Public Reason, Partisanship and the Containment of the Populist Radical Right

This article discusses the growth of the populist radical right as a concrete example of the scenario where liberal democratic ideas are losing support in broadly liberal democratic societies. Our goal is to enrich John Rawls’ influential theory of political liberalism. We argue that even in that underexplored scenario, Rawlsian political liberalism can offer an appealing account of how to promote the legitimacy and stability of liberal democratic institutions provided it places partisanship centre stage. Specifically, we propose a brand-new moral duty binding ‘reasonable’ partisans committed to pluralism. This duty establishes conditions where partisans must strategically transform society’s public reason (i.e. transform the visions for society their parties campaign on) in ways that promise to attract back support from illiberal and antidemocratic competitors. While this strategic behaviour might seem impermissible, we show that Rawls’ distinctive account of sincerity in democratic deliberation is uniquely placed to justify it as perfectly ethical.

by authoritarian beliefs (Mudde, 2007: 22-23). Examples of populist radical-right parties and leaders that have recently attracted considerable support include the Italian La Lega, Rassemblement National in France, Alternative für Deutschland in Germany and the 45th US President Donald Trump (Mudde and Rovira Kaltwasser, 2017).
This article discusses PRR from within normative political theory. We use PRR as a case study to develop John Rawls' prominent account of political liberalism in new directions concerning the containment of possible threats to liberal democracy. Our argument is that a Rawlsian approach involves that under certain conditions, partisans of mainstream parties have a moral duty to react against competitors that, like PRR, reject key liberal democratic commitments. Those partisans must employ their characteristic creative abilities to collectively change the whole range of mainstream political platforms on offer (or, in Rawlsian language, to transform society's public reason). While staying within limits we set, partisans must be strategic in developing visions for society that promise to be effective in inspiring back the support lost to PRR and similar competitors.
Our article offers several contributions, directed at any political theorist interested in the tenability of Rawlsian political liberalism, either as supporters or critics. Rawls investigates how liberal democratic institutions can create their own support over time, moving from strength to strength (Rawls, 2005: 158-168). However, he never focuses on the case in which this trend slows down, stops and eventually reverses, with beliefs rejecting liberal democracy gaining ground. The growth of PRR exemplifies that case, highlighting difficult questions around whether Rawlsian political liberalism can demonstrate that, even under these circumstances, the legitimacy and stability of liberal democratic societies can be protected. Political liberals have only just started to address these large questions, and we aim to make substantial progress by drawing attention to partisanship.
Our goal is to suggest that Rawlsian political liberalism can be made able to handle cases where liberal democracy is losing supporters. We argue that at least at a general level, a Rawlsian framework can provide partisans with normative guidance in those cases. Specifically, it excludes both self-righteous political paralysis and the appropriation of illiberal themes. Importantly, Rawls' distinctive account of sincerity in public reasoning makes Rawlsian political liberalism uniquely well-suited to justify as ethical some forms of strategic behaviour that we intuitively expect of mainstream partisans transforming their platforms in this context. Moreover, this investigation allows us to make original theoretical points about Rawlsian political liberalism at large. Our novel argument that there are scenarios where public reason must be transformed gives the full picture of the dynamic nature of Rawlsian public reason, providing the strongest rebuttal to date of the criticism that public reason is a hopelessly conservative ideal. Also, by building on Rawls's work, we argue that our duty applies to a variety of partisans across the left-right spectrum, who should experiment with different visions. This highlights a brand-new source of appeal of Rawls's characteristic appreciation of wide 'reasonable' pluralism of political views as a fact of liberal democratic life -that is, its contribution to a theory of containment of possible threats to liberal democracy.
The article is organised into five sections. The first reconstructs Rawls' theory and the few existing political liberal analyses of PRR. Next, it argues that to remedy the incompleteness of those analyses, partisanship must be placed centre-stage. The second introduces the duty of partisans to transform public reason, justifies it and outlines the conditions where it applies. The third further specifies the duty, drawing on real-world examples of partisan action. The fourth replies to two possible objections, while the fifth argues for the surprising conclusion that the strategic behaviour we prescribe is sincere in the relevant Rawlsian sense. Rawls' political liberalism aims to demonstrate that a liberal democratic state can be legitimate and stable, where legitimacy is understood as wide justifiability of the principles behind that state, despite the disagreement about religion, the good life and philosophical issues that results from the free exercise of our burdened reason (Rawls, 2005: 54-58). Legitimacy and stability are possible despite this 'reasonable pluralism' of comprehensive views because, if duly reduced to their political core, liberal democratic commitments can be the subject of an 'overlapping consensus' among the most diverse comprehensive doctrines.

Political Liberalism, PRR and Reasonable Partisanship
Rawls' two basic political ideas of (1) society as a cooperative system for everyone's benefit and (2) persons as free and equal cooperators in that system are consistent with otherwise extremely different belief systems concerning God, excellence in life and other comprehensive matters. Endorsing the two basic political ideas commits you to liberalism, but Rawls stresses that they can be interpreted differently, within certain limits. Therefore, different citizens can plausibly hold very different conceptions of justice, although they will all include: first, a provision of basic rights and opportunities for all; second, special priority for such rights and opportunities; and third, adequate all-purpose means for all citizens to make use of them (Rawls, 1997: 773-775).
When political power is formally exercised in settling constitutional essentials and other issues of basic justice, legitimacy requires that decision-makers only support decisions grounded in 'public reason'. They must be able to offer at least one argument that makes no reference to deep comprehensive commitments and is built from within a framework provided by the two shared basic political ideas and a liberal conception of justice they can plausibly take to convincingly interpret those ideas. This reason-giving requirement makes Rawls' account of democracy deliberative (Rawls, 1997: 771-773).
Given Rawls' acknowledgement that reasonable pluralism extends to political matters, the content of a society's public reason is 'given by a family of [liberal] political conceptions of justice, and not by a single one' (Rawls, 1997: 773). Specifically, there is a constant debate in society regarding which conception should be adopted among the many compatible with the two basic ideas; public reason's content is provided by several conceptions that dominate these sorts of discussions at present in society (Rawls, 2005: xlviii-xlix and 226-227).
The two basic ideas are shared by all 'reasonable' persons but are rejected by everyone else. Only reasonable persons will therefore approach important political decisions as required by liberal legitimacy, that is, from within one of the liberal conceptions plausibly interpreting the two basic ideas. On Rawls' definition, reasonable persons seek terms of social cooperation that produce fair benefits for all citizens, understood as free and equal in virtue of their sheer ability to be fully cooperating members of society (Rawls, 2005: 49-50). Also, reasonableness implies acceptance of the 'burdens of judgement' and therefore that the free exercise of reason produces disagreement (Rawls, 2005: 54).
On Rawls' account, the stability of legitimate liberal democratic institutions hinges on enough persons in society being reasonable. However, unreasonableness is never erased, not even in the most well-ordered liberal democracy possible. Rawls mentions, but leaves unexplored, the possibility that the growth of unreasonable views might then give society the key 'task of containing them − like war and disease − so that they do not overturn political justice' (Rawls, 2005: 64). 1 Jonathan Quong has provided an early analysis of this task, proposing strong containment measures. Basic individual rights, including free speech, should be denied to the followers of unreasonable views provided that they pose a 'real threat' to the stability of liberal democratic institutions (Quong, 2011: 290-314). However, in previous work, we criticised Quong's account as providing an incomplete picture of liberal containment. Although rights violation appears justified in extreme circumstances, complementing it with softer containment measures, to be enacted earlier on in the build-up to a real threat, is necessary to express real care for everyone in society enjoying equal rights (Badano and Nuti, 2018: 153-154). 2 Indeed, the more recent literature discussing containment primarily concentrates on those softer measures (e.g. Clayton and Stevens, 2014), which provide the main focus of both our previous work and the rest of the Rawlsian literature looking at PRR.
In this area, Rawlsian scholars aim to explain why PRR cannot be accepted as a player like any other in democratic politics, using unreasonableness to explain why its success should be concerning. Populist radical-right parties and leaders exclude from their vision of society groups whose members have all the necessary abilities to become full participants in social cooperation, making PRR's supporters unreasonable in a first sense (Badano and Nuti, 2018: 149-151). 3 Indeed, PRR spells out populism's 'us/other' logic so that members of ethnic, religious and cultural minorities 'are systematically singled out qua members of that group' and unreasonably portrayed as 'lack[ing] the sheer capacity to even become fully cooperating members of society' (Badano and Nuti, 2018: 150).
Moreover, PRR is unreasonable in a second sense because of its populist commitment to a 'moralised form of antipluralism' (Müller, 2016: 20). By portraying themselves as the only legitimate representatives of the homogeneous people, populists (and, therefore, radical-right populists) dismiss any opposition as illegitimate, thereby 'eliminating the space for reasonable disagreement' and unreasonably rejecting the burdens of judgement (Ferrara, 2018: 470). To highlight that PRR is twice unreasonable is worrisome because, as explained above, reasonableness is necessary to embark on a process of public reasoning that, speaking to a sufficiently inclusive audience, can contribute to liberal legitimacy. Moreover, the greater the traction of unreasonableness in society, the fewer participants in the overlapping consensus over basic liberal commitments, undermining the stability of liberal democratic institutions.
PRR supporters fail their duty to participate in public reasoning and otherwise be reasonable. They are morally blameworthy for this, not their reasonable fellow citizens. However, according to the political liberals discussing PRR, the protection of legitimacy and stability is so crucial as to create remedial moral obligations for at least some reasonable actors to serve as agents of containment. In other work, we carve out a role specifically for ordinary citizens. Reasonable citizens have a duty to 'press the unreasonable they know (e.g. relatives, friends and colleagues) on their political views to change their mind and push them towards greater reasonableness' (Badano and Nuti, 2018: 157). These discussions should be sought at gatherings with friends and in other 'nonpublic' forums, where reasonable citizens can use their individualised knowledge of their unreasonable interlocutors to be persuasive. According to Alessandro Ferrara, a major cause of PRR's success is the powerlessness that citizens experience towards transnational 'disembedded financial markets' (Ferrara, 2018: 472). Liberal democracies should therefore employ the law to empower citizens as consumers; facilitating class actions against banks and corporations becomes a containment strategy (Ferrara, 2018: 474).
Here we do not wish to dispute that there might well be potential for the law to reduce citizens' anxiety or for ordinary citizens to contribute to pushing acquaintances away from unreasonableness. However, existing theories of containment of PRR are significantly incomplete. There is a crucial sense in which neither legal proceedings nor individual efforts are political, falling short of any large-scale competition for popular support and, in turn, power. Politics is a significant resource that must be exploited because it contains powerful agents (i.e. partisans of reasonable political parties) able to directly counter unreasonable parties and leaders and weaken their support.
Political liberals interested in containment should now turn to (reasonable) parties and partisans because of their ability to change public opinion. 4 As influentially described by Giovanni Sartori, parties have the ability to work as 'transmission belts', connecting the general public and the state. They can scan for, capture and systematise ideas that are already present in civil society. Parties are unique in that they can translate such ideas into policy (Sartori, 1976: 25). Importantly for us, this process can also actively reshape public opinion so that the platforms parties run on and gather support around always go beyond those ideas (Sartori, 1976: 27-29). Similarly, Nancy Rosenblum stresses the creative function of partisans in generating the very terms of reference of political debate (Rosenblum, 2008: 7 and 366; see also Muirhead, 2014: 85-91). As argued by Jonathan White and Lea Ypi, 'partisans contribute to shaping the political culture around them' (White and Ypi, 2016: 68); the argument styles acceptable in political debate, the premises from which to argue and the problems to be prioritised are to a large extent determined (and changed over time) by partisans.
The Italian Partito Radicale well exemplifies the creative abilities of partisans we aim to put centre-stage. In the 1970s and 1980s, they placed in the public eye a phenomenon that had not yet been widely perceived as important -a system where the major parties received substantial public funding and clientelistically distributed key appointments in public administration and the private sector. Partito Radicale did so by framing their denunciations ingeniously, creating terms of reference that are still important in Italian politics; they spoke of 'partitocrazia' ('partitocracy'), strongly conveying the inconsistency between democracy and the system they were attacking, and 'cupola', comparing Italian party politics with the commission leading the Sicilian mafia (Petti, 2012).
In sum, 'ethical partisanship' -the normative exploration of what partisans should do -should take centre-stage in the political liberal literature on PRR and containment more in general. 5 What the intersection between political liberalism and ethical partisanship can teach us about containment is the focus of the rest of the article.

Reasonable Partisans' Duty to Transform Public Reason
This section aims to introduce a novel duty held by reasonable partisans, generated not by a real threat to stability, but by a process likely to lead to one. Under such circumstances, reasonable partisans are morally required to transform society's public reason by changing the programme that their parties each propose to voters, introducing into the political debate alternative conceptions of justice that do not presently have wide currency but fall within the limits of reasonableness. The reasonable conceptions of justice that presently dominate mainstream public discourse have started losing support, and this 'moral duty to transform' (MDT) public reason is meant to stop that. In effecting change, partisans should be driven by strategic considerations explaining which conceptions of justice have a chance of winning back support from unreasonable competitors. We will now justify and describe MDT at a general level. The next section will provide concrete examples to better clarify it.

The Justification for MDT
To understand the idea of partisans transforming public reason, let us start with the notion that public reason can change. As seen, given that reasonable disagreement extends to political matters, the content of a society's public reason is provided by several conceptions of justice, which can be significantly divided on important issues. Specifically, such content is given by the conceptions that are currently prominent in political discourse out of those in principle compatible with reasonableness. Consequently, public reason's content may change over time. '[N]ew variations [of reasonable conceptions of justice and therefore of public reasons] may be proposed from time to time and older ones may cease to be represented. It is important that this be so; otherwise the claims of groups or interests arising from social change might be repressed and fail to gain their appropriate political voice' (Rawls, 1997: 775). This is the fact of public reason's 'mutability', which Rawls employs to reply to Jeremy Waldron's prominent objection that public reason is a static and therefore conservative ideal. 6 Matteo Bonotti builds on Rawls' quote to explain why parties are useful political actors for political liberalism. Due to partisans' position between the general public and the institutions bound by public reason, partisans are uniquely able to scan civil society for emerging reasonable conceptions of justice and consider them for inclusion in their public reasoning. In his language, thanks to partisans, 'potentially new conceptions of public reason are granted a hearing' and 'given the chance to become part of the public reason vocabulary that political representatives ultimately ought to employ when justifying legislation' (Bonotti, 2017: 134). For Bonotti, partisans are important enablers of change in public reason, serving, as originally stressed in Rawls' quote, inclusivity.
We are sympathetic to Bonotti's idea of partisans as enablers of change, but we importantly differ from him. While for Bonotti partisans' ability to change society's public reason is valuable because it fosters inclusivity, on our account that ability should be put in service of the stability of legitimate liberal democratic institutions. We call on reasonable partisans to develop fresh reasonable conceptions of justice, which should then form the core of their parties' political campaigns, to inspire back the support lost to worrisome unreasonable competitors. Relatedly, for us, the fact that partisans can enable change in public reason's content is not simply a welcome feature of parties. Reasonable partisans have a duty to transform public reason under specific circumstances. Our idea that there are circumstances in which public reason must change is completely original and theoretically interesting. While Waldron's suspicions that public reason is a rigidly conservative ideal are eased by Rawls' and Bonotti's point that public reason can change, the full extent of public reason's dynamism is only captured by MDT.
How can we justify this duty? On a political liberal account, legitimate power equals reasonable rule. To the extent that PRR or other unreasonable actors exercise any formal political power, legitimacy suffers, and any sign that they might end up in that position must be taken seriously. Legitimacy is so important for political liberalism that it traditionally justifies moral duties, even for common citizens, aimed at securing it -most prominently, the duty of public reason (Rawls, 2005: 217). In analogy with public reason, our previous work derives from legitimacy an additional duty falling on common reasonable citizens when a threat to stability is on the horizon -the duty of pressure reconstructed in the previous section (Badano and Nuti, 2018: 156-157).
As citizens, partisans are subject to the directives of the duty of pressure. Therefore, they should engage one-to-one with their acquaintances attracted to unreasonable views, by using their knowledge of the specific life history and belief system of each interlocutor to persuade them that those views are wrong. We now proceed to demonstrate that partisans qua partisans also have at least another moral duty aimed at protecting liberal legitimacy. One cornerstone of the literature over ethical partisanship is the idea that partisans have special obligations besides those that all citizens have -obligations that normally place centre-stage (critical forms of) loyalty to one's party and its core values. Different grounds have been proposed for such obligations, and one strength of MDT is that its justification can be built on any of several such grounds. Therefore, we will remain agnostic between them.
One possible justification for MDT employs the argument stressing that partisans publicly affirm political commitments they share as integral to what their party stands for. This creates a legitimate expectation that they be true to their word, contributing concretely to the pursuit of those commitments (White and Ypi, 2016: 110-113). Reasonable partisans across different parties disagree over many important issues. However, they are by definition publicly committed to an idea of society as fair system of cooperation among all persons and to reasonable pluralism, which, for example, are both rejected by PRR. Therefore, the expectation is that, if needed, reasonable partisans concretely defend these commitments they publicly express as fundamental components of their partisan shared projects. In sum, reasonable partisans are required to fight unreasonableness beyond any duty common citizens might have. Another possible justification builds on the argument underlining that partisans receive distinctive benefits (e.g. increased political influence) from belonging in their community of fellow partisans. According to White and Ypi, this generates an obligation to give something back, including loyalty to one's party and its fundamental commitments (White and Ypi, 2016: 110-113; see also Bonotti, 2017: 15-19). Arguably, given that for reasonable partisans such commitments centrally include reasonableness, a special requirement to fight unreasonableness follows. Alternatively, one could explore the possibility that the justification for such special requirement comes down to the fact that among reasonable citizens, partisans are particularly well-placed to sway persons away from unreasonableness. 7 Whichever justification we pick, reasonable partisans have a generic additional obligation to protect the stability of broadly reasonable rule. How can we move from it to MDT? We do not deny that there might be further tasks partisans should undertake to satisfy that generic obligation, but we leave them for another day. Our claim is that whatever else might be required, MDT is justified in the cases discussed in this article, where progressively fewer persons vote for or otherwise display support for the range of reasonable visions for society (or, in Rawlsian language, conceptions of justice) making up the platforms reasonable parties offer to the electorate and therefore place centre-stage in mainstream political debate. Faced with this declining demand for the range of platforms on offer from reasonable parties, partisans worried about the prospect of unreasonable actors increasing their political power have no choice but to shake themselves out of this trend by proposing something different. Specifically, reasonable partisans should aim to change their platforms in ways that, while staying within the limits of reasonableness, promise to intercept some of the demands that are presently catered for by unreasonable competitors. 8 As explored in the 'Some Hope for MDT's Success' section, there is reason to hope that these changes might be effective, given both demand-and supply-side drivers of the success of unreasonable party platforms.

Differences between Partisans and MDT's Costs
An important implication of the task that MDT assigns to partisans is that different partisans will have different roles to play. In turn, such differences will help us to understand the costs that partisans must endure in discharging MDT. In the literature on ethical partisanship, partisans are normally conceptualised as a broad group, extending beyond party members; they include local activists, think-tankers, journalists and others who might lack any formal affiliation to a party but pursue largely its same agenda and are interested in influencing its future direction (White and Ypi, 2016: 26-30). Consequently, different partisans of the same party occupy widely different positions, ranging from elected officials to members of the party's central office, broadly aligned journalists, mere members and so forth.
Differently positioned partisans have to do different things to succeed in collectively transforming a party's platform as prescribed by MDT. We leave for another day a full specification of the roles of different partisans, also because it would require entering debates over how best to organise a party to pursue certain goals. However, to clarify how different those roles might be, let us first think of a composite group of partisans, including think-tankers, bloggers and public intellectuals, whose main task is to lead the ideational work involved in MDT, perhaps especially at the early stage of identifying fresh ideas that might serve as the basis for their party's transformed conception of justice. This role is very different from, say, that of a party's national leadership, whose main tasks include placing the search for a different vision high on the party's agenda and, once this vision has been developed, working hard to convince reluctant fellow partisans of its value.
The idea that different partisans should play different roles also helps us to highlight how MDT can involve considerable costs. Such costs confirm that the transformation of society's public reason is indeed a matter of duty, not mere self-interest for partisans who want their parties to succeed. Let us consider again a party's public intellectuals. Some of them might be strongly associated, both within the party and in society at large, with promoting if not developing the current ideological platform that, based on MDT, should now be transformed. Even if they were convinced by the need to fulfil MDT, their work towards a different ideological platform might look untrustworthy. Consequently, those public intellectuals might need to accept that they must step out of the limelight, into a much-reduced supporting role. Turning to the party's national leadership, some rotation might also be required. Other costs for them include the risk of straining if not severing long-standing ties of comradeship and friendship by leaning on fellow partisans reluctant to fall in line behind the party's revamped political platform.
The Rawlsian framework is helpful for reiterating that partisans can be morally required to perform MDT also when it is so costly. Such framework links the containment of unreasonableness to the protection of liberal legitimacy, which is so fundamental among the values reasonable partisans must be loyal to that significant personal costs might well have to be endured in its service.

What Are the Circumstances under which MDT Applies?
Calling for a transformation of society's public reason to revamp its appeal, MDT applies only in circumstances where reasonableness has been losing support. Moreover, the objection to Quong we reconstructed in the 'Political Liberalism, PRR and Reasonable Partisanship' section, criticising any exclusive focus on last-ditch coercive containment, implies that those circumstances must be broader than 'real threats' to the liberal democratic order's stability. Signs of a process likely to result in a real threat must suffice, but how can we identify them? Our idea of a process has important similarities with that of 'deconsolidation' of liberal democracy, which has recently attracted attention in political science. We will now draw on the criteria defining when deconsolidation is occurring to specify the conditions under which MDT applies.
As defined by Stefan Foa and Yascha Mounk, the question of deconsolidation refers to the future 'durability' of liberal democratic rule in a country rather than the 'extent' to which that country is liberal democratic today (Foa and Mounk, 2017: 10). A country that still counts as liberal democratic and whose institutions are under no immediate threat might be deconsolidating if a process is under way that makes it progressively more likely that the liberal democratic order will one day break down.
For Foa and Mounk, three criteria must be jointly satisfied for deconsolidation to be occurring. Analysing empirical data, they suggest that those criteria apply in many countries in 'the West' and around the world, signalling that liberal democracy is deconsolidating in all of them. If this specific suggestion is correct, reasonable partisans are bound by MDT in those countries. However, we do not need to embrace it. Even if their empirical analysis was incorrect, their criteria would still be theoretically useful in specifying the general conditions under which MDT arises. 9 First, it must be the case that 'a sizable minority of citizens loses its belief in [liberal] democratic values' (Foa and Mounk, 2017: 9), for example, coming to doubt that a democratic government or civil rights are essential. Second, 'a sizable minority [. . .] becomes attracted to authoritarian alternatives' (Foa and Mounk, 2017: 9), like a strongman leader who does not have to worry about parliament. Third, 'political parties and movements with real power' do not all share a commitment to liberal democracy (Mounk, 2018: 104). As a possible specification of this criterion, think of cases in which populist radical-right or other unreasonable parties have joined coalition governments, especially as major players.
Having defined MDT, its justification and the conditions under which it applies, we will now further elaborate what it means to change public reason as required by it. Realworld examples of partisan action in the face of PRR will provide illustrations of different elements of MDT's general content.

How Partisans Should Change Public Reason
MDT prescribes that reasonable partisans steer away from two unpalatable extremes. First, when losing support to other parties' unreasonable messages, they must resist the temptation of integrating those messages into their platform, even after whitewashing them to make them sound less extreme. The strategy that yields to this temptation, which we call 'unreasonableness-lite', has often been adopted by mainstream parties reacting to PRR, as exemplified by hostility towards Islam. Mainstream partisans adopting an unreasonable-lite language might not outrightly insult Muslim citizens. However, as exemplified by repeated comments by former French president Nicolas Sarkozy during the 2016 Les Républicains presidential primaries, Islam is unreasonably depicted as posing unique 'difficulties' for the country, created by an alleged lack of effort from Islam at large to integrate into society (Chassany, 2016). To give another example, Mette Frederiksen, the leader of Socialdemokratiet, was widely described as adopting the anti-Muslim rhetoric typical of PRR in the run-up to the 2019 Danish general election. Frederiksen described Islam as a 'barrier to integration', proposing or otherwise supporting discriminatory policies, such as closing all Muslim but no other religious schools (Orange, 2018).
Given that MDT's goal is to combat the power of unreasonable views in society, our rejection of unreasonableness-lite might come as a surprise. If whitewashed anti-Muslim rhetoric could greatly help a mainstream candidate win an election against an out-and-out populist radical-right competitor, it would seem perfectly permissible to opt for the lesser evil of unreasonable-lite messages, especially if the mainstream candidate is not planning to turn them into policies. However, MDT must be understood from within our broader political liberal framework, which imposes certain constraints on it.
Singling out all members of an ethnic or religious minority as posing a problem to social cooperation is a key feature defining PRR as unreasonable. Normalising the problematic generalisations that define unreasonable language is off limits whatever its payoff, because on a political liberal theory, built on the rejection of unreasonableness, there can be no moral licence to employ unreasonable messages. This is most clearly encapsulated in the fact that if those messages were used during electoral campaigns, they would formally violate Rawls' duty of public reason (Rawls, 2005: 213-216).
Put differently, political liberalism shares with Rawls' earlier work a commitment to the 'priority of right', that is, the idea that individuals should not cross certain limits when pursuing their projects (Rawls, 2005: 173-211; see also Maffettone, 2010: 19-21). In political liberalism, such limits are imposed by reasonableness, for example through the duty to refrain from supporting any important decisions political decision-makers find appealing solely on the basis of comprehensive reasons. Although Rawls is primarily interested in how reasonableness and, in turn, public reason impose limits on the pursuit of individuals' comprehensive doctrines and, therefore, their conceptions of the good, they also effectively constrain purely political projects -specifically, by forbidding that any such project be publicly backed by unreasonable arguments.
Second, reasonable partisans must avoid 'reasonable paralysis'. When losing supporters to political alternatives they consider beyond the pale, like PRR, reasonable partisans might be tempted to take a position of moral superiority by ignoring the lost support. Therefore, they might stick to the same reasonable political platform they have campaigned or governed on thus far. An example of reasonable paralysis is provided by Partito Democratico, which only received 18.9% of the vote in the 2018 Italian general election after peaking in the 2014 European elections with 40.8%. For long after the defeat, many commentators, including some close to the party, criticised Partito Democratico's inability to advance any alternative political proposal that might reverse its decline, caught in an endless stalemate among its undercurrents (e.g. Esposito, 2019).
Political liberalism highlights how threatening growing support for unreasonable ideas is for stability and legitimacy, proving that reasonable paralysis is no position of moral superiority. Given partisans' ability to build new consensus through their creative capacities, reasonable partisans should exercise them, breaking with their party's current message while staying within the limits of reasonableness. Overcoming reasonable paralysis can be costly for many partisans, if the party's present platform encapsulates their favoured conception of justice or, at least, a close approximation to it. In this case, MDT requires that they support in practice a different conception that, although reasonable, is less appealing to them. Still, legitimacy as reasonable rule, which MDT is meant to protect, is so fundamental among the values reasonable partisans should be loyal to that it might well take priority over these sorts of costs.
Like unreasonableness-lite, MDT requires change but, like reasonable paralysis, it remains within reasonableness. We will now explain that partisans occupying different positions on the left-right spectrum should seek this change by exploring different visions for society. Moreover, in developing such visions, partisans of different parties should focus on different explanations of the support currently enjoyed by unreasonable competitors, to strive to understand how to win back supporters from them.
Rawls appreciates that there is a plurality of reasonable conceptions of justice, underlining the importance of an 'orderly contest' among them for support in society (Rawls, 2005: 227). This contest is fought by actors including a multiplicity of reasonable partisans, which we have all called on to discharge MDT. Consequently, MDT prescribes a distinctly combined containment effort, where partisans of various progressive and conservative parties all play their part by renewing their messages. This marks an important difference with recent work in political theory that appears to assign a role only to the renewal of social democratic ideology broadly construed (e.g. Fraser, 2019;Mouffe, 2018).
Consider again the case of PRR. The pluralism of partisans subject to MDT can be connected with the multiple existing explanations in political science for the current demand for PRR, which have led many to note that there is not one, but many electorates of PRR (Arzheimer, 2018;Mudde, 2007: 201-231). Reasonable partisans of different parties can transform their platforms by focusing on different explanations of the support for their unreasonable competitors, so that, in this case, different parties can hope to win voters from different electorates of PRR.
It is not our role to prescribe to any real-world reasonable partisans how exactly they should transform their political platform to tackle PRR -even assuming, for argument's sake, that the conditions under which MDT applies are satisfied in their country. Our argument has stressed partisans', not political theorists', creative abilities. Moreover, any candidate as transformed party platform should be evaluated contextually. However, we wish to mention two general directions of change that (at least some kinds of) progressive and conservative partisans fulfilling MDT might explore to see for themselves whether they are worth pursuing.
One possibility worth considering for broadly social democratic partisans is to draw on the prominent 'economic losers of globalisation' explanation of PRR's success (Betz, 1994;Kriesi et al., 2012). In brief, this explanation highlights that many workers, who, for example, might lack flexibility or the right sorts of specialised skills, were badly hit by the transition from industrial to post-industrial societies. Consequently, they are unwilling to vote for mainstream parties that appear to embrace such transition. In light of this explanation, social democratic partisans might explore whether a socialist platform, promising to attract voters drawn to PRR because of their socio-economic circumstances, is the way to go.
As a concrete example of such platform, take the vision for society behind the Green New Deal (GND). The GND, as formalised in the resolution presented in February 2019 by Alexandria Ocasio-Cortez and Ed Markey, is the product of the work of American politicians from the Democratic Party and organisations that are either internal or close to it, such as Justice Democrats and Sunrise Movement. GND's goal is to decarbonise the US economy while actively addressing in the process existing injustices against racial minorities and other disadvantaged groups. Importantly, the GND includes a 'federal job guarantee' ensuring good pay and benefits and ring-fenced public investments for the benefit of locally-owned firms, to support the transformation of the energy, manufacturing and transport sectors to render them powered by clean energy (Gunn-Wright and Hockett, 2019). Such policy proposals reflect a transformed conception of justice, bringing concerns for the environment together with a socialist commitment to economic redistribution and the goal of urgently redressing substantive racial inequalities. 10 Rawls himself mentions an ideological tradition that, in our view, provides a possible example of an intellectual resource to be considered by reasonable conservative partisans performing MDT. For him, the neo-Thomistic tradition of investigating 'common good and solidarity' can be expressed 'in terms of political values' and provide a reasonable conception of justice (Rawls, 1997: 775). This is the tradition normally labelled 'Christian democracy', which has recently enjoyed renewed academic interest. Carlo Invernizzi Accetti analyses many intellectual resources provided by such ideology, including a Christian democratic vision of how to limit capitalism that is different from social democracy. One pillar of this vision is a sufficientarian commitment to everyone having enough that, for example, could provide strong support to universal basic income (Invernizzi Accetti, 2019: 242-243).
Invernizzi Accetti also discusses PRR. Drawing on empirical research, he suggests that religiously-motivated Christian constituencies often largely support PRR, whose partisans are generally eager to express their religious beliefs in politics. However, a large part of such constituencies only supports PRR as a costly ideological compromise and for lack of an alternative more consistent with their beliefs. Given Christianity's universalistic message, PRR's exclusionary unreasonable features contribute to this tension (Invernizzi Accetti, 2019: 335-337). This religious explanation for some of the demand for PRR seems worth exploring for reasonable conservative partisans. As argued by Invernizzi Accetti, if properly refreshed, Christian democratic ideology seems wellplaced to attract voters from this and perhaps other electorates of PRR.

Some Hope for MDT's Success
This section addresses two interrelated worries. The first worry is that we have not drawn attention to any ground for affirming that MDT would be effective in turning citizens away from unreasonableness. Our reply is threefold. First, this is a work in political theory, which places any conclusive evidence that MDT would work in practice beyond the scope of our argument. Second, at least two points we have made actually create some hope for MDT's success. The 'Political Liberalism, PRR and Reasonable Partisanship' section reconstructed partisans' characteristic ability to reshape public opinion, and reshaping public opinion (so that unreasonable views become less popular) is precisely what MDT is aimed at. Also, the 'How Partisans Should Change Public Reason' section specified that a plurality of partisans on the left-right spectrum is required to transform their platforms by tapping into different explanations of the success of their unreasonable competitors. In part, our hope for MDT's effectiveness is based on the simple fact that the required change to each platform is strategic, using existing research to understand how best to intercept the demand for unreasonable competitors while staying within reasonableness. Moreover, the fact that MDT applies to a variety of partisans becomes important whenever unreasonable parties have many different electorates, as with PRR. We do not expect partisans of any single party to find the key to swaying all those electorates. More realistically, we hope that the aggregate effect of disparate attempts by a plurality of reasonable partisans might be to attract significant support away from unreasonableness.
Before turning to our third point, we aim to address the possible objection that MDT's ability to mobilise a plurality of partisans targeting different electorates of unreasonable parties can only be fully exploited in proportional electoral systems, characterised by many different parties. To a certain extent, we are ready to accept this point, which reveals an interesting theoretical implication of our argument. Indeed, the observation that the two large parties normally dominating majoritarian systems are less well-equipped to employ the precious resources for containment identified by our Rawlsian framework provides an original pro tanto reason that a proportional system fits better within Rawlsian political liberalism. 11 Moreover, this objection underappreciates that even in majoritarian systems, parties are not internally homogeneous and different groups within each party have some agency of their own. Different partisans from the same party can speak with somewhat different voices, attempting to reach out to different electorates of unreasonable competitors. For example, consider the 'presidential tickets' chosen ahead of US presidential elections. If the three criteria specifying when MDT applies are satisfied, partisans close to the presidential nominee and the nominee themselves for the party in question could strive to choose a running mate who can appeal to different potential constituencies of unreasonable competitors from those that the nominee most naturally speaks to. This point can be extended to other selection processes for highly visible offices. For instance, if the party wins both the presidential election and the majority in the House of Representatives, its partisans could try to get a Speaker elected who is in turn somewhat different.
The third point zeroes in most closely on PRR as the case we use to illustrate MDT. There is literature in political science, studying the supply-side determinants of PRR's success that chimes with our general recommendations about containment. Existing scholarship on 'political opportunity structures' -that is, environmental factors external to populist radical-right parties that affect their success -pays considerable attention to mainstream parties' behaviour towards each other and populist radical-right competitors (Mudde, 2007: 237-243). Two influential positions are that (1) convergence among and (2) copy-catting by mainstream parties and leaders constitute important opportunity structures facilitating PRR's breakthrough and persistence. No position in this debate is beyond criticism, and this is not the place to prove empirical hypotheses right. However, insofar as (1) and (2) are sound, they support the idea that MDT might significantly contribute to the containment of PRR if MDT is triggered by it.
'Convergence' refers to the idea that voters dislike situations where mainstream parties adopt ideologically similar platforms, with PRR consequently gaining electoral support (Arzheimer and Carter, 2006;Kitschelt and McGann, 1995). Some scholars suggest that the present state of many 'Western' countries is one of convergence, especially over socio-economic matters, with mainstream parties described as forming a neoliberal consensus, agreeing on globalisation as essentially positive, or similar. For them, this state is an important contributor to PRR's recent electoral success (e.g. Hobolt and Tilley, 2012;Kriesi et al., 2012). Now, if reasonable parties' messages are indeed converging, performing MDT would end this state because MDT prescribes that reasonable partisans provide alternatives to those messages. If the convergence thesis is correct, MDT would remove an important political opportunity structure favouring PRR.
'Copy-catting' refers to mainstream parties, especially from the centre-right, adopting frames, rhetoric or policies typical of PRR to gain an electoral advantage. According to some scholars, by making their platform even more salient, copy-catting actually increases support for populist radical-right parties, either in the short or in the medium term (Bale, 2003;Dahlström and Sundell, 2012). As suggested by the wider literature on the risks posed by the polarisation of mainstream parties in a broadly nativist direction, even where short-term electoral gains materialise, mainstream parties by nature struggle to implement their polarised platforms and will likely circle back to more moderate positions. Having by then normalised those platforms, they thereby create the chance for PRR to make large gains (Ignazi, 1992). 12 Specified in contrast to unreasonableness-lite, MDT outrightly forbids copy-catting, which is, for some, a significant contributor to PRR's success. Like the one on convergence, the literature on copy-catting therefore creates some further hope that if required, MDT might contribute to undermining support for PRR. 13 The second worry we discuss in this section specifically concerns the current state of mainstream parties in 'Western' countries. According to the prominent 'cartel-party thesis', the second half of the twentieth century saw a progressive movement of parties towards the state, with professionalised party elites striving to hold on to public office and minimise the loss of state funding generated by electoral defeat. A key consequence of this movement is that the mere members constituting the 'party on the ground' became less and less influential. In turn, this appears to have compromised parties' ability to influence civil society, which MDT relies on (Katz and Mair, 2009). This seems to suggest that if MDT is triggered by PRR in present-day societies, it is bound to be hopeless.
Our reply draws on other theorists of ethical partisanship, who also tackle this worry that existing parties can no longer act as transmission belts between state and civil society. White and Ypi underline the possibility of creating new parties on the back of a desire to break with the cartel form, reconnecting with activists on the ground and, in turn, civil society. Such new parties can at least partially be the work of tired partisans of existing cartelised parties themselves (White and Ypi, 2016: 72). Fabio Wolkenstein stresses that well-established parties should not be written off either, as exemplified by British Labour under Jeremy Corbyn, which saw partisan grassroot organisation Momentum exercise great influence in the party (Wolkenstein, 2020: 7-8). According to Wolkenstein, these positive developments can be facilitated by implementing intra-party deliberative measures, which we welcome, aimed at reconnecting parties with civil society (Wolkenstein, 2020; see also Ignazi, 2017: 233-260).
Moreover, the inclusive definition of partisans that we have adopted provides further help. Partisans might have no formal affiliation to their party, suggesting that partisans and parties must not be conflated. Given that partisans include members of grassroots organisations, public intellectuals, journalists and others who might simply share a party's commitments and be interested in influencing it, MDT would mobilise actors with agency and power that are largely independent from the party machine strictly understood. Such actors could therefore throw their considerable weight behind the strategies for addressing cartelisation that we have just discussed. 14

Sincerity and Strategic Behaviour Reconciled
In sum, at least when liberal democracy is deconsolidating, reasonable partisans haemorrhaging support to unreasonable competitors should refashion their messages in an appealing direction without mimicking those deeply problematic competitors. This recommendation might seem highly intuitive, and the previous section suggested that there is some hope for its success. However, our discussion of unreasonableness-lite showed that political liberalism imposes strict side constraints on political action. Therefore, that recommendation might seem unavailable to political liberals because unethical. Our claim that reasonable partisans should develop new reasonable platforms that have a good chance of winning back supporters involves that many partisans should practically support a vision for society they do not regard as fully just. Asking partisans to constantly clarify that the platform they campaign on is not their preferred one risks being selfdefeating. Excluding this, MDT seems to violate norms of sincerity that Rawls, along with many other deliberative democrats, holds dear (e.g. Habermas, 1984: 318-320). This section aims to highlight the distinctive features of Rawlsian sincerity, which lead to the surprising conclusion that reasonable partisans' strategic behaviour does not violate it and is therefore ethical.
The key here is that Rawlsian political liberals require that public reasoners sincerely believe that the arguments they provide are reasonable, not correct. In Rawls' words, 'we should sincerely think that our view of the matter is based on political values that everyone can reasonably be expected to endorse' and also that 'the particular view we propose [. . .] expresses a reasonable combination and balance of the political values' (Rawls, 2005: 241; see also l and 253). 15 Further specifying Rawls' definition, Micah Schwartzman claims that sincerity is satisfied when a citizen believes that her proffered arguments are both public (i.e. based on those reasonable political values) and 'sufficient' to justify their conclusions (Schwartzman, 2011: 385). Schwartzman draws on Gerald Gaus to define sufficiency, which sets a much lower bar than 'best justified' argument, essentially excluding obvious mistakes in reasoning (Gaus, 1995: 253). To be sincere, public reasoners need neither to be motivated by the arguments they propose nor to actually believe them (Schwartzman, 2011: 387-390 and 386 fn. 34, respectively).
Quong endorses an analogous understanding of sincere public reasoning as belief that one's arguments are justified to oneself and others as reasonable persons. By providing justifications to others, we show respect to and avoid manipulating them. Specifically, as clarified by Quong when using sincerity to criticise 'convergence' approaches to public reason, these values are only satisfied when justifications are ultimately grounded in reasonable political ideas of society and persons that the reason-giver accepts (Quong, 2011: 265-273).
Remaining within reasonableness, MDT satisfies those values and Rawlsian sincerity in general. The values highlighted by Quong also allow us to explain why MDT does not need to work like Rawls' 'reasoning from conjecture'. Falling outside public reason, conjecture involves reasoning from within our interlocutor's comprehensive doctrine, aiming to show that their doctrine is consistent with reasonable political values. Reasoning from comprehensive premises we do not share, conjecture can only avoid manipulation if we disclose 'that we do not assert the premises from which we argue' (Rawls, 1997: 787). Building on basic reasonable political ideas they accept, partisans fulfilling MDT constitute a different case and can therefore be respectful and non-manipulative without any further disclosure.
Rawls' understanding of sincerity might seem idiosyncratic but fits perfectly with a basic tenet of Rawlsian political liberalism, which famously 'does without the concept of truth' (Rawls, 2005: 94). According to Rawls, given that truth is singular, asking citizens to regard the truth of a consideration as either necessary or sufficient for its admissibility into their public reasoning would be sectarian and divisive (Rawls, 2005: 116 and 129). Therefore, even those Rawlsian political liberals who aim to carve out a role for truth agree with Rawls that reasonableness, which is plural, should be the sole standard regulating inclusion into public reason. David Estlund explains that public reasoners 'need not believe' the arguments they use, only that they are 'acceptable to all reasonable citizens' (Estlund, 1998: 271). Similarly, Joshua Cohen claims that 'what matters for democracy's public reason is reasonableness, not truth' (Cohen, 2009: 31). Partisans performing MDT will likely deviate from what they see as the truth about justice but will still propose a reasonable view of it, which is all that matters. Rawls' framework is therefore extremely well-suited to justify forms of strategic behaviour we intuitively expect of partisans protecting liberal legitimacy and other crucial values.

Conclusion
This article has used PRR to enrich Rawlsian political liberalism, so as to deal with the underexplored scenario where reasonableness is losing supporters. It has brought together for the first time political liberalism, populism and ethical partisanship as discussed in political theory while at the same time engaging with relevant political science literature. The result has been a duty for reasonable partisans to collectively change the content of society's public reason, by strategically transforming the visions for society their parties campaign on in ways they can hope might intercept at least part of the demand for unreasonable competitors. We have highlighted how Rawls' distinctive understanding of sincere public reasoning provides the necessary resources to vindicate this strategic behaviour as ethical. Moreover, in developing our duty, we have made original points about political liberalism at large, for example, by identifying the full extent of public reason's dynamism and a novel source of appeal of Rawls' characteristic appreciation of reasonable disagreement about politics. Notes 1. Rawls' deliberative account of democracy is procedural in that democratic institutions are meant to implement intrinsically important values like equal liberty or autonomy (Saffon and Urbinati, 2013). However, Rawls' hints at preventive containment distinguish him from the specific procedural approach associated with Hans Kelsen, according to which democracies should never limit the political rights of anyone, including democracy's enemies (Kelsen, 2006). 2. This criticism also applies to theories of 'militant democracy', which argue for repressive state intervention to protect democracy from political extremists. For example, see Kirshner (2014). 3. Some might object that mere supporters of PRR do not count as unreasonable; different parts of the electorate vote for PRR for different reasons, suggesting that many supporters might not share PRR's core ideological beliefs (Mudde, 2007: 225-226). However, being unreasonable is essentially a practical matter. Rawls does not describe unreasonable persons as those failing to hold certain political beliefs.
Unreasonable persons fail to act consistently with them, for example by neglecting to 'propose' and 'abide by' fair terms of cooperation among free and equal persons (Rawls, 2005: 49). Given that PRR's message is unreasonable, the practical acts of voting and showing support for it, especially over time, suffice for citizens to count as unreasonable. 4. This neglect of partisanship also characterises the broader debate over democratic self-defence. While this debate was split for long time between the supporters of militant and procedural democracy, several scholars now explore a middle space between the two, investigating measures that actively oppose extreme views while minimising repression (Brettschneider, 2012;Malkopoulou and Norman, 2018;Rummens and Abts, 2010). The project of exploring that middle space is appealing, but thus far, its proponents have overlooked the role of partisans committed to liberal democratic values. An important exception is provided by Stefan Rummens and Koen Abts's model of 'concentric' containment, stressing the importance of a cordon sanitaire against illiberal and antidemocratic parties that reach parliament (Rummens and Abts, 2010: 655-663). However, building alliances against competing parties is but one activity that partisans can undertake. In particular, they are well-placed to reshape society's public opinion and even its 'common sense' (White and Ypi, 2016: 67), creating the room to theorise how to change the minds of supporters of problematic political ideas. These creative abilities are what our proposal concentrates on. 5. For two classic explorations of ethical partisanship, see Muirhead (2014) and Rosenblum (2008). 6. Waldron (1993). On public reason's mutability, see also Flanders (2012). 7. We thank an anonymous reviewer for suggesting this final justification. 8. Readers might wonder whether it is really necessary to restrict MDT to partisans. In other words, why are not all reasonable citizens bound by it? Transforming the reasonable visions for society that have wide currency in political debates can only be a collective effort. Therefore, assigning MDT to all reasonable citizens as individuals would mean making an impossible demand on them. To solve this problem, it could be suggested that reasonable citizens have an obligation to become partisans of reasonable parties (or alternatively to join reasonable social movements). Although intriguing, we find this suggestion inconsistent with the typical liberal wariness, which political liberalism shares, to place heavy demands on citizens regarding political participation (e.g. Badano and Nuti, 2018: 167). 9. For some criticisms of Foa and Mounk's empirical analysis, see Alexander and Welzel (2017) and Voeten (2017). 10. We have discussed revamped socialist platforms merely as one possibility that progressive partisans might consider, exploring whether such platforms are worth pursuing in their country's specific context. However, readers might object that the November 2020 victory of Joe Biden over populist radical-right incumbent US President Donald Trump proved that socialist platforms are not even worth considering. This would be because the platform he ran on did not depart nearly as much as the GND from the centrist ideology characterising the last few decades of the Democratic Party. However, at least at this stage, one should be cautious in drawing any general lessons from Biden's victory. First, the 2020 US presidential election and its preceding campaign occurred under truly exceptional circumstances, that is, COVID-19 and Trump's disastrous management of the pandemic. According to several commentators, the electoral outcome would have likely been different if the pandemic had not happened (e.g. Parker et al., 2020). Second, on our account of reasonable partisanship, since the victory of any unreasonable competitor would diminish liberal legitimacy, winning an upcoming election is important. However, it is not everything, because reasonable partisans are also committed to the stability of legitimate rule. Therefore, elections should not be won in a way that risks undermining reasonable partisans' ability to keep winning in the future. We are not suggesting this is clearly the way in which Biden won, but it might be; precisely because Biden's platform does not address the deepest (including socio-economic) determinants of much