens to supplant liberal institutions and culture with something more akin to a crude utilitarianism.

In fact, in this context, it is something of a distraction to frame the problem of stability only with reference to minority groups. The lack of a strong sense of citizenship extends to the general population too (at least in the UK). Voter turnout continues to dwindle, along with declining support for the shared institutions and solidarity of a strong welfare state. ²⁸ There is insufficient space here to pursue the implications of this last claim. The relationship between liberalism and the modern welfare state cannot be neatly summed up and there are of course a variety of welfare ideals and a variety of liberalisms. Nevertheless, one in particular stands out in this context – the New Liberalism of the heirs to Mill, Hobson and Hobhouse. As Michael Freeden says of the New Liberals in his account of the origins of the welfare state, ‘the notion of human need as flourishing became central to conceptions of human nature, and to the raison d’être of social organisation’. ²⁹

This amounts to a vision that is about more than the narrow provision of services. The New Liberals, we should not forget, were also concerned with the ‘organic’ quality of a liberal society – a sense of cohesion and solidarity, based upon and reflecting liberal principles. Of course, this brings with it a particular view of individual freedom and the relationship of citizen to state. In terms of the classic distinction between positive and negative freedom, we would say that the New Liberal vision is ‘positive’ and developmental. It is the role of the state to enhance the positive freedoms of its citizens, not just to protect negative freedoms and to provide a minimal safety net to prevent material poverty. This means that welfare ‘services’ are to be construed broadly. The Open University, for example, is the kind of institution that the New Liberals could be expected to have embraced. This view of welfare and human flourishing also has a binding function: it represents a sense of shared common purpose – a unifying sense of individual virtue – that our common institutions should be designed to gently but openly encourage the pursuit of.

Conclusion

It will be of little surprise that I suggest that this unifying liberal value is in fact the virtue of balance. Equally unsurprising is that the brief sketch I have offered of the institutional status of balance leaves many questions unanswered. A key area for development, I have suggested, is the relationship between human flourishing and the welfare state, an issue that brings with it a range of conceptual and empirical controversies regarding different welfare systems and what we expect of them. These cannot be broached here, though it is a topic that deserves more attention. But what I hope to have achieved is a positive step in this direction with the articulation of a normative liberalism that is nevertheless open to a continual process of interpretation and dialogical interaction with the variety of non-liberal traditions that have a central place in the collective life of modern, plural societies.