Plural reconstruction: A method of critical theory for the analysis of emerging and contested political practices

In this article, I argue that Habermas’s method of rational reconstruction faces limitations when it comes to analysing newly emerging and contested political practices. As rational reconstruction aims to criticize existing practices by determining their normative meaning as reflected in the participants’ idealizing presuppositions, it reaches its limits where emerging and contested practices make it impossible to identify a shared self-understanding and a single participants’ perspective. Using the example of membership politics, I argue that this is often the case where nationally constituted forms of politics become controversial or are fundamentally questioned. Building on the work of Benhabib and Fraser, I develop an alternative reconstructive method of plural reconstruction, which modifies the basic premises of rational reconstruction, adjusting it to emerging and contested political contexts.

First, it discusses whether the method of rational reconstruction is suitable for a critical analysis of emerging or contested practices whose fundamental principles are challenged, using the example of membership politics as a test case. Membership politics, that is, political decision-making on the boundaries of political communities, is traditionally a 'domestic' issue, which is, however, increasingly transnationalized and politically contested (Bauböck 2017;Lafont 2019, 355). Second, building on the works of Seyla Benhabib and Nancy Fraser, the article develops a new reconstructive method called 'plural reconstruction', which modifies basic premises of rational reconstruction, adjusting it to the requirements of emerging and contested practices.
The method of rational reconstruction as developed by Habermas aims to criticize existing practices by reconstructing their normative meaning embodied in the participants' idealizing presuppositions. But what do we mean by 'existing practices'? When it comes to politics, Habermas focuses on institutionalized practices, as enabled by constitutions or treaties, such as constitutional democracy or the politics of the European Union (EU). Increasingly, however, many of the most pressing political problems are negotiated in contexts that are not (yet) in the same way institutionalized or whose legitimacy is fundamentally challenged. When we look at issues such as migration or climate change, for example, we find political practices where transformative change is currently taking place: established institutions are contested, democratic innovations emerge, legal norms are in flux.
It is unclear what the emergence of new practices and the contestation of existing practices imply for the method of rational reconstruction. Should we still only reconstruct the normative potential inherent in dominant institutional structures and legal documents or also include new forms of protest or emerging counter-narratives? The challenge that arises here is to delineate in a non-arbitrary way the practices that are to be reconstructed. Thus, emerging and contested practices give rise to the following dilemma: Either a political practice is delineated narrowly (focusing on established institutions), which might lead to a problematic status quo bias, or it is delineated broadly (including contestatory and newly emerging elements), which threatens to undermine the method's key assumption -that we can identify a shared practice with constitutive rules and a single participants' perspective. Does this mean that rational reconstruction is unable to deal with emerging and contested practices?
My claim is that the method is ill-equipped to deal with practices that are only just beginning to emerge or where transformative change is currently taking place. As I will argue, rational reconstruction is incapable of criticizing such political practices without either showing a problematic status quo bias or undermining the method's key assumptions. However, this does not mean that we have to bid farewell to reconstructive approaches in general. I argue that in Benhabib's and Fraser's contributions to democratic theory, we can identify the building blocks of an alternative reconstructive method, which I describe as 'plural reconstruction'. The method of plural reconstruction includes four steps: (1) Remapping the political landscape: It starts by describing and framing a changing political context, focusing on both established institutions and emerging practices. (2) Reconstructing plural presuppositions: In the next step, the presuppositions of these practices are reconstructed from different participants' perspectives, including different social positions. (3) Reconstructing emancipatory potentials: Subsequently, emancipatory elements in existing political contexts are identified that point beyond the dominant institutional set-up. (4) Re-imagining institutions and the division of labour between theorists and citizens: Finally, these elements are systematized into an institutional proposal, within which the reconstructed plural perspectives can be brought into dialogue. The method of plural reconstruction serves both to criticize existing political conditions and to sketch ways in which they can be overcome by the participants themselves.

The method of rational reconstruction
In general, reconstructive methods can be described as 'a form of critique that aims to employ normative potentials [that] transcend the agreed-upon norms of a society, but are, in some way or another, nevertheless already "immanent" in social reality' (Stahl 2013, 534). In the following, I will focus on Habermas's method of rational reconstruction, which has recently received renewed attention in the context of political theory's 'methodological turn' (Gaus 2016;Iser 2008;Patberg 2014;Pedersen 2008; on the methodological turn, see Erman and Möller 2015). The aim of rational reconstruction is to identify 'the implicitly assumed normative contents of empirically established practices [ . . . ] from the participant perspective, i.e. in a performative attitude' (Habermas 2012, 291, n. 16). My focus is on rational reconstruction of political practices such as constitutional democracy, EU politics or global constitutionalism. 1 The method of rational reconstruction specifies in two respects the broad definition of reconstructive methods introduced above. First, Habermas has a precise idea of the 'normative potentials' that existing practices are supposed to possess. He tries to 'identify particles and fragments of an "existing reason" already incorporated in political practices, however distorted these may be' (Habermas 1996, 287;cf. Cooke 2012, 814). Thus, the focus is on the normative content that is already embodied in a practice, that is, on the idealized meaning of the practice from the participants' perspective. In discourse theory, this means that we have to identify (counterfactual) assumptions that participants need to make to make sense of a specific practice. In Between Facts and Norms, Habermas explains: A reconstructive legal theory follows a methodology premised on the idea that the counterfactual self-understanding of constitutional democracy finds expression in unavoidable, yet factually efficacious idealizations that are presupposed by the relevant practices. (Habermas 1996, 462) For a reconstruction of the democratic constitutional state as an institutionalized political practice, a variety of more specific practices are relevant, such as the electoral act, parliamentary debates or the written constitution. For Habermas, rational elements can sometimes be found in what seems like irrational behaviour at first glance. As he explains referring to the 'voter's paradox', when people cast their vote although it is explained to them that their individual vote won't make a difference, this reflects the participants' idealizing assumption that democracy is a practice of free and equal citizens where 'every vote counts'. Citizens' behaviour thus points at the rational (in the discourse-theoretical sense of discursively justifiable) core of the practice (Habermas 2006a, 413;Patberg 2018, 114). In this way, Habermas seeks to identify 'the normative self-understanding of the legal system' (Habermas 1996, 69;cf. Fossen 2015cf. Fossen , 1078. In a second specification of the general definition of the reconstructive method, rational reconstruction proposes a procedure meant to uncover constitutive rules or presuppositions of practices. A 'thought experiment' is to be carried out which simulates a rational discourse between the participants of a practice (Habermas 1996, 132). The aim of this thought experiment is to make 'explicit their intuitive knowledge of the performative meaning of this practice' (Habermas 2001, 777). To explicate the meaning of the democratic constitutional state, Habermas proposes the thought experiment of a fictitious founding act: 'Similar to its social-contract predecessors, discourse theory simulates an original condition: an arbitrary number of persons freely enter into a constitution-making practice' (Habermas 2001, 776). In Habermas's legal theory, it is the 'system of rights' that spells out these assumptions: With the system of rights, we have assured ourselves of the presuppositions that members of a modern legal community must take as their starting point if they are to consider their legal order legitimate but cannot base this legitimacy on religious or metaphysical arguments. (Habermas 1996, 132) This step is crucial to understand why a rational reconstruction can serve to criticize existing practices. The results of the hypothetical discourse are meant to show what the practices would look like if they fully realized their normative potential: 'The reconstruction of tacitly made counterfactual assumptions provides an objective evaluation standard that is rooted in the observed practices themselves' (Habermas 2008a, 151, my translation). The method of rational reconstruction thus seems to be 'rational' in two respects. First, it aims at identifying rational elements already embodied in practices, and, second, it 'purifies' and spells them out using a rational procedure. 2 Habermas does not only employ the method of rational reconstruction at the domestic level but also uses it for political contexts beyond the state: in the context of the constitutionalization of international law (Habermas 2008b) and in the context of the EU (Habermas 2012). 3 In both cases, he aims to identify conditions of legitimacy for these postnational orders. He closely follows the procedure of rational reconstruction as previously applied to the domestic realm. In the context of world society, Habermas imagines a 'thought experiment' in which 'two categories of founding subjects' -nation states and citizens -come together to decide on the outlines of a global constitution (Habermas 2008b, 449, original emphasis). This thought experiment leads to a description of the institutional, procedural and substantive features of a global constitution and to an overview of the 'conditions under which [it] could be democratically legitimated without assuming the character of a state' (Habermas 2008b, 451). In the European case, Habermas argues that looking at the EU 'from the perspective of a rationally reconstructed constitution-building process' (Habermas 2012, 26, original emphasis), we can see it 'as if it had been created for good reasons by two constitution-founding subjects endowed with equal rights' (Habermas 2012, xi). Habermas's goal is to show why the current institutional design of the EU, most importantly the role of citizens and peoples, is justified: A suitable way of clarifying the constitutional and legal structure of this peculiar formation is to reconstruct its history of emergence, interpreted in teleological terms, as though the more or less contingent historical outcome had been the deliberate result of a regular constitutional convention. (Habermas 2012, 31) What these examples show is that Habermas, in applying rational reconstruction to national as well as international politics, always focuses on those practices or elements of these practices that are well institutionalized and clearly legally circumscribed, for example, by the UN Charter or the EU treaties. In the next section, I will point to a series of objections against rational reconstruction and argue that the method faces limitations -especially once we turn to practices that are newly emerging, fundamentally contested or undergoing transformative change. Before I do so, however, I would like to address a common misconception concerning rational reconstruction's supposed conservative tendencies.
Some critics suggest that the method is unable to guide the fundamental critique of problematic political practices because it tends to defend them with reference to their rational core, understood in a discourse-theoretical sense. In particular, there is the worry that the method is inadequate when it comes to regressive practices, as it abstracts from existing contexts and seems to overly idealize their empirical forms. If we look at current examples from the area of membership politics, such as practices of denationalizing unwanted citizens, of selling citizenship or of securitizing access to citizenship, it might be objected that the idealizing assumptions that participants have to make in order for the practice to appear meaningful to them are not justifiable at all. Is there, we might ask, a problematic 'statesman-like optimism' at work in Habermas's legal and democratic theory (Benhabib 1997, 726)?
It is important to note that not all practices can be rationally reconstructed. Instead, a practice needs to have a core of existing reason to justify the attempt to determine under what conditions the practice could be accepted by all participants in a rational discourse (Patberg 2018, 183). A certain 'rational minimum' needs to be met before we can employ the method. When we look at the current practice of unilateral denationalizations, it is not at all obvious whether this practice meets such a baseline (Lenard 2018). Just as Habermas does not rationally reconstruct the practice of election fraud in Between Facts and Norms but the more general practice of constitutional democracy -which would enable him to criticize election fraud -we should not assume that denationalization practices are an adequate object of rational reconstruction. Neither election fraud nor unilateral denationalizations can be rationally reconstructed as discursively justifiable. In other words, simple status quo bias objections to rational reconstruction do not hold water. However, while it is possible to criticize regressive phenomena with the help of the method, we would have to identify a broader practice whose performative meaning contradicts election fraud or denationalizations. I will address the challenge of delineating practices in the following section.

The method of rational reconstruction in the context of emerging and contested political practices
Despite its merits, the method of rational reconstruction has been subject to critique for some time. The main objection is similar to what the novelist Chimamanda Ngozi Adichie calls 'the danger of a single story' in the context of literature: single stories are 'not wrong, but they are incomplete' (Adichie 2009). Critics argue that rational reconstruction unduly focuses on a single interpretation of political practices, ignoring alternative viewpoints and counter-narratives (Honig 2009: 33-35;Olson 2003). Even in the context of clearly defined practices whose fundamental principles are relatively uncontroversial, such as constitutional democracy or the UN Charter as a 'world constitution', the method risks presenting too smooth a picture of their idealized meaning. According to this objection, rational reconstruction neglects the complexities and plural voices of political practices by focusing on a shared self-understanding and a single participants' perspective and by leaving aside the particularities and 'unruly practices' (Honig 2009, 4) of specific political contexts, reflected, for example, in protests and counternarratives. Furthermore, rational reconstruction, relying on hypothetical founding scenarios, also seems vulnerable to objections that have been raised against thought experiments as they are used in the social contract tradition more generally (Pateman and Mills 2007). The worry is that rational reconstruction results in an idealized interpretation or normative outlook 'too detached from the practices in which it claims to be based' (Olson 2003, 288).
While I do not think that these objections fully discredit the method of rational reconstruction, they point to important limitations. As I will argue now, the problems just mentioned amplify in contexts where there is no clear institutionalization or of practices whose legitimacy is fundamentally challenged. The method of rational reconstruction runs into problems in cases where it becomes implausible to identify a single participants' perspective. When we deal with emerging or contested practices, we encounter a plurality of perspectives, which makes it impossible to identify shared presuppositions. I argue that the main problem for the method of rational reconstruction in such contexts is to delineate in a non-arbitrary way the practices that are to be reconstructed. For example: Should we focus on the established elements of practices or also include contestatory movements criticizing these practices? When delineating a practice, that is, when determining the object of a rational reconstruction, the following dilemma arises. Either the practice is delineated narrowly, which can lead to a problematic status quo bias, or the practice is delineated broadly, which threatens to undermine the method's key assumption, that is, that we can identify a shared practice with constitutive rules and a single participants' perspective.
Let me illustrate this dilemma using the example of membership politics. Membership politics describes decision-making on the boundaries of political communities, including decisions on voting rights and citizenship. Where would we begin if our aim was to rationally reconstruct membership politics to criticize its various exclusionary and discriminatory aspects? In the last section, I already indicated that instead of singling out individual regressive elements, the method of rational reconstruction would have to take a look at the broader political practices of which they form part. But what would such a 'broadening' entail? The first option would be to look at the dominant legal and political practice of membership politics by analysing national constitutions and legal doctrine. Such an analysis would presumably reflect the normative idea of 'sovereign membership politics', according to which it is up to existing political communities to remake their boundaries. The practice of sovereign membership politics does indeed seem to have a 'rational core', as it realizes the idea of national self-determination and democratic selfgovernment: Who should be included and who should be excluded is decided by 'the people' (Miller 2016;Wellman 2008). One could argue that the fact that democratic states increasingly securitize their membership boundaries only confirms that their citizens share the idea that sovereign membership politics is legitimate, that is, that decisions about citizenship and voting rights should be in the hands of national politicians. Now, while this may be an accurate reconstruction, it builds on a limited picture of current membership politics and, more importantly, risks defending a problematic status quo. If the principle of unilateral self-determination of political communities were indeed the rational core of membership politics, what would it mean to 'fully realize' this principle? Is the existing practice perhaps not yet sufficiently exclusive? Based on the narrow interpretation of membership politics, the method of rational reconstruction seems to produce counter-intuitive conclusions. In this case, critics would be right to object that the method of rational reconstruction is unable to question current political developments.
But, again, this conclusion, which suggests that the method of rational reconstruction cannot fulfil a critical function with a view to current membership politics, seems too quick. After all, to reconstruct the current practice as sovereign membership politics is not our only option. Instead, we can take an even broader look at the practice of membership politics to include other types of political action by members and nonmembers, such as protests, challenges in the courts, media statements or solidarity movements. From this perspective, it would unduly limit the scope of rational reconstruction to focus exclusively on existing laws, citizens' demands or the actions of elected politicians -as a focus on the firmly institutionalized elements of a practice, which are usually taken as the most important indicator for the inherent normativity of a political practice, would require. If we take the 'post-sovereign' reality of today's membership politics into account, including elements such as protests against existing policies, we might find a rational core that does not simply affirm the ideal of unilateral self-determination of nation-state peoples.
Even though such a move would be in line with Habermas's method, it raises difficult questions regarding the status of emerging and contested practices. Of course, we might consider 'no borders'-demonstrations as part of the practice of membership politics -but what does that mean for the idea of a 'shared practice'? Can there be a thought experiment about membership politics that identifies the 'rational core' of both sovereign decisions on membership boundaries and of fundamental challenges to the legitimacy of this practice? Is there a normative difference, from the perspective of rational reconstruction, between institutionalized practices and emergent counter-practices? In today's membership politics, non-sovereign elements are less institutionalized and less consolidated. So far, there have only been a few cases in which domestic membership norms have been institutionally contested, for example, when EU institutions objected to the introduction of an investor citizenship program in Malta or when supranational courts declared certain membership norms incompatible with human rights (e.g. in the Hirst Case before the European Court of Human Rights on the right to vote for prisoners in the United Kingdom). The problems of the method of rational reconstruction seem to be especially challenging for practices that are only just beginning to emerge or where transformative change seems to be currently taking place.
Habermas is aware of the danger that the gap between the participants' idealizing presuppositions and the actual form of the practice may become so wide that the participants 'feel alienated from an established practice' (Habermas 2008a: 151, my translation). However, he has in mind processes of decay which have the consequence that a practice no longer lives up to the high normative demands of participants. Where institutional constellations are in flux due to challenges of their fundamental principles, the situation is more complex: the presuppositions of some participants still correspond to the 'old structures', while new practices are already emerging; some participants are contesting what they consider outdated ideals, while others defend these ideals against emerging counter-practices. In such situations, the critical goal of a reconstructive method cannot consist in realigning the practice with its underlying presuppositions. Instead, it is necessary to take a critical stance towards them.
However, this cannot be achieved within the method of rational reconstruction, where a key step consists in 'purifying' the participants' idealizing assumptions through thought experiments. In the hypothetical scenario of a rational discourse, it cannot be determined whether the constitutive rules that the participants must assume in order for the practice of membership politics to appear meaningful to them are not themselves problematic. The normative and methodological premises of the idea of the thought experiment reflect the idea of contract theories that have been problematized in feminist theory and critical race theory (Pateman and Mills 2007). In his works on the world constitution, for example, Habermas explicitly builds on the 'reconstructive program of social contract theory' (Habermas 2006b, 131), imagining a 'second state of nature' including states and world citizens, instead of an anarchic state of nature among individuals (Habermas 2008b: 449). Thought experiments abstract from concrete experiences that people make because of their gender, race or social position. They explicate intuitions about just political orders -but whose intuitions? When Habermas reconstructs the meaning of a practice from the participants' perspective, he does not question who these participants are. For the context of constitutional democracy, for example, Habermas thinks of the perspective of members who, as a cross-generational community, are 'in the same boat' (Habermas 2001, 775;cf. Ferrara 2001). Although critics have pointed out that this assumption is problematic even in this case of a relatively well-established political practice (Honig 2009), it seems all the more controversial once we move to practices whose boundaries are in flux and whose participants are not clearly defined. Questioning the idea of the participants' perspective also raises the question of what counts as a 'shared practice' with 'constitutive presuppositions'. As Maeve Cooke points out: Ideas [ . . . ] are constitutive in the sense that they are formative of individual and collective identities, both constraining and enabling thought and behaviour. This is not to say that individuals and collectives must accept the validity of such formative ideas; it implies, rather, that those who reject them must take on the task of reorienting self-interpretations so that the normative assumptions in question no longer seem convincing. (Cooke 2005, 380) In the current context of membership politics, it is unclear to what extent we are dealing with a shared practice for which, from the perspective of the participants, the same presuppositions are constitutive. Rather, it seems that some participants have already taken on the task of 'reorienting their self-interpretations so that the normative assumptions in question no longer seem convincing'. Critics of current membership politics not only question specific problematic policies (such as denationalization) but fundamentally oppose the idea that states have the right to unilaterally determine their membership boundaries -or that there should be membership boundaries at all (Ataç, Rygiel and Stierl, 2016). Take the example of citizenship tests in Europe. In the early years of the 21st century, many European states have introduced citizenship tests as a requirement for naturalization. While many citizens, politicians and migrants support the idea that new citizens should show their interest in and commitment to a state's history, politics and culture, opponents find the idea of citizenship tests exclusive, outdated and even 'nationalist'. They do not only contest specific discriminatory questions but the overall idea that a national 'we' should set the terms and conditions for joining a political community (Ahlhaus 2020, chap. 1). In such a case, it seems difficult to speak of shared self-interpretations and a shared practice.
To sum up, the method of rational reconstruction faces a dilemma: Either we reconstruct a narrowly circumscribed social practice, which can lead to a problematic status quo bias, or we reconstruct a broader social practice including emerging and contested elements, undermining the method's key ideas about a shared practice with constitutive assumptions and a single participants' perspective. Against this background, I now turn to Benhabib's and Fraser's contributions to membership theory to develop an alternative reconstructive approach.

Benhabib and Fraser: Rational reconstructivists?
Benhabib's and Fraser's contributions to democratic theory and membership theory are well known. Their turn to questions of cosmopolitanism and global justice has been characterized as an 'emerging trend in critical theory' (Allen 2014, 144). Put briefly, Benhabib starts by analysing the changing context of multiculturalism and international politics. She describes the 'disaggregation' of citizenship and the importance of international human rights norms. Her aim is to present a new democratic theory of human rights and citizenship at the centre of which is the idea of 'democratic iterations'political processes in which popular sovereignty and universal norms are incrementally mediated (Benhabib 2002(Benhabib , 2004a(Benhabib , 2011a. Fraser, on the other hand, argues that the new political context created by globalization requires a post-Westphalian political theory that leaves traditional nation-state assumptions behind. To this end, she develops a multidimensional theory of democratic justice questioning the what, who and how of justice (Fraser 2005(Fraser , 2008(Fraser , 2009. Despite the prominence of Benhabib's and Fraser's political theories, there has not yet been a sustained discussion of their respective methodological approaches. Even though I do not want to claim that it is possible to fully separate a theory's method from its substance (Freyenhagen 2017, 458), I want to argue that based on Benhabib's and Fraser's contributions to democratic theory we can develop a new methodological approach. Usually the two authors are portrayed as taking their shared heritage as feminist critical theorists into different directions. Fraser has been characterized as a 'boldly eclectic theorist', while Benhabib is seen as a 'direct descendent of Habermas' (Canaday 2003, 51, 52). For Amy Allen, Benhabib belongs to those theorists who develop the 'left-Hegelian strand of Habermas's work', while Fraser presents a more contextualist approach similar to the idea of a 'problematizing genealogy' (Allen 2017, 250, 256). Despite these differences, Allen agrees that both authors follow a reconstructive approach (Allen 2013). Fraser explicitly states that 'feminists need both deconstruction and reconstruction, destabilizing of meaning and projection of utopian hope' (Fraser 1995a, 71, original emphasis). Indeed, in the rare moments when Benhabib and Fraser comment on the methodological underpinnings of their work, they self-identify as reconstructivists. Benhabib describes her approach as follows: [C]ritical social theory turns to those structures of autonomy and rationality which, in however distorted and imperfect fashion, continue in the lifeworld of our societies, while allying itself with the struggles of those for whom the hope of a better future provides the courage to live in the present. (Benhabib 1986, 15) She refers to rational traces or elements of realized autonomy that are already embedded in social practices. According to her, the reconstructive method aims 'to analyze and distill the rational principles of existing practices and institutions in such a fashion that we can then use these rational reconstructions as critical guidelines for measuring really existing democracies' (Benhabib 2002, 134; emphasis added). Fraser describes her methodological approach in a similar way: On the one hand, one should avoid an empiricist approach that simply adapts the theory to the existing realities, as that approach risks sacrificing its normative force. On the other hand, one should also avoid an externalist approach that invokes ideal theory to condemn social reality, as that approach risks forfeiting critical traction. The alternative, rather, is a critical-theoretical approach that seeks to locate normative standards and emancipatory political possibilities precisely within the historically unfolding constellation. (Fraser 2007, 8) Like Benhabib, Fraser embarks on a search for 'normative standards and emancipatory potentials' that have already been established in existing practices. She describes this procedure as 'immanent' (Fraser 2014, 131).
It is important to note that Benhabib and Fraser do not only take a reconstructivist approach but also follow Habermas in defending a communicative conception of rationality. Fraser takes 'dialogic, communicative rationality that goes with a post positivist understanding of social knowledge' as a starting point, arguing that participants have 'the reflexive capacity of communicative reason' (Fraser 2009, 42). Benhabib defends Habermas's idea that persons have the 'capacity for communicative reason' and adapts his discourse-ethical framework to make room for feminist conceptions of the self (Benhabib 2004a, 144;cf. Benhabib 1992). The normative principles that guide their analyses -participatory parity (Fraser) and respect for the capacity of communicative reason (Benhabib) -both build on Habermas's discourse-theoretical framework. Addressing democracy as a specific practice, Benhabib not only argues that 'institutions of liberal democracies embody the idealized content of a form of practical reason' but also that 'the task of a philosophical theory of democracy would consist in the clarification and articulation of the form of practical rationality represented by democratic rule' (Benhabib 1996, 68f.).
Yet neither Benhabib nor Fraser faithfully follow Habermas's method of rational reconstruction step by step. In particular, we don't find in their works hypothetical founding scenarios to distill normative principles. Benhabib describes her work as following Habermas but taking it in a feminist, Arendtian, and contestatory direction (Benhabib 2011b, 49). Similarly, Fraser rethinks Habermasian principles from a feminist and neo-pragmatist perspective (Fraser 1995b, 167). If we take a closer look at where they depart from Habermas, we find that they challenge the idea of a single participants' perspective and that they rethink the role of hypothetical rational discourses. My claim is that in their contributions to democratic theory we can identify elements of an alternative methodological approach that allows us to modify Habermas's method of rational reconstruction.

The method of plural reconstruction
Building on Benhabib's and Fraser's contributions to democratic theory, I now develop the method of plural reconstruction, which provides a way out of the described dilemma of rational reconstruction by systematically integrating emerging and contested practices. Using the example of membership politics, I distinguish four steps of the method of plural reconstruction: (1) Remapping the political landscape, (2) reconstructing plural presuppositions, (3) reconstructing emancipatory potentials, and (4) re-imagining institutions and the division of labour between theorists and citizens.

Remapping the political landscape
The main problem that arises when applying the method of rational reconstruction to emerging and contested practices is that it is unclear how to delineate the practices under investigation. Fraser and Benhabib do not claim to present an easy answer to this dilemma but instead they acknowledge the indeterminate nature of emerging political practices and their boundaries: '[V]ery often we do not know what type of practice is in question, for we do not share a common understanding of the disputed practice itself' (Benhabib 2002, 13, emphasis in original). They see themselves as engaged in a cartography of a new political context: a 'mapping of political space' (Fraser 2009, 4) is needed as the 'old maps' (Benhabib 2004a, 6) are increasingly inadequate.
Both authors embark on this project of redescribing the new political landscape's ambivalences, carefully reconsidering dominant empirical and normative assumptions. They hope to avoid the risk of '[r]ushing to assimilate a new phenomenon to a familiar past model', missing 'its genuine novelty' (Fraser 2009, 138). In their contributions to membership theory, they analyse political membership as a practice of both inclusion and exclusion, that is, as a precondition for democratic equality and as an instrument of inequality in the name of democracy. Fraser begins her description of membership relations with the inadequacy of the 'Westphalian frame', arguing that the idea that political communities are congruent with those subject to binding decisions is no longer a reality in a globalized age (Fraser 2005). Benhabib similarly shows that the congruence of ethnos and demos is an outdated ideal and that citizenship should instead be understood as a practice that is increasingly 'disaggregated' (Benhabib 2004a, 171f.). Both theorists see contestatory practices as part of the new political landscape, practices that 'emerge through the struggles of the oppressed, the exploited, and the humiliated' (Benhabib 2011a, 189), or more generally, 'emancipatory counter-tendencies and contrary and oppositional elements' (Fraser 2009, 139). Focusing on women's and immigrants' rights, they often identify these contestatory and emancipatory practices in social movements (Benhabib 2011a, 188;cf. Fraser 2009, 26).
If we describe this approach in general terms, we can say that the first step of a plural reconstruction consists in redescribing the political context under investigation with a particular focus on emerging and contestatory practices. By contestatory practices, I primarily mean different kinds of protests against dominant political norms. Emerging practices include new legal developments or institutions acting in a novel way, for example by gradually assuming new competences. The cartographic project includes a remapping of existing phenomena and a mapping of emerging developments, a description of former ideals and current demands, of dominant frames and alternative attempts to 'reframe' the political constellation. This is a crucial task, as only those elements of the political context that are identified, described and analysed can play a role in the further steps of plural reconstruction. Depending on our initial description of a practice, different presuppositions may appear constitutive for it. Indeed, just like maps are not mere neutral reflections but are created to serve a specific purpose, this redescription is a constructive and creative endeavour. It is up to the theorist to decide how to describe a practice, what aspects to focus on and what aspects to ignore. We have to take into account the 'creative reflexivity of reconstruction' (Fossen 2015(Fossen , 1079, as competing reconstructions are possible for every political practice. The practice of membership politics, for example, could be reconstructed as an 'instrument of national self-determination' or as a 'democratic practice of reciprocal boundary-making', meaning that we could either focus our reconstruction on the sovereign aspects of membership politics (e.g. decisions taken by domestic institutions, where only citizens have a say) or on the post-sovereign, boundary-blurring elements of contemporary membership politics (e.g. including protests and statements of non-citizens, court decisions strengthening the rights of non-members etc.). Of course, descriptions can be more or less plausible. A good map indicates where we stand and which routes lead to which destination. Similarly, a description of practices 'that flies over the heads of the social actors' is problematic (Fraser 2014, 144). But who are the addressees of plural reconstruction and whose perspectives count?

Reconstructing plural presuppositions
The key idea of Habermas's method of rational reconstruction is the analysis of idealized presuppositions from the participants' perspective. The normativity in his theory stems from analysing the performative meaning of social practices. When it comes to emerging and contested practices, however, where the point and purpose of a practice is controversial, or established fundamental principles are called into question, we cannot assume that there is a single participants' perspective. For example, once we include protests against the institutionalized forms of a practice in our description of it, we cannot speak of a shared perspective anymore, as the presuppositions of those who criticize a dominant practice might be in direct opposition to the presuppositions of those who do not see the need to change existing structures. While Habermas aims to reconstruct 'tacitly made counterfactual assumptions' (Habermas 2008, 151, my translation), a broader and more conflictual description of the relevant practice highlights the loss of the 'ordering force of shared presuppositions', which leads to an increase in 'metadisputes over constitutive assumptions concerning who counts and what is at stake' (Fraser 2008, 395). Against this background, it becomes clear that we need to refine the method such that conflicting presuppositions can be taken into account.
For example, Fraser reconstructs different perspectives of non-members in the case of membership politics. She paints a picture of the voices of the excluded, such as subaltern counter-publics that '[are] aimed at empowering subordinate voices in the battle for hearts and minds in wider publics' (Fraser 2014, 142). She problematizes 'the exclusion of the "global poor" from the universe of those who can press justice claims against "us"' (Fraser 2010, 367). Similarly, Benhabib's goal is to reconstruct both the perspectives of existing political communities and the views of 'outsiders'. In her membership theory, she diagnoses a 'constitutive dilemma at the heart of liberal democracies: between sovereign self-determination claims on the one hand and adherence to universal human rights principles on the other' (Benhabib 2004a, 2). Based on this analysis, she argues that 'practices of political membership are best illuminated through an internal reconstruction of these dual commitments' (Benhabib 2004a, 2;original emphasis). The reconstruction must therefore extend to the perspectives of both members and nonmembers.
But how do we identify the performative meaning of a practice from plural perspectives? If we adopt Habermas's strategy of searching for seemingly irrational behaviour with regard to the context of membership politics, we find an example in the participation in 'symbolic elections'. The European initiative 'We Vote' campaigns for voting rights for non-citizens in local, national and European elections by setting up symbolic polling stations and presenting the results to politicians. Thousands of non-citizens take part in these symbolic elections, hundreds of citizens set up polling stations and count votes (We Vote 2020). Which presuppositions do the participants of this practice of formally inconsequential voting share? For their actions to make sense, they must assume that the exclusion of non-citizens is problematic, that the current system is not immune to change, that symbolic elections might make this problem visible to others and that it might eventually lead to political change.
Other opponents of current membership practices instead act on the assumption that boundaries and borders can never be democratic and that more direct disobedience is necessary, such as crossing borders "illegally" or protesting in the streets (Ataç, Rygiel and Stierl, 2016). More abstractly, we might say that the participants of such contestatory practices challenge dominant answers to the question of who counts and how this should be decided (Fraser 2008, 398). The majority of citizens, by contrast, tacitly share the "old" idealizing presupposition that the current system of membership politics is legitimate, that is, that citizens have a right to set and defend their membership boundaries (Dumbrava 2014, 94).
To sum up, in the second step of the method of plural reconstruction, the normative meaning of a practice is reconstructed from plural perspectives. This means that different idealizing presuppositions are identified to understand a practice's performative meaning. These reconstructed presuppositions often contradict each other, 'interpretive conflicts' arise (Fraser 2014, 147) when formerly tacit assumptions are politically problematized and contested and when dominant narratives and self-understandings of democratic communities are challenged.

Reconstructing emancipatory potentials
While it is a crucial insight that participants of contested and newly emerging practices hold competing idealizing presuppositions, Benhabib and Fraser go beyond this diagnostic step. They analyse these practices to reconstruct their emancipatory potential. Which practices and emerging institutions already realize 'participatory parity' on a small scale? Where do we find practices that embody respect for the communicative freedom of all individuals? Fraser identifies different contestatory and innovative political practices such as 'direct action by the independent militants associated with Occupy, WikiLeaks, and the World Social Forum [whose] radical criticism has managed to pierce the veil of economistic and militaristic apologetics that dominates official public discourse in the present era' (Fraser 2014, 142). Of course, these might only be 'apparently emancipatory transnational movements [ . . . ] which may contain elite biases and do not always manage to live up to their own democratic aspirations' (Fraser 2009, 140), but they might also include the seeds for imagining emancipatory alternatives: In the World Social Forum, for example, some practitioners of transformative politics have fashioned a transnational public sphere where they can participate on a par with others in airing and resolving disputes about the frame. In this way, they are prefiguring the possibility of new institutions of postwestphalian democratic justice. (Fraser 2009, 26, emphasis added) Fraser uses the concept of 'prefiguration', which generally describes 'the deliberate experimental implementation of desired future social relations and practices in the hereand-now' (Raekstad and Gradin 2020, 10; emphasis removed). The idea is not that all future political institutions should work like the World Social Forum but that it is a practice in which the ideal of participatory parity is already partly realized. Prefigurative politics 'aspires to fuse the "against" -our rejection of ruling relations and institutionswith the "beyond" -our creation of new ways of being, relating, and doing' (Dixon 2014, 83). In other words, we find emancipatory elements in political practices, such as instances of free and fair discussions, examples of new forms of economic equality or alternative modes of listening to marginalized voices. This does not mean that we have to replicate such institutions at a higher political level but that we build on these emancipatory instances when thinking about alternatives to the institutional status quo.
Benhabib introduces the concept of 'democratic iterations' to identify processes in which human rights are contextualized and reinterpreted by political communities. Within such processes she looks for 'those moments of rupture and possible transformation when social actors reappropriate new norms, such as to enable new subjectivities to enter the public sphere and to alter the very meaning of claims making in the public sphere itself' (Benhabib 2011a, 136). Democratic iterations can have emancipatory potential when they show how 'political agency is possible beyond the member/nonmember divide' (Benhabib 2011a, 111), when they enable new subjectivities, when more individuals become authors of laws that apply to them. In claiming rights that they do not yet have, non-citizens act as political agents, as if they were citizens (Rygiel 2010, 12). Emancipatory practices also include the changing role of human rights and international law. Supranational courts defend the rights of non-members against isolationist tendencies, as well as legal norms that are 'translated' for local contexts and 'migrate' through different legal systems (Benhabib 2011a, 132). In such processes, human rights 'become permeable and fluid enough to absorb new semantic contexts, which, in turn, enable the augmentation of the meaning of rights' (Benhabib 2011a, 183).
If we describe this way of approaching post-sovereign membership politics in general methodological terms, we can say that in the third step of a plural reconstruction, emerging and contested practices are reconstructed to identify emancipatory potentials that indicate how existing structures of domination could be overcome. These might be moments in which we can catch a glimpse of radical change, instances showing that political practices could look different. We try to identify institutions and practices that realize autonomy in an unexpected way: when citizens (and non-citizens) create new forums to claim a say in political decisions, when innovative modes of building solidarity are tested, when new slogans repoliticize an issue or when unwritten rules are contested and changed. Of course, it is often difficult to determine what may count as 'emancipatory' and there is the risk of blindly affirming any protest that looks innovative or radical, but plural reconstructivists look for moments in which people reclaim their communicative power to change political structures.

Re-imagining institutions and the division of labour between theorists and citizens
But how do we get from plural presuppositions and emancipatory potentials to proposals for political change? As I have mentioned above, in the method of rational reconstruction, a thought experiment helps to 'purify' the rational fragments of existing practices. In his thought experiment in Between Facts and Norms, for example, Habermas asks: 'What rights must citizens mutually grant one another if they decide to constitute themselves as a voluntary association of legal consociates and legitimately to regulate their living together by means of positive law?' (Habermas 1996, 453). The 'system of rights' is his answer to this question -and at the same time his attempt at spelling out the practice's constitutive rules from the participants' perspective. Benhabib and Fraser, again, depart from Habermas's method. Benhabib worries that '[w]e seem to have already presupposed what democracy means and what democratic citizenship entails' and that Habermas's reconstruction 'takes the teeth out of the experience of social struggles in history' (Benhabib 2013, 50, n. 42).
For both, the idea of a thought experiment, especially in the form of a hypothetical founding scenario, must be revised in such a way that the perspectives of 'real others' are also taken into account. Benhabib includes the perspective of the participants as 'concrete others' and 'generalized others' (Benhabib 2011a, 69). The idea is to go beyond Habermas's 'retroactive wisdom' (Benhabib 1997, 726) by focusing less on the hypothetical genesis of certain institutions but on the different perspectives of actual individuals. Fraser criticizes 'monological theories' (Fraser 2005, 17), which do not make room for political decision-making by the participants, and proposes 'to rethink the classic division of labour between theorist and demos' (Fraser 2005, 18). The revised division of labour should primarily ensure that important aspects are not determined by theorists, but 'by the citizens themselves, through democratic deliberation' (Fraser 2005, 18).
For Fraser and Benhabib, plural and competing presuppositions do not have to stand in the way of normative guidelines for re-imagining the political constellation. They refuse to give in to the 'specter of incommensurability' (Fraser 2009, 3), that is, to the view that there can be no meaningful communication between those who hold competing worldviews (Benhabib 2002, 29). Instead, they propose new political institutions in which diverse positions and plural voices can enter into actual discourses. Fraser builds on the emancipatory potential she identifies in practices such as the World Social Forum but also criticizes that it lacks democratic legitimacy and efficacy. Thus, she develops the idea of a 'meta-democracy'. Decisions on inclusion and exclusion, on identifying the appropriate 'frame' for justice claims, should be taken in 'new global representative institutions' (Fraser 2008, 416). While Fraser does not elaborate on these institutions' 'shape, scope, scale, location, or character' (Zurn 2012, 169), she suggests that their main function is to decide whether claims against 'unjust' exclusions are indeed justified and to propose adequate changes. Benhabib's idea of 'democratic iterations' plays a similar role. It is a descriptive and a normative concept: describing processes in which human rights norms are contextualized and reappropriated by democratic communities, but also offering criteria for evaluating such processes in terms of fairness, openness and justice (Benhabib 2007, 449). Building on the cosmopolitan elements of existing practices, Benhabib develops the normative proposal that membership politics should take place in transnational discourses in which not only members but also non-members are represented (Benhabib 2011a, 151-2).
At this point, it is important to note that there are also crucial differences between Benhabib's and Fraser's approaches and aims. In the context of their contributions to democratic theory, however, these differences are often overstated when the idea of 'democratic iterations' is only interpreted as a description of discursive processes of incremental (multi)cultural change, while Fraser is seen as primarily proposing institutional innovations. In my view, we need to take seriously the institutional alternatives envisaged in relation to the concept of 'democratic iterations'. Benhabib's idea of a 'demotic community' (encompassing citizens and resident non-citizens) and her inclusion of transnational NGOs, social movements and international human rights courts as agents of 'jurisgenerative politics' indicate that she also aims at alternatives to the institutional status quo (Benhabib 2011a, 151-2). Both authors argue that it is not up to theorists to prescribe membership boundaries but the task of members and nonmembers of political communities to debate and decide these questions.
The fourth step of a plural reconstruction thus builds a bridge between plural perspectives and emancipatory potentials by proposing a new shared (democratic) practice. Through the pluralization of reconstructed perspectives, tensions between the different understandings become visible and their respective 'one-sidedness' is revealed. However, these tensions should not be resolved theoretically but negotiated politically. The idea is that in the absence of shared presuppositions, shared institutions are necessary to formulate new versions of a democratic self-understanding. The constructive contribution of the method of plural reconstruction is to outline institutional forums to democratize conflict in the context of emerging and contested practices. This negotiation and deliberation is modelled after those practices that are identified as having emancipatory potential. This step also helps to clarify the ambivalent relationship between critical theory and institutional design. On the one hand, many critical theorists agree that '[a]long with social theory, moral philosophy, and psychology, critical theory today needs to focus on [ . . . ] macro-questions of institutional design' (Benhabib 2004b, 299, original emphasis). On the other hand, they do not want to offer 'a blueprint for changing institutions and practices' (Benhabib 2002, 115). The last step of the method of plural reconstruction suggests that the goal is to outline ways of democratizing political practices. Instead of prescribing new institutions in detail, alternative institutional avenues are discussed that could help democratizing the status quo 'if and when the democratic will of the participants to do so exists' (Benhabib 2002, 115; on the notion of 'democratization' in the context of critical theory see also Bohman 2007, 36). Plural reconstructions do not present a fully fledged institutional model but they combine the plurality of perspectives with emancipatory elements already realized in existing practices to propose modes of democratization.

Addressing objections against the method of plural reconstruction
Let me now address two potential objections against my claim that the method of plural reconstruction is more suitable to analyse emerging and contested political practices than rational reconstruction. A first objection could be that the method of plural reconstruction, despite its attempt to escape it, inherits the 'single story'-problem of Habermas's method, that is, that it glosses over the empirical messiness of a political practice and as a result formulates overly idealized principles. Indeed, as plural reconstruction relies on communicative rationality, the idea of the participants' perspective, and idealized presuppositions, the building blocks of Habermas's method remain intact. While I agree that the method of plural reconstruction could be described as a variant of rather than an alternative to rational reconstruction, it is important to keep in mind the two main differences between the methods.
First, plural reconstruction leaves behind the idea of a single participants' perspective, meaning that we no longer have to assume that participants have the same selfunderstanding. As plural reconstruction does not seek to identify a single set of shared presuppositions, it can fully take into account the complex and contradictory nature of political practices. This means that objections as the ones raised by Honig and Olson mentioned earlier lose traction in the case of plural reconstruction. Furthermore, the method of plural reconstruction replaces the idea of a thought experiment with actual deliberation in democratic forums. Taking seriously the long-standing objections against hypothetical founding scenarios and the more recent critiques raised by Pateman and Mills as mentioned above, the method of plural reconstruction builds on the idea that a hypothetical founding act cannot accommodate the plurality of perspectives we find in political practices. Instead, it proposes a framework for actual deliberation among participants and indicates criteria for fair and inclusive processes. Extending the idea of what a 'method' in democratic theory should accomplish, plural reconstruction identifies a place for 'the people' in criticizing and overcoming problematic structures of domination. Making visible and building on a variety of participants' perspectives, the method is more suitable for emerging and contested practices than the method of rational reconstruction. At the same time, it could also allow for a more nuanced analysis of wellinstitutionalized practices such as constitutional democracy, highlighting the plurality of perspectives hidden in seemingly 'uncontroversial' practices.
While the first objection sees plural reconstruction faced with the same problems as the method of rational reconstruction, a second objection could be that plural reconstruction does away with the most important features of rational reconstruction and therefore forfeits the advantages of Habermas's method. The worry here is that while rational reconstruction provides a clear way of formulating normative recommendations, the results of plural reconstruction are too vague to offer practical guidance. Two things can be said in response to this. On the one hand, it is true that the method of plural reconstruction does not necessarily lead to clear normative guidelines as to how a specific practice (e.g. selling citizenship) should be evaluated and changed. Instead, the method reframes the question and shows that some normative puzzles ('Should citizenship be for sale?') transform into political problems once we consider the plurality of perspectives involved ('Who should decide on citizenship boundaries?'). In other words, the method of plural reconstruction implies that certain problems should no longer be considered issues to be solved by political theorists but conflicts to be addressed by citizens in political processes. On the other hand, this does not mean that the method of plural reconstruction stays clear of any normative recommendations. As compared to rational reconstruction, its main advantage is to expose the 'interpretive conflicts' that are at play in contemporary political practices and to indicate whose perspectives need to be included in fair, open and inclusive deliberations and decisions on these normative puzzles. Plural reconstruction is a method that reframes normative problems as political challenges for political communities: it aims at proceduralizing and, ultimately, democratizing problems by showing the myriad of diverging perspectives that need to be taken into account.
Finally, the question is not whether we should defend the method of rational reconstruction or the method of plural reconstruction but rather when and for which contexts and research questions these methods should be used. I have argued that there are a number of cases the method of rational reconstruction is unable to address adequately. Wherever institutional constellations are in flux and contested, the core ideas of rational reconstruction are challenged. Other cases, besides membership politics, could be climate politics or internet governance. In such cases, we need to pay attention to both established institutions and emerging and contestatory developments, which is why the more exploratory and experimental nature of plural reconstruction seems adequate.

Conclusion
I have argued that the method of plural reconstruction provides a way to rethink the method of rational reconstruction for emerging and contested political practices. The method of rational reconstruction as developed by Habermas aims to criticize existing practices by reconstructing their normative meaning embodied in the participants' idealizing presuppositions. Where we are confronted with emerging and contested practices, this method faces challenges that question its adequacy. I argued that in Benhabib's and Fraser's contributions to democratic theory, we can identify elements of an alternative reconstructive method -which I call 'plural reconstruction'. The method of plural reconstruction proceeds in four steps: Remapping the political landscape, reconstructing plural presuppositions, reconstructing emancipatory potentials and re-imaging institutions and the division of labour between theorists and citizens. The method of plural reconstruction is particularly suitable for political contexts whose fundamental parameters -What is the problem? Who counts? How can decisions be legitimate? -are unclear.

Notes
This article has benefitted from many comments and suggestions. I would like to thank Amy Allen, Hubertus Buchstein, Robin Celikates, Jeanette Ehrmann, Rainer Forst, Ervin Kondakciu, Peter Niesen, Martin Nonhoff, Markus Patberg, Thomas Saretzki, Rainer Schmalz-Bruns, Karsten Schubert, Frieder Vogelmann, and the participants of workshops in Bremen, Freiburg and Prag in 2019. I am especially grateful for the anonymous reviewer's helpful comments.
1. This entails that I do not engage with the question of whether Habermas uses the method of rational reconstruction in a similar way when applying it to other kinds of social practices, that is, basic social practices such as communicative action (cf. Gaus 2019, 369;Patberg 2014). 2. I use the term 'rational' here in Habermas's communicative sense (see Strecker 2019). 3. I thank the reviewer for pressing me on this point.