Book review: Rationalism, Pluralism, and Freedom

Jacob T. Levy, Oxford University Press, 2015

Over the last few weeks, I’ve been reading Jacob T. Levy’s 2015 book Rationalism, Pluralism, and Freedom alongside several other books on the topic of freedom: revisiting F. A. Hayek’s Road to Serfdom, reading Lea Ypi’s Free, and so on. In hindsight, I think this was unfair, as these other works were necessarily lessened by comparison with Levy’s.

Rationalism, Pluralism, and Freedom is probably the best twenty-first century book about liberalism that I have read. For those of you who read book reviews for recommendations, then, you already have the highest recommendation I can give: go buy this book. The rest of this review will be a record of my thoughts in response to this wonderfully valuable book. Many of them will not be original, and indeed I think many of them are implicitly present in the text; that’s just more praise for a wonderful book.

Levy’s dichotomy

Levy’s book is ostensibly about the place of ‘intermediate groups’ in the liberal state. But really, I think this is a much deeper book, that’s about the nature of liberty itself: the analysis of intermediate groups is present in large part just because it’s fertile ground on which to explore a tension Levy identifies within the very idea of liberal freedom.

This is the tension between ‘rationalism’ and ‘pluralism’. The former name is ‘meant to encourage the reader to think of Weber, not Descartes’: rationalism is a tradition that seeks rationalisation, that demands intelligible justifications for practices, norms, and hierarchies in order to guard against oppression and coercion. This demand for justification is designed to protect individuals against the local tyrannies that can emerge if power-hungry actors are left alone, their potential victims left without protection. And, as Levy astutely points out, it ‘is a demand that is made in a particular institutional context, i.e. states demanding justification of the practices of non-state groups’.

Pluralism emphasises the varieties of norms and practices, organisations and identities that human beings create and share in. The pluralist insists that ‘free persons act together’, and that (in particular) free persons seek and achieve their ideas of the good life together. If freedom is supposed to allow us all to seek out the good life as we conceive of it, then pluralists insist that freedom of association is the crucial freedom—it allows us to act as the social beings that we are.

Rationalism and pluralism are not natural opposites, and are sometimes mutually reinforcing. But Levy wants us to think about these two traditions in the context of ‘intermediate groups’: religious organisations, clubs, families, ethnic groups, cities and towns, homeowners’ associations, and so on. In this context, rationalism and pluralism begin to offer sharply diverging prescriptions.

Rationalists tend to view these with suspicion: they are potential sites of unjustified power, where authority can become concentrated and local despotism can begin to exercise its power. Churches produce superstition and censorship, ethnic groups whip up hatred and bigotry, and the family—well, the family! As such, rationalists demand that intermediate groups must be accountable for their abuses to the public, which (again) de facto means the state; and they tend to understand the legitimacy of intermediate groups (when they are indeed legitimate) as dependent on the democratic will expressed by the state.

But pluralists take a very different view. For them, these intermediates are themselves the site of freedom, and are necessary part of the good life for social animals: churches are where individuals seek transcendence in group worship, ethnic groups produce ties of kinship and solidarity, and the family—well, the family! For the pluralist, these groups do not depend on the sign-off of the state to be legitimate: they exist independently of it, often as a kind of Hayekian spontaneous order. It is the state that we need to worry about, using the pretext of rationalisation as an excuse to overreach and restrict freedom; strong intermediate groups are important not just in themselves, but also because they can serve as checks on state power.

By not contrasting rationalism to irrationalism, or pluralism to monism, Levy is encouraging the reader to not simply come down on one side or the other—to be sympathetic to both traditions, and to see that this is not a contradiction in need of resolution. He is clear that ‘rationalist’ does not just mean ‘anti-intermediate’, and ‘pluralist’ does not just mean ‘pro-intermediate’: both are attitudes that are broader than just this one question. (This is one reason why I read this book as fundamentally being about liberal freedom, not just narrowly about intermediate groups.) But by focussing on the case of intermediate groups, Levy demonstrates that these two attractive and not-inherently-contradictory conceptions of liberal freedom can be in deep practical tension.

In the eighteenth century, which Levy identifies as the birth of liberalism, pluralism and rationalism found their archetypal thinkers: for the former, the Baron de Montesquieu (the book’s de facto hero); for the latter, Voltaire. But more often, the two tendencies are mixed together in liberal discourse, leading to the confusion that Levy diagnoses in liberal debates over multiculturalism or religion. Rationalism, Pluralism, and Freedom alternates between an intellectual history of liberalism and a philosophical analysis of freedom to show the degree to which liberal thought has always been both rationalist and pluralist, and the reasonableness of both perspectives, without ever letting up on the fact that these are often incompatible ideals.

Levy himself is more pluralist than rationalist, and claims to have written this book because contemporary liberal theory leans too hard into rationalism. (The book itself, and its presentation of contemporary debates, may belie this claim: it might not be that modern pluralist arguments are nonexistent, just that they aren’t good enough for Levy.) But he is also very capable of coming up with powerful arguments for rationalism, as with his Joseph Raz–inspired account of how group life generates authority and authority generates power in chapter 2. Ultimately, his goal is simply to make the differences between pluralism and rationalism more salient, and to help people better understand the ideas and arguments they themselves are making—to admit to, rather than cover up, the difficulties inherent to the idea of liberal freedom.

The Weberian trifecta

‘Rationalism’, in the sense inspired by Max Weber of aiming for the rationalisation of culture, society, and practices, is a pretty common idea. But Levy’s idea of ‘pluralism’ is unusual. It is not the Isaiah Berlin–style value-pluralism that is sometimes said to ground liberalism;¹ it’s closer to early–twentieth century British pluralism, if you know that history.² But by making us think about rationalism in terms of Weber, Levy opens the door to also thinking about pluralism in terms of Weber.

Weber drew a distinction between different types of rule, or (more specifically) different ways that rule can be legitimate. One of these was ‘legal-rational’ rule, which is legitimate by virtue of ‘“competence” founded on rationally devised rules’. This is the type of rule that state bureaucrats and judges exercise, one that is justified with reference to public and intelligible justification, rules and precedent. Another was ‘traditional’ rule, based in ‘the habitual disposition to preserve’ customs and traditions. This type of rule is much more common in intermediate groups: it is the rule of the de facto spokesman for an ethnic group, or of presbyters, or of statutes observed not for their rationality but simply because they have been handed down from time immemorial.

Rationalists demand that intermediary groups are legitimated on legal-rational rather than traditional grounds. Now, the modern state has a powerful tendency to bureaucratisation and ‘rationalisation’; this was one of Weber’s main theoretical concerns. But few intermediary groups actually can give fully public and rational justifications of their customs and internal organisation. Thus, there are structural barriers making it harder for intermediate groups to meet the rationalist’s demand for public legitimation, while structural forces actively make it easier for the state to do so. And as such, rationalists (often unwittingly) tend to end up on the side of the bureaucratic state against associations and groups. But meanwhile, pluralists seek to preserve the variety of associations and forms of social living against the bureaucratic onslaught of the modern state; and so they (again, often unwittingly) end up on the side of unquestioned tradition against liberal intervention.

Levy’s rationalist/pluralist distinction does not map exactly onto Weber’s legal-rational/traditional distinction: for one thing, not all groups are founded on tradition, and some (like the Catholic Church) are deeply bureaucratised. Indeed, Levy thinks that local government is a type of intermediate group, and as someone who has spent a lot of time campaigning in local politics I can tell you that local government is just as bureaucratised as national government. But these two distinctions are, at very least, closely related.

But this suggests a way in which Levy’s description is incomplete. For Weber’s distinction was threefold: he distinguished between legal-rational rule, traditional rule, and charismatic rule. It was this last one that Weber pinned his hopes on: rule by the ‘gift of grace’, the leader who blows away accumulated bureaucracy and tradition and who redefines the terms in which politics is discussed. As the state became increasingly bureaucratised, and as its monopoly on force shrank the space for tradition, Weber became more and more pessimistic about the prospects for liberal freedom: his only hope was that successful political leadership would be able to break the stranglehold of bureaucrats and special interests over the state.

Legal-rational rule is primarily (though not exclusively!) associated with the state, and traditional rule primarily (though not exclusively!) associated with intermediate groups. But charisma can legitimate both states and intermediate groups equally; we apply the word to religious leaders as easily as politicians. (Indeed, many charismatic speakers fall into both categories.) Charisma can create publicly-justifiable rule that is ‘thick’ and meaningful; but it also creates its own risks of tyranny, demagoguery, and subjugation. The main problem that Weber’s political thought aimed to solve was how to create institutions that could foster the potential for charismatic leaders to emerge, and that could allow them to exercise their rule fully to enhance individual freedom.

Weber’s own proposals were failures, to put it mildly: he was one of the strongest proponents of Article 48, the emergency powers provision in the Weimar constitution that was invoked by Adolf Hitler in 1933 after the Reichstag fire. But whatever the strength of his answers, it seems to me that his questions still haunt us. What is the place of leadership in remaking the grounds of politics? Can charisma be used as a tool for enhancing freedom, without sliding into demagoguery and despotism? Is there a way to avoid falling into both the trap of rationalists being overly sympathetic to bureaucracy, and the trap of pluralists being overly sympathetic to tradition?

But Weber’s concerns seem largely absent from Levy’s book. Levy rightly dismisses the idea that the state is a machine for dispensing justice, but (on my reading) in the philosophical sections of the book he all-too-often falls into the opposite trap of seeing it simply as a hulking mass that is just there, in opposition to individuals and groups.³ The question of how political leadership might remake the state, or remake the culture in which groups participate, is absent. This is a deep shame, but an understandable one: for the last century, liberal political thought has failed (or often not even tried) to get beyond Weber.⁴ I cannot claim to do any better than Levy has.

The historical trifecta

So that’s one way that Levy’s description of a twofold distinction can be analysed as a limited perspective on a more fundamental threefold distinction. I now want to propose another way to do this, that may or may not be compatible with the previous section.

Levy’s history begins with a study of the mediaeval origins of the intermediate groups he cares about, and then the ancient constitutionalist tradition that re-theorised them during and after the rise of the modern Weberian state. He tries to be fairly even-handed about these, discussing guilds, universities, courts and parlements, parliaments and Estates, and giving most space to the Catholic Church—which, after all, was by far the most contemporaneously important of any of these bodies.

But for this reader, at least, what stood out in Levy’s account was his account of cities as intermediaries. In the mediaeval period, cities were not just thought of as places: they were corporate bodies, subject to a struggle over whether their legitimacy came from the prince or from the people. This is important because it was the city that was ultimately the most consequential of all the intermediate bodies, when the city-dwelling bourgeoisie began to dominate European politics in the early modern period. Bourgeois political thought, or at least part of it, is what has been handed down to us today as ‘liberalism’.

Two of the most important traditions of early modern political thought were both (in rather deep ways) political theories of the city: the social contract and civic republicanism. Levy astutely notes this in the case of contractarianism. He observes that mediaeval cities were often quite literally founded on a contract, and that their offer of significant individual freedom to their citizens was the basis of the contractarian understanding of liberty. ‘The civil society of contractarianism,’ Levy writes, ‘was the city write large, as large as the new modern state.’ And, while he is less explicit about this, the same is true of civic republicanism, which (as the name implies) thought about politics in terms of the civitas. David Hume knew that ‘even under absolute princes, the subordinate government of cities is commonly republican’; those who made republicanism their political project sought to extend the subordinate level upwards. Republicans and contractarianism differed on whether to focus on the classical or Germanic city, but for both the free city was the ideal that they sought to realise in the state as a whole.

Ancient constitutionalism, however, was a political theory of the city in a different sense. Rather than analysing the nature of civic freedom in itself, so as to understand how to best extend it beyond the city, the ancient constitutionalists analysed civic freedom in context. They placed cities alongside other intermediate powers to analyse the wider conditions that allowed urban liberty to emerge. Of course, the ancient constitutionalists and those they inspired were often in fact not primarily concerned with the city! The city played a supporting role in accounts that gave more attention to other intermediaries: consider the minor role of cities in Montesquieu, who attributes much more importance to the parlements and noble bodies. But when the bourgeois began their rise to prominence, they were attracted primarily to the way the ancient constitutionalists thought about the city in particular; and so before too long, we had Adam Smith’s Montesquieu-inspired history of liberty emerging from the commercial towns of Europe.

Republicanism and the social contract were concerned with understanding the city in itself so as to extend its freedom to the state as a whole, while ancient constitutionalism analysed the city in its wider context in order to better understand and protect that freedom. This suggests one reason to proceed as Levy seemingly has, and combine republicanism with the social contract under the headline of ‘rationalism’: these two share a central analytical emphasis that ancient constitutionalism, and the pluralism it inspired, lacks. But this seems to be a rather one-sided view of the matter.

As Levy is well aware, none of these three political languages survived the eighteenth century unchanged, and in many cases they were united together for various political purposes by the bourgeois of the age. (This bringing together of different political languages in the late eighteenth century is, I think, is a more plausible dating for the birth of liberal political theory than Levy’s implicit proposal of the publication of The Spirit of the Laws.) And this creates, not a single tension between two distinct perspectives, but multiple tensions between three political languages; and simultaneously, it creates the possibility of combining more ‘pluralist’ with more ‘rationalist’ traditions.

And so, yes, it is true that figures like Thomas Paine or Jean-Jacques Rousseau combined republicanism with the social contract to create various highly-rationalistic forms of (liberal and illiberal) political theory, standing in sharp contrast to any form of pluralism. But these were not the only options: Hume and James Madison famously used an account of intermediate groups—in their case, political factions—to defend republics against Montesquieu’s charge of anachronism; and many forms of nationalist political thought used the language of the social contract to propose the existence of a nation that was both independent of and prior to rationalised state power. These three languages could be, and were, remixed and mashed-up in innumerable different ways as liberalism began its difficult ascent to political prominence.⁵

This created many tensions within liberalism, of which the issue of group life was only one. Liberalism absorbed two different perspectives on the city: one that (with significant internal disagreements about how best to achieve this) sought to extend civic freedom up to the level of the state, and one that sought to protect civic freedom against encroachments from the state. And so, as liberals gained power over the state, their complicated intellectual legacy led them to split over how best to deal with intermediate powers: whether to see them as sites of freedom that need defending from the potentially illiberal state, or anachronistic sites of coercion that previously nurtured liberty but must now be brought into line by the newly-liberal state.

The tensions within liberalism that Levy identifies, then, might be best understood not in terms of a divide between rationalism and pluralism that spans the centuries. Instead, it is a historically-located tension, just one part of the coming together of three partly-incompatible political languages in the eighteenth century, as the city-dwelling bourgeois were forced to respond to radically new political challenges with all the resources at their disposal. The difficulty of properly accounting for group life was merely one among many difficulties facing this faithful marriage of convenience (in Judith Shklar’s sense) between different views of the city.

Political economy

Focussing on liberalism as a bourgeois political theory (literally—a political theory of the burgh) also brings another dimension into focus that Levy deliberately ignores. For the important fact about the cities of the eighteenth century was that they were engaged in commerce. This, after all, was what enriched and empowered the bourgeoisie to the point that they could even consider usurping monarchs and nobles. And it was commerce and political economy that started the ‘ideological war within the bourgeoisie’, between those who sought to capture or recreate the concentrated power of the nobility and those who sought to overthrow it entirely, that Eric Schliesser has (convincingly, in my opinion) argued was the birth of liberalism.

If we are concerned with a history of liberalism, then, commerce and political economy should be absolutely central. But Levy attempts to leave these aside. In his introduction, he tells us that he plans to mostly ignore those groups such as unions or corporations that are primarily oriented towards the ‘common or external space’, like the market or state; his focus, instead, is on groups that are oriented towards their own internal life. Corporations care only instrumentally about their internal governance structure, while churches have split over it.

This distinction, however, won’t work as it stands. To give a simple—perhaps too simple—example, consider when my parents used to volunteer to bring sandwiches and snacks to a race organised by a local athletics club (an intermediate group if ever there was one!). While my parents used their own labour, they could not provide all the inputs: they had to purchase them. They bought the mayonnaise, they bought the bread, they bought the chicken, they bought the knives and the breadboards and the salt and pepper and the tupperwares to store the finished sandwiches. Each of these items was the product of an ‘I, Pencil’–style story of exchange leading to further exchange leading to… Partaking in internal group life required my parents to turn outwards, to rely on the cooperation of people outside the group. As such, the ‘external’ orientation is not simply contrary to the ‘internal’, and in some ways the former is prior to the latter.

Another case is that of property. As Levy astutely notes, groups’ ability to own property corporately (rather than just using the property of their members) is often crucial to their long-term survival and success, and their robustness against state challenges. But the ability to purchase, or affordably hold, property is set in large part by market forces that go far beyond the group’s internal structure. To ensure their own internal permanence, then, large religious or social groups must partake in the market at least as much as many small or medium-sized businesses.

As these examples show, the distinction between ‘internal’ and ‘external’ orientations cannot easily be maintained in the modern world. When churches no longer brew their own wine, and when their ability to hold property becomes subject to market forces, then even the most basic ‘internal’ aim of congregating together in a shared space to partake in the Eucharist becomes reliant on their partaking in the ‘external’ market. And (on the other side) when political parties become more responsive to their members than their voters when selecting a leader or proposing policy, they are compromising on the ‘external’ goal of political power in favour of an ‘internal’ orientation towards expressing their members’ values. The two things cannot be torn apart.

Rather than just being a technical philosophical argument, this points towards a defining feature of modern life. As Smith observed, we live in a society where the division of labour has become so complicated that, in the pursuit of all our needs and desires—including the needs and desires we have qua members of intermediate groups—we rely on exchange and trade. As he described it, everyone has become (to some measure) a merchant; and so society becomes alike to the place where merchants trade, the name for which is ‘market’. G. W. F. Hegel (for one) understood this quite deeply, which is why he defined ‘civil society’ in such a way that it could not be distinguished from the market. The twentieth-century attempt to so distinguish the two concepts was a step backwards: our analysis of civil society, and the intermediate groups that make it up, cannot ultimately pretend that it is a sphere apart from the market.

Levy’s introduction suggests that he already knows all this, but is engaging in some useful simplification to avoid the complex ‘four-body problems’ that arise when multiple different groups in a market simultaneously interact with the state and the individual. The question then becomes whether his simplification is actually useful, rather than harmfully reductive. In one way, the answer can only be ‘yes’: Levy has produced a magnificent book using this method of simplification. But I will end by suggesting three places where I think Levy is led astray by his method.

First, he simply misses some relevant ways to extend his own insights. While Levy is perhaps right that too much focus on the market would require him going too deep into the topic of economic justice, the arguments of his book suggest that existing writings on that topic (and related topics) are limited in important ways. I might provocatively suggest that his ideas open up the space for someone to play the pluralist to Elizabeth Anderson’s rationalist, and defend corporations as crucial intermediaries against state power. Less provocatively, Levy gives us important reasons for centring interest groups and corporations in our account of political economy, moving away from the ‘econ 101’ focus on individuals while also providing a crucial corrective to some of the hostility (whether from social democracy or public choice) that these intermediate economic groups face in more sophisticated accounts.

Second, Levy cannot account for how the dynamics of the market have shaped the behaviour of intermediate groups. Smith’s observation that ‘every man becomes to some measure a merchant’ in the modern world should be, but all-too-rarely is, placed in the context of his aggressive and violent attack on merchants and the intellectual system he attributed to them, ‘mercantilism’. Today, the state often acts as if captured by merchants (a standard public choice insight); but also, all intermediate groups can come to resemble groups of merchants. To give just one example close to my heart, associations of local residents or homeowners in the West tend to act as local cartels, using their group power to restrict supply and raise prices—exactly as Smith would have predicted of any group of merchants coming together in the same place. Levy discusses some political analogues of these observations in chapter 10, but fails to see the crucial economic side of the issue. Precisely because they exist in a market, intermediate groups are no longer more than roughly analogously to their mediaeval counterparts; a proper attention to groups requires a proper attention to markets.⁶

And finally, and most importantly, Levy’s attempt to abstract away from the market leaves him unable to answer some important questions about liberalism’s own success. In Levy’s chapter on the Age of Revolution, he says Smith is unique among liberal thinkers in managing to find an ‘unusual’ balance between pluralism and rationalism. The closest comparison Levy can think of is Hegel; but Levy thinks Hegel ultimately failed to overcome the dichotomy, and he gives one of the best arguments I’ve seen for why Hegel-as-statist is still a legitimate interpretation. Thus, Smith alone in Levy’s history successfully balanced the two tendencies. The impression one gets from Rationalism, Pluralism, and Freedom is that this was almost magical, a product either of Smith’s unique and incomprehensible talent or (better) of his particularly incisive scepticism. But I do not think that Levy actually believes this, and so we are left with the question of why Smith was so able to unite these divisions.

I want to suggest that it was precisely Smith’s focus on political economy that helped him do this. The core of Smith’s attack on mercantilism, and the basis of his own liberalism, was a focus on what we would today call non–zero sum games. He saw how exchange, trade, and economic development could make everyone involved better off. And, unlike the narrow, instrumental, and materialist conception of ‘gains from trade’ that is so prevalent in today’s economic discourse, Smith meant this in the broadest sense. In The Theory of Moral Sentiments, for example, he discusses how status games can be made non–zero sum, how an increase in esteem and self-respect for one individual or group can boost (rather than undermine) the esteem or self-respect of another.

Formalised, the argument might be something like this: whatever you value, it is likely that exchange and growth can help build a world where you are better able to achieve it, by increasing the range of opportunities and materials available to each citizen to do whatever they wish with. The most explicit version of this argument was made by Tyler Cowen in Stubborn Attachments, but in more implicit forms it is perhaps older than liberalism itself.⁷ There may be some values or ways of life that are necessarily harmed by economic growth, rather than just harmed by the contingent form that growth has taken (monastic virtues might be one example; republican virtue, understood in terms of Montesquieuan self-denial, might be another). But these are rare, and in a market society increasingly so. And in particular, neither rationalists nor pluralists have values that are inherently incompatible with market exchange or economic growth, and both have ends that can be served by sustainable development.

My suggestion is that this insight was at the basis of Smith’s balance between pluralism and rationalism, and more broadly the foundation of liberalism’s rise. Smith saw the harms of concentrated group power and of concentrated state power. But rather than seeking to capture one source of power to undermine the other, he argued that the ‘system of natural liberty’ offered a way to defuse both while giving both groups and individuals more of what they valued. Increased wealth (again, understood as broadly as Smith understood it) offered to individuals both the resources to enrich their group life and the resources to escape that group life, whichever they would so choose. The tension that Levy identifies might be theoretically insurmountable; but in thinking about it practically, in the context of a market of increasing extent, liberalism can come to understand its own greatest strength.

1. In a concluding chapter, Levy suggests that Berlin’s monism/pluralism distinction is logically prior to his rationalism/pluralism distinction.

2. If you don’t, Levy gives the British pluralists a lot of space in his book, although I feel he falls prey to the all-too-common failure to grasp that their ideas were already irrelevant once Max Weber began to write on politics in a more sustained way. Perhaps that is part of why Rationalism, Pluralism, and Freedom feels marked by a general lack of engagement with Weber (on which more below).

3. There is no such problem in Levy’s historical account, which necessarily has to engage with a period before the modern Weberian state emerged and thus cannot just take it as a given.

4. For an example: so much of public choice theory is simply an analysis of how the state acts in the absence of charismatic leadership, and as such goes beyond Weber only in formal proficiency.

5. And all of this, of course, was carried on with the spectre of a third trifecta lurking in the background: Montesquieu’s distinction between republics, monarchies, and despotisms, and his warning that attempts to hybridise the first two would likely slide into the third. As nineteenth-century liberals attempted to combine various seemingly-incompatible forms of politics in response to unprecedented challenges, there was a constant worry that they could overreach and slide beyond themselves: that the Weberian state might return to what it originally had been, the Hobbesian state. (It is very plausible to me that Montesquieu’s account of the form of despotism drew explicitly on Hobbes.)

Levy identify fear of state despotism as a pluralist tendency, but to me it seems much more general. By its nature, liberalism aims to combining different political ideals to deal with the problems of the present age. But it also absorbed the lesson from Montesquieu that such combinations can often put us on the road to serfdom. And so the liberalism of fear, the terror of despotism and (later) totalitarianism, is a permanent and universal form of liberalism precisely because liberalism needs to be wary of itself in order to maintain its own stability and not slide towards concentrated power. (Think of John Maynard Keynes’ response to Hayek.)

6. This is one reason why, despite my love for The Spirit of the Laws, I find Montesquieu less important for contemporary liberalism than Levy does. Montesquieu’s preferred intermediate group was the nobility, who were explicitly banned from engaging in commerce in his ideal monarchy. As such, in spite of his emphasis on the importance of both intermediate groups and commerce in modern society, he cannot coherently analyse both at the same time.

7. Cowen says that this is a pluralist argument, but he means that in the sense of Berlinian value pluralism.