Ricardo Fabrino Mendonça

Federal University of Minas Gerais (UFMG), Brazil

Replicated from Brazilian Political Science Review (Online), Rio de Janeiro, v.2, n.2, July/Dec. 2008.

ABSTRACT

This paper discusses the issue of political representation, by arguing the necessity of re-envisioning it so as to consider non-electoral forms of representation. It claims that civil society associations can be conceived of as representatives of a series of discourses, voices, opinions, perspectives and ideas. Whilst this type of representation lacks formal mechanisms of authorization and accountability, its legitimacy may emerge from the effects of such associations and from their porosity to several interactional loci. The paper suggests that associations that are open to several discursive spheres are more prone to foster a discursive accountability, built within a broad process in which discourses clash in several communicative contexts. The idea of a deliberative system helps to understand the interconnections among these interactional loci, as it points to the possibility of a dynamic between partiality and generality, which is at the heart of political representation.

Introduction

Philosophers and political scientists have been conceptualizing the idea of political representation and its role in the constitution of legitimate governments for centuries now. Adopting very different approaches, they have theorized about the ways through which collectively valid decisions should be taken. From Hobbes to James Mill or Madison, and including Rousseau, Burke and Condorcet, several canonical thinkers have reflected upon the adequacy of representation, and the forms through which such practice should (or should not) be implemented. Scholars concerned with the viability of democracy in large and complex societies are particularly preoccupied with the issue, claiming that representation is the only feasible way to exercise popular sovereignty in contemporary polities. Some conceive of representation not as a defective substitute for direct democracy, but rather as democracy in action, defending its advantages even in small communities.

In a very broad sense, the concept of representation denotes a form of political action in which a person or group acts in the place of another or others with a certain kind of authorization to do so. Representation is, by definition, a relation between represented and representative(s), which can take a wide range of forms (Dryzek and Niemeyer 2007; Castiglione and Warren 2005; Urbinati 2006; Rehfeld 2006; Avritzer 2007). As stated by Pitkin, the idea of representation itself has changed a lot throughout history, in parallel with the changes in institutions by which representative practices have been brought up to date (Pitkin 2006, 21).

Nowadays, there is a certain consensus around the idea that a representative should not defend only the interests of the faction that directly supports him or her. It is necessary that s/he aims at the best for the whole polity. This idea has raised a series of new questions, especially in times in which it seems more difficult to demarcate 'political communities'. On the level of macro-relations, the expansion of transnational interactions evince that the consequences of States' decisions (and their members' actions) have impacts that go beyond territorial borders (Giddens 1990; Dryzek and Niemeyer 2007; Castiglione and Warren 2005; Rehfeld 2006; Runciman 2007; Avritzer 2007). At the level of internal relations, territorial unity is constantly questioned by a plurality of cultural and social cleavages. There are, also, doubts about the basic units to be represented, with minorities demanding the creation of alternative mechanisms so that they may make themselves heard (Young 2000).

Hence, the necessity of thinking of non-electoral possibilities for the constitution of representation becomes evident, as the whole idea of representation loses its territorial basis (Urbinati 2005a). Even if we have become used to thinking of elections as the manner to institutionalize relations of representation in contemporary democracies, such relations may be redesigned and made more complex if other legitimizing procedures and accountability mechanisms are adopted (Dryzek and Niemeyer 2007). This does not mean electoral representation is dreadful and unnecessary. All I am defending is that representation and democratic elections are not indissoluble. Like Michael Saward, I believe that

the idea that electoral institutions themselves, while indispensable to contemporary democracy, by their very structure leave open the possibility for non-elective representative claims that can call on criteria of democratic legitimacy which in some ways echo but in important other ways are distinct from electoral criteria (Saward 2009, 2-3).

However, to think of these associations as political representatives also raises a set of thorny questions. What transforms them into representatives if there are no formal mechanisms of authorization? To whom should they be accountable? How can such accountability be instituted? What grants the legitimacy and quality of this form of representation? In short, could representatives be legitimate in the absence of clear mechanisms of authorization and accountability?

This paper aims at reflecting on these questions, regarding the possibility of conceiving of actors from civil society as political representatives. In order to do so, I will start by briefly discussing the idea of representation and some of the contemporary proposals to reframe it. The fertility of notions that broaden the focus from individuals will be defended. I will, then, discuss the potential of civic associations to act as representatives of discourses and perspectives. I argue that the formats of these associations, as well as their pragmatic effects, are at the core of their evaluation. This is a first condition for considering associations as democratic representatives.

Lastly, I will discuss the idea of deliberation in a diversity of interactional loci. ¹ Understanding public deliberation as a macro-process that happens in several spheres of communication, I defend that the porosity among these spheres is at the heart of broader types of representation. A variety of interactive contexts is essential if an association is to foster a dynamic between partiality and generality that is vital to political representation. I argue that the idea of interactional loci (i.e. spheres of communicative interaction) is of central relevance to the constitution of associations that dynamically renew their representativeness. Interactions in a variety of discursive arenas help to make this collectivity (and its leaders) discursively accountable, and thus more legitimate and more capable of exerting qualified representation. The openness of an association to several internal and external interactional loci is thus the second condition suggested for considering it as a democratic political representative.

Representation as a Dynamic Concept: Expanding the Focus Beyond both Elections and Individuals

Representation has shown itself to be a dynamic concept throughout history. Its roots are in the Latin notion of repraesentare, which literally means make present something that is actually absent (Runciman 2007). As discussed by Pitkin, the concept was initially reserved for inanimate objects (Pitkin 2006). It did not mean acting for, or on behalf of, others. It was only in the Middle Ages that the word started to be employed in reference to human beings. But that was just the beginning of the elaboration of the concept of political representation. Afterwards, the idea of agency had to be connected to representation, and several debates have sought to define what this type of agency actually meant. As pointed out by Pitkin (2006), the dichotomy delegate X trustee has been a major issue focused both by political philosophy and practice. She argues that representation can neither be seen as pure authorization (as Hobbes would defend), nor as simple delegation.

Such an idea is widely accepted nowadays. Nadia Urbinati, for instance, starts by criticizing both the notions of imperative mandate and complete autonomy, and by defining political representation as a relationship in which both representatives and represented must have their autonomy safeguarded (Urbinati 2005a, 2005b, 2006). Accepting this proposition, however, does not solve all the problems posed by contemporary polities. And Urbinati's work goes on to raise other key issues that are central to the definition of what representation can currently mean (Urbinati 2006). In so doing, she suggests that this political practice cannot be conceived of in purely electoral terms.

Urbinati defines political representation as "a circular process (susceptible to friction) between state institutions and social political practices" (Urbinati 2005b, 1). Representation connects institutions and society, in a cyclical process in which both representatives and represented are free to act, although being required to give reciprocal justifications concerning the interests, opinions and ideas they defend. To represent is, therefore, to be in a "relation of sympathetic similarity or communication with those in the place of whom the representatives act in the legislature" (Urbinati 2005a, 211). Representation, according to this approach, is a relationship that may be embodied in several ways. Besides elections, there are many possibilities to engender links between inputs and outputs in a political system. There are different ways to foster circularity between state and society.

From this perspective, popular sovereignty does not emerge only through electoral authorization. The exercises of prospective accountability, political surveillance and of influence through informal venues are also fundamental. In order to advance such a perspective, Urbinati resorts to the Kantian notion of judgment. In her framework, citizens should constantly evaluate the behaviour of representatives, thinking as if they were in their place. In such dynamics, constituents are oriented by principles, opinions, values and ideologies, with which representatives should establish dialogue. Sovereignty would emerge from public processes of opinion formation. It depends on "the activation of a communicative current between civil and political society" (Urbinati 2005b, 12-13).

There is a growing literature devoted to this possibility of turning democratic representation into something more complex and diverse (Avritzer 2007; Abers and Keck 2006; Bang and Dyrberg 2000; Castiglione and Warren 2005; Dryzek and Niemeyer 2007; Eckersley 2000; Gurza Lavalle et al. 2006; Keck 2004; Mansbridge 2003; Meier 2000; Parkinson 2003; Saward 2009; Squires 2000; Young 2000). Some scholars have advocated the necessity of representative procedures that are not centred on the representation of individuals. Their proposals involve a shift in the basic political unit to be represented, which would depend on, and concomitantly imply, other political mechanisms besides elections. Urbinati herself points out that the selection of representatives is not simply a choice of specific persons to represent particular individuals. It involves the expression of support for ideas, values, beliefs and publicly manifested positions.

Jane Mansbridge also defends the importance of forms of representation that are not characterized by electoral bonds (Mansbridge 2003). When proposing a political model that combines different forms of representation, she argues that surrogate representation also has its place. Usually exercised through informal ways, surrogate representation is based on the advancement of opinions, interests and perspectives. Although Mansbridge focuses specifically on the exercise of surrogate representation by elected representatives, her idea could be extended to representatives who are not authorized through elections (Saward 2009, 2).

When analyzing some challenges faced by contemporary democracies, Castiglione and Warren argue along similar lines. They claim that the object of representation is not individuals as such. Representatives act in defence of certain wants, ideas, understandings, interests and values. And they participate in the process in which those are constructed, once they frame and vocalize specific social perspectives. According to them,

it is precisely this detachment of collective entities from persons that enables representatives to represent positions in public discourse and argument, in this way serving as both conduit and structure of public spheres. Without this detachment from specific persons and interests, politics would fail to have a discursive locus, and would be reduced to the aggregation and bargaining of interests and identities (Castiglione and Warren 2005, 16).

Clearly, their proposal is to shift the focus of representation away from individuals, and this requires non-electoral means. It also requires a type of symbolic or discursive action in the public sphere. This is so especially in societies characterized by the decentralization of certain powers and the globalization of others. "On the one hand, politics is increasingly spilling out of formal, electoral politics into non-electoral and informal domains […] On the other hand, modes of influence are expanding" (Castiglione and Warren 2005, 17). There are, hence, several forms of non-electoral representation, which goes from interest groups to social movements and includes a wide range of associations and ascriptive groups. These representatives may act in spheres of participatory decision-making (Gastil and Levine 2005; Wampler and Avritzer 2004), in the exercise of influence over elected representatives (Habermas 1996; Parkinson 2003), and in the formation of public opinion by the vocalization of certain discourses (Dryzek 2000a). In this way, such representatives raise actual possibilities for citizen participation and for the maintenance of the circularity between state and society.

Another interesting approach is the one adopted by Iris Young, who argues that representation should be understood "as a differentiated relationship among political actors engaged in a process extending over space and time" (Young 2000, 123). When arguing in favour of democratic representation of minorities and marginalized sub-groups, Young distinguishes perspectives from both interests and opinions.² Perspectives, which are a product of social structures, do not have a specific content and would thus be plural. The representation of them is neither focused on individuals nor on a group common essence.

According to Young, democracy may be deepened by the pluralization of formats and spheres of representation because "systems of political representation cannot make individuals present in their individuality, but rather should represent aspects of a person's life experience, identity, beliefs, or activity where she or he has affinity with others" (Young 2000, 133). A complex web of representative mechanisms has a greater chance of representing more aspects of individuals. For this reason, Young values both formal and informal representatives, and she indicates that the representation of social perspectives must occur in several contexts besides parliaments, including civil society associations.

A similar point is made by Michael Saward (2009), who acknowledges that representation is always partial and incomplete. In order to deal with the constitutive plurality of identities and constituencies, he recommends one should think of representation as a claim, instead of a possession. Such claim has to be redeemed by audiences, and elections are just one procedure to conduce this process of redemption. Saward (2009, 7-8) argues that "despite its undoubted strengths elective representation contains structural weaknesses that some forms of non-elective representation may be able to exploit, by offering different sorts of representative claims which may resonate well with specific audiences". He suggests hence that the deepening of democracy may require different types of claims of representation. Although he does not specify the objects of these different sorts of representation, one can assume they should not be restricted to individuals.

Keck's approach is, nevertheless, still imprecise when attempting to define the idea of discursive representation. This is so because she is somewhat vague in her definition of discourse. A more developed account is the one advanced by Dryzek and Niemeyer (2007), who connect the idea of discursive representation to deliberative democracy. For them, a system solely based on the representation of individuals has a homogenizing feature, as it is unable to capture the nuances of socially existent discourses. For this reason, they suggest additional modes of representation, which would be more appropriate to deal with the constitutive multiplicity of selves. They argue representation should be centred on discourses, as they are essential for the constitution of selves and social life. From this perspective, individuals are not the most basic unit of the political world. At the kernel of politics is a constellation of multiple and contesting discourses. To Dryzek and Niemeyer (2007), civil society associations are very important in vocalizing several discourses. Their action is at the heart of a lively public sphere. But the scholars also propose an institutional arrangement for the consolidation of discursive representation. To them, the formation of discursive chambers would promote the connection between processes of opinion formation and decision-making.

The approaches discussed throughout this section point to the necessity of conceiving of representatives in ways that go beyond the idea of formally elected politicians. They suggest that civil society associations may be thought of as political representatives, which are organized around certain collectivities and that foster specific interests, perspectives or discourses. In so doing, such approaches also state the need to understand the object of representation beyond the classical definitions centred on individuals. While opening a fruitful discussion, these suggestions pose new questions. Civil society representation is not marked by formal procedures of authorization and accountability. In addition to that, such representation does not often have decision-making capacities.

I argue, nonetheless, that even if idiosyncratic, civil society representation plays a significant role in the collective construction of decisions made by a society. Representation is a political practice that may acquire existence in institutions of different formats. Doubtless, civic associations are one of them, as they can foster political inclusion of individuals (by advocating their discourses, perspectives, opinions or identitarian features) in processes of opinion formation and decision-making. These associations may enhance the circularity between state and society. As Young puts it, "Strong, autonomous, and plural activities of civic associations offer individuals and social groups maximum opportunity in their own diversity to be represented in public life" (Young 2000, 153).

It is important, however, to refine this argument and discuss in depth the whole idea of civil associations as political representatives. This is what I will do in the following sections of this paper. I argue that considering an association as a legitimate political representative requires analyzing its features, its pragmatic effects and its relationships with other social actors. I do not agree, therefore, with neo-Tocquevillean perspectives which simply take for granted that the redemption of democracy can emerge by the simple existence of a strong and organized civil society. There are conditions to evaluate if an association exerts (or not) democratic political representation.

My central argument, as should be clear in the final section, is that legitimacy and accountability are also central features of this type of representation. They do not nevertheless emerge from formal procedures of authorization. Legitimate representation from civil society can be enhanced by accountability processes constituted through communicative interchanges that take place in several interactional loci (i.e. discursive contexts). The absence of formal mechanisms of authorization does not imply the total absence of mechanisms of legitimation. There are, as a matter of fact, informal procedures to generate accountability and legitimacy. Such procedures, I contend, can be fostered by a deliberative conception of democracy that emphasizes the importance of communicative exchanges in different contexts.

Representation in Civil Society: Associations as a Possibility of Transit between the General and the Partial

Thinking of civil society associations through the lenses of political representation requires caution, so as to avoid an overestimation of their potentials. Simply propagating the qualities and wonders that emanate from civil society does not take one very far. Civil society must be seen as complex and heterogeneous. And such complexity cannot be put in uncomplicated terms as if there were a good and a bad civil society which could be simply defined by the goals of certain associations (Armony 2004).

An interesting route, in this sense, is the one opened by Mark Warren, who defines the practice of association as a "form of social organization that thrives on talk, normative agreement, cultural similarity and shared ambitions – that is, forms of communication that are rooted in speech, gesture, self-presentation, and related forms of social interaction (Warren 2001, 39). In this approach, associations are not so much an empirical reality, but more a medium of social organization. Anchored in Parsonian sociology, this scheme claims that the principle of association is based on communication and normative influence, not on money and power. As such, it can be found in organizations and institutions also permeated by other steering media.

Warren subdivides potential democratic effects of associations into three main categories (Warren 2001, 61): 1) they may contribute to the formation and strengthening of citizens' capacities; 2) they may be important in creating an infrastructure for public spheres; and 3) they may contribute to generate institutional conditions for the transformation of autonomous judgments into collective decisions. These three types of effects are central to democracy, because they help to strengthen the exercise of autonomy, both in its individual and political dimensions.³

One might ask, however, what these effects have to do with the exercise of democratic political representation by civic associations. If one is interested in proposals that expand the concept of representation, so as to allow the deepening of democracy, and if one claims that associations may play a significant role in such expansion, this can only be the case if associations themselves have democratic effects. These effects cannot simply be attributed to associations; neither can they be deduced from the stated goals of associations. They emerge pragmatically as the result of several factors, including the context in which such associations are immersed (Armony 2004). It is only when strengthening democracy – by fostering individual and political autonomy – that associations may be considered as democratic political representatives. This is so because only then may associations be able to nurture the dynamic circularity between state and society.

Take, for instance, the second type of democratic effect mentioned by Warren, i.e. creating the structures for public spheres. Such structures are essential for the exercise of autonomy. The communicative interactions established by an association with other actors, as well as the interlocutive flows that constitute the association itself, are crucial to representative democracy. These internal and external discursive exchanges are fundamental to enhance the accountability of actors from the formal political system. Also, and perhaps most importantly, they may enable the publicization of perspectives and arguments which might not have been heard otherwise. Associations may capture, organize and amplify the public frames of specific debates (Habermas 1996). In this dynamic, they publicly advocate perspectives and discourses of specific groups, representing them before broader publics. Such a process is not harmonious and easygoing, but full of tensions (Parkinson 2003).

My point here is that such communicative role means that associations advocate specific causes, promoting the public existence of a great number of traditionally excluded discourses. As argued by Saward (2009, 12) "a representative claim may be based on the fact that an important perspective within a debate is not being heard or even voiced". It is worth mentioning that this public advocacy cannot be simply presented as a struggle for particular interests. It also involves the generalization of perspectives and arguments. If associations are to act as political representatives, they may be partial-yet-communal actors, as Urbinati refers to representatives. That is, they must foster the connection between particular positions and general principles, which does not mean they must (or should) be impartial. "Advocacy is not blind partisanship; advocates are expected to be passionate and intelligent defenders" (Urbinati 2006, 46).

The whole idea of associations as representatives is not restricted, however, to the vocalization of discourses in the public sphere, which would be a very informal conception of representation. It is relevant to recall that most of the proposals we have discussed in the previous section are mostly concerned with processes of decision-making. This leads us to the third category in Warren's typology, namely, the institutional effects of associations. Civic associations can both exert pressure upon formal arenas of decision-making and get involved in more participatory institutional designs.

In the first case, one must take into consideration that associations may put pressure on elected politicians, acting on behalf of discourses, interests, opinions and perspectives. As discussed by Habermas, the centre of the political system cannot be thought of as an autonomous and autopoietic sphere. It cannot produce legitimate decisions if isolated from other systems and society as a whole (Habermas 1996).⁴ The pressure upon the actors of this system can happen in a variety of ways, ranging from rhetoric to cultural change (Dryzek 2000a). These activities are fundamental for the maintenance of communicative flows linking state and society. Extra-parliamentary forms of representation are permanently in practice.

In the second case, i.e. participatory forums, associations may have a direct voice in processes of decision-making. This happens in arenas in which members of the government establish dialogues with other social actors in order to produce more complex and participatory decisions. Some contemporary practices that could be mentioned are participatory budgeting, deliberative councils, and thematic committees that sometimes have legislative authority.⁵ In these forums, associations act in defence of certain policies, alleging they benefit both the ones they represent and society as a whole. There, civil society actors may present discourses and perspectives in ways that promote the connections between specific and general. They may foster communicative processes that lead to the consideration of the positions of all those potentially affected by a specific decision.

In this section, I have argued that a first condition for conceiving of civil society associations as democratic political representatives is their pragmatic effects. If such associations are to be understood as democratic representatives, they must enhance democracy and they may do so by providing communicative structures for the public sphere and by enabling citizens to have a say in decisions that affect their lives. I also suggested that these democratic effects cannot simply be taken for granted, nor can they be deduced from the goals of associations. One must research, empirically, the manifestation of these effects in specific contexts, as argued by Armony (2004). It is only through the observation of these associations and their effects that one may evaluate whether they promote the public representation of a greater number of discourses or suffocate the pluralization of the public sphere.⁶ This empirical observation must take into consideration the constitution of these representatives, since they are collective, heterogeneous and multifaceted actors. This is the issue I will be dealing with in the final section of the present paper. My claim is that an association's openness to several interactional loci is a significant way to promote accountability and legitimate representation.

As already mentioned, the main problem of considering civil society associations as political representatives refers to the legitimacy of these organizations.⁷ Usually, there are no formal mechanisms of authorization, accountability and punishment to ensure that a representative relationship will be an actual relationship. Such absence of representative bonds poses questions as to the adequacy of these actors having an active role in processes of decision-making, for instance (Parkinson 2003).

However, as pointed out by Castiglione and Warren (2005, 20) "What counts as authorization and accountability will, of course, depend upon the kind of representative". If associations are not elected by a whole political community, this does not mean they are not submitted to accountability. There are other kinds of practices that permeate an association and that may point to (or deny) its legitimacy. Such practices guarantee a strong and tight connection between the represented (their discourses, ideas, perspectives, opinions) and the representatives (in this case, associations). Castiglione and Warren argue that, in these cases, authorization may emerge by the capacity of a group to attract members, by a convergence with characteristics of the represented, by public visibility or by success in building public justifications.

I believe informal mechanisms of legitimacy and accountability are directly connected to the organizing structure of an association and to its strategies. In this sense, I agree with the argument that internal inclusive communication between subjects and those who claim to act on their behalf is central to the constitution of representation (Warren 2001, 166; Parkinson 2003, 84). It is of fundamental relevance that an association structures itself in a way that fosters a series of interactional loci, so as to increase communicative flows. Associations must guarantee the existence of several spheres of interlocution, which enable a permanent encounter and confrontation of discourses and ideas. This is the only way, an association may show its plurality and its adjusting capacity, which are essential attributes for the exercise of effective representation. A representative must be in permanent metamorphosis so as to reconstruct its bonds with the represented. Through internal communication, an association's claims of representation may be endorsed or questioned by those that are at its basis (Runciman 2007).

It must be clear, though, that just internal communication is not enough. Note, therefore, that I am not arguing that a horizontal internal structure entitles an association to act as a political representative. Although the existing literature stresses the importance of internal communication, it is also important to emphasize the relevance of communication with social actors that are not part of the association. The above-mentioned adjusting capacity is not only in reference to the aspirations of those an association claims to represent. There must be an adjustment to society as a whole, because representation is not mere delegation. If it is to act as a democratic representative, an association must insert itself in a web of discourses, building its own utterances, and testing their adjustment in a variety of internal and external interactional loci. In this way, it can fine-tune its relationship both to the perspectives, interests, opinions and discourses it claims to represent and to the broader constellation of discourses available in the public sphere.

The point I make here is that these several spheres of interlocution allow not only the construction of an association's discourse, but also the encounter of this discourse with those from other social actors. This enables the dynamics between the partial and the general, which is at the core of representation. "The political process of representation filters and sorts out the irreducible partiality of social or cultural identities by making them issues of political alliances and programs" (Urbinati 2006, 37). Representation is therefore an important component of a type of politics characterized by the confrontation of discourses.

By enabling the vocalization of certain world views and by fomenting the exercise of reciprocal evaluations, representation may instigate a fruitful dynamic between partiality and generality: a representative speaks from a specific perspective, but s/he may do so in the name of the entire collectivity. Thus, democratic political representation must promote plurality and divergence in the political field, without leading to sectarianism. This is so because it demands a permanent movement between the general and the specific. Representation depends on the translation of specific points of view into a general language, a job with which some associations are daily engaged (Alexander 1996).

Hence, my claim is that the capacity for certain associations to act as political representatives mostly emerges in the process of construction of the discourses they publicly defend. These discourses will only be legitimate if they remain open to dialogue both with those they claim to represent and with society in a broader sense. Following Avritzer (2007), I thus acknowledge that civil society representation should not be thought of in terms of authorization, as it is their legitimacy that justifies their importance as political representatives. It is through the public exchange of arguments in different discursive arenas that an association may build its legitimacy. It is also through such back-and-forth of non-coerced communication that an association may justify its actions and utterances. The central aspect for representation in civil society is the maintenance of an ongoing discursive process in a diversity of spheres.

In this sense, the idea of public deliberation⁸ in different arenas seems to be a central element in constructing political representation and in fomenting accountability. Herreros has already proposed the capacity to promote deliberation as a criterion to differentiate associations, but he does so because he thinks this can produce virtuous citizens (Herreros 2000). Instead, I propose that the openness of an association to deliberation in several arenas can stimulate a movement between partiality and generality which is essential to representation.

The idea of a deliberative system, as advanced by several authors (Mansbridge 1999; Conover and Searing 2005; Hendriks 2006; Parkinson 2003; Marques et al., 2007),⁹ is at the kernel of this proposal. Such a system is formed by the crossing over of informal spheres of conversation and formal arenas of decision-making. This model "recognizes that public deliberation is not an activity restricted to either micro or macro venues, but something that takes place in all sorts of institutions, arenas and spaces in social life" (Hendriks 2006, 497). If a deliberative system is formed by several loci where people interact with each other, it is central that these loci are connected, so as to promote a social circulation of discourses. How tight this articulation should be and how it can be promoted or endangered is a matter of empirical research, but there must be an articulation of different spheres, if deliberation is to be effective in fostering the flow of discourses.

Recently, Habermas (2006) has also come to emphasize that a deliberative process spread over society promotes the generalization of arguments. He thus sustains the relevance of a clash of discourses produced in different social arenas.

Political communication, circulating from the bottom up and the top down throughout a multilevel system (from everyday talk in civil society, through public discourse and mediated communication in weak publics, to the institutionalized discourses at the center of the political system), takes on quite different forms in different arenas. (Habermas 2006, 415).

Summing up, the idea advocated is that these interactional loci (or communicative contexts) that constitute the process of public deliberation permeate an association. Such interactional loci may range from informal conversations in a bus stop to formal public assemblies. Associations that remain more open to such crossing, building their foundations on internal and external argumentative exchanges, are more prone to play an actual role as political representatives. If representation always raises the question of who should be accountable to whom, as submitted by Gutmann and Thompson (1996, 128), the proposal advanced here tries to decentralize the process of accountability. It does so by arguing that such accountability occurs in the encounter of multiple discourses processed in several contexts. The exchange of justifications constrained by publicity appears as the quintessential form of accountability. From this perspective, associations

need not have direct principal-agent link with the relatively inactive citizenry to have a legitimate role in a deliberative democracy: they are the essential facilitators who do have time, resources and expertise to facilitate communication throughout the macro deliberative system (Parkinson 2003, 117).

Thus, my approach assumes that accountability is not restricted to isolated actions, such as voting in regular elections. Furthermore, it suggests that processes of accountability may not be centred on individuals, as they emerge in the confrontation of discourses in the public sphere. If deliberative accountability goes beyond elections and requires that representatives justify their actions in moral terms (Gutmann and Thompson 1996, 129), what I am advocating goes one step further, since it disembodies these justifications by focusing on communicative exchanges instead of on subjects who produce utterances.¹¹ It is, therefore, a discursive accountability, which emerges in the give-and-take of arguments.¹² Associations may give good reasons for their perspectives, and the publics with which they interact (both internal publics and external publics) may evaluate the adequacy of these reasons in a discursive process.

Civil society associations, as democratic political representatives, need to constantly justify their actions and utterances in several discursive arenas. In this way, they foster a process of back-and-forth of communication which advances the connections between partiality and generality and the circularity between state and society.¹³ In this discursive justificatory process, associations may (or may not) constitute themselves as legitimate representatives for the propagation of specific discourses and the defence of certain causes. Representation becomes thus a discursive process in which claims of representation are always subjected to redemption or denial (Saward 2009). Only associations that are able to sustain a link with the discourses and opinions of those they claim to represent, and also publicize them in socially acceptable terms, can be taken as genuine and legitimate democratic political representatives.

In this paper, I have discussed the issue of political representation, defending the idea that the actions of civil society associations can be thought of as a form of representation of discourses, voices, opinions, perspectives and ideas. I have defended the necessity of distinguishing between different types of associations, claiming that the ones that actually have democratic effects and whose structures are open to several crossings in a diversity of interactional loci are more apt to act as democratic political representatives. I have gone on to advocate that the notion of deliberation and, more precisely, of a deliberative system, helps to conceive the cyclical process of accountability through which associations may acquire or lose legitimacy to act as representatives.

I am well aware of the perspective that insists that deliberation cannot be thought of under a representative regime of democracy, but believe this to be mistaken. Deliberation is not opposed to representation. The former may even fuel the latter, since it can extend accountability mechanisms beyond the formality of voting. From my perspective, associations that foster deliberative processes in several arenas foment a political context propitious to the spread of participation. They may, therefore, lead to a more inclusive representative democracy, as the process of public justification helps to supplant non-reflexive forms of power and promotes the public evaluation of discourses. Once opened to public scrutiny and justifying themselves in several argumentative arenas, civil society associations may have the legitimacy to act in defence of certain discourses.

I do not claim, however, that associations are the whole basis of representative democracy, in some kind of simplified version of associative democracy. All I am saying is that they may play important roles as representatives, thus leading to further democratization of democracies. Associations are an important component of a system of multi-layered types of representation, as argued by Parkinson and Urbinati. All that is necessary is finding a balance among these different representative agencies, through communicative flows that traverse and connect them.

Notes

1 I use the term interactional locus to refer to any sort of context where people interact with each other through language. I do not call these contexts deliberative arenas because most of what goes on in communicative exchanges is not deliberative. I claim, however, that fragments of communicative exchanges in several spheres may constitute amplified deliberative processes. Interactional loci can be formal or informal. They can happen in face-to-face meetings or through any sort of mediated communication. Different interactive contexts allow the emergence of different sorts of interaction, and these different types of communicative exchanges can bring different contributions to deliberation. Deliberation can be enriched if it is thought of as something that intersects everyday conversations, meetings in associations, media discourses, parliamentary debates and technical committees, for instance.

2 Young (2000, 134-135), defines "interest as what affects or is important to the life prospects of individuals, or the goals of organizations". Opinions are "principles, values, and priorities held by a person as these bear on and condition his or her judgement about what policies should be pursued and ends sought". Perspectives, on the other hand, refer to a certain way of looking at the world and comprehending it. They are shared by individuals who have similar experiences, biographical histories and frameworks, which are generated by the structure of social locations.

3 Warren (2001, 63-65) explains that individual autonomy does not imply isolation and individualism. It has to do with the inter-subjectively built capacity of participating in reasoning processes and of arriving at judgments that can be defended in public. It refers to individuals' capacity of agency. Political autonomy on the other hand, transfers this idea into collectivities, by suggesting that collective judgment should be the outcome of public reasoning.

4 It is interesting to mention that Urbinati (2006) criticizes Habermas, by alleging that his model explains the harmonic relations between state and society better than the critical periods when such circularity is obstructed. Nevertheless, this criticism seems inappropriate, since it does not recognize the great effort made by the German philosopher on his model of circulation of power. This model is mostly concerned with situations of crisis, when the outside initiative model may be implemented.

6 Ariel Armony (2004) warns that associations are not always good for democracy. They may even hinder its development. And it is not a matter of just distinguishing a good from a bad civil society, as if only totalitarian groups offered some risk. Armony reminds that several types of associations may deepen social cleavages. He bases his argument on historical examples, showing how this happened in Germany during the Weimar Republic, in postwar USA and during Argentina's dictatorship. In these contexts, several spheres that neo-Tocquevilleans would interpret as sources of social capital were essential to destroy citizenship rights and democratic institutions. Rejecting generic overviews, Armony claims civil society can only be analysed in context. For other examples of discussions for a more cautious analysis of civil society, see Chambers and Kopstein (2001); Dryzek (2005); Gomes (2006); and Marques, Mendonça and Maia (2007).

7 In his formulation of a general theory of political representation, Rehfeld (2006, 4) has argued that representation, in itself, does not have to be legitimate, equal and fair. However, if one thinks of democratic political representation, and if one faces the problem of having to decide which among several actors is best suited to exert representation, legitimacy emerges as a key concept.

8 The idea of public deliberation has a long and varied trajectory, ranging from traditions inspired by Habermasian discourse ethics to ones guided by Rawls's concepts of public reason and overlapping consensus. There are deeply philosophical perspectives and rather empiricist ones. Some focus on argumentative exchanges in formal decision-making arenas, while others are more concerned with a broader societal deliberation. I argue here for a definition similar to the broad perspective advanced by Dryzek (2000b, 86), who seeks to "redefine deliberation in terms of any kind of communication that induces reflection on preferences in non-coercive fashion". For an overview of perspectives on deliberative democracy, see Habermas (1996; 2005); Dryzek (2000a); Bohman and Rehg (1997); Bohman (1998); Chambers (2003); Gutmann and Thompson (2004); Elster (1998); Benhabib (1996); Avritzer (2000); and Maia (2008).

9 It is important to point out that the ideas of Habermas and Dryzek are at the heart of proposals for a deliberative system.

10 Urbinati claims not to work under the framework of deliberative democracy, as she criticizes the proponents of the model for their presumed cognitivist rationalism. However, her interpretation seems mistaken, since the whole proposal of deliberation is to escape the cognitivism advanced by the philosophy of conscience. In addition, Urbinati's idea of judgment could be enriched and deepened if inscribed under a deliberative approach.

11 Although broadly defining accountability as the act of reason demanding and giving, Gutmann and Thompson still somehow tie such acts to elected representatives. They do not consider, for instance, the requirement of deliberative accountability in civic associations. In their own words, "Because deliberative democracy seeks to justify only decisions that collectively bind people, decisions in truly voluntary associations should be less subject to its demands" (Gutmann and Thompson 1996, 34).

12 I am thankful to John Dryzek for the suggestion of a distinction between deliberative accountability and discursive accountability.

13 Note that this connection is produced in the process of communication as a result of the clash of discourses. It is not a pre-condition for the public expression of positions as defended by the Rawlsian differentiation between private and public reason.

Abers, R. N., and M. Keck. 2006. Muddy waters: Decentralization, coordination and power struggle in the Brazilian water management reform. International Journal for Urban and Regional Research 30 (3): 601-622.

Alexander, J. 1996. Collective action, culture and civil society: Secularizing, updating, inverting, revising and displacing the classical model of social movements. In Alain Touraine, ed. M. Diani and J. Clarke. London: Falmer Press.

Armony, A. 2004. The dubious link - Civic engagement and democratization. Stanford: Stanford University Press.

Avritzer, L. 2000. Teoria democrática e deliberação pública. Lua Nova, no. 49:25-46.

______. 2006. New public spheres in Brazil: Local democracy and deliberative politics. International Journal of Urban and Regional Research 30 (3): 623-637.

______. 2007. Sociedade civil, instituições participativas e representação: da autorização à legitimidade da ação. Dados 50 (3): 443-464.

Baiocchi, G. 2005. Citizens and militants. Stanford: Stanford University Press.

Bang, H., and T. B. Dyrberg. 2000. Governance, self-representation and democratic imagination. In Democratic innovation – Deliberation, representation and association, ed. M. Saward, 146-157. London: Routledge.

Benhabib, S. 1996. Toward a deliberative model of democratic legitimacy. In Democracy and difference: Contesting the boundaries of the political, ed. S. Benhabib, 67-94. Princeton: Princeton University Press.

Bohman, J. 1998. The coming age of deliberative democracy. The Journal of Political Philosophy 6 (4): 400-425.

Bohman, J., and W. Rehg, eds. 1997. Deliberative democracy: Essays on reason and politics. Cambridge, Mass: MIT Press.

Castiglione, D., and M. Warren. 2005. Rethinking representation: Seven theoretical issues. Paper presented at the Midwest Political Science Association Annual Conference, 6-10 April, in Chicago, EUA.

Chambers, S. 2003. Deliberative democratic theory. Annual Review of Political Science 6:307–326.

Chambers, S., and J. Kopstein. 2001. Bad civil society. Political Theory 29 (6): 837-865.

Coelho, V. S. P., and M. Nobre, eds. 2004. Participação e deliberação: teoria democrática e experiências institucionais no Brasil contemporâneo. São Paulo: Editora 34.

Conover, P. J., and D. D. Searing. 2005. Studying 'Everyday political talk' in the deliberative system. Acta Politica, International Journal of Political Science, no. 40:269-283.

Dryzek, J. S. 2000a. Deliberative democracy and beyond: Liberals, critics, contestations. Oxford: Oxford University Press.

______. 2000b. Discursive democracy vs. liberal constitutionalism. In Democratic innovation – Deliberation, representation and association, ed. M. Saward, 78-89. London: Routledge.

______. 2005. Deliberative democracy in divided societies: Alternatives to agonism and analgesia. Political Theory 33 (2): 218-242.

Dryzek, J. S., and S. Niemeyer. 2007. Discursive representation. Paper presented at the annual meeting for the American Political Science Association, 30 Aug-2 Sept, in Chicago, EUA.

Eckersley, R. 2000. Deliberative democracy, ecological representation and risk. In Democratic innovation – Deliberation, representation and association, ed. M. Saward, 117-132. London: Routledge.

Elster, J. 1998. Deliberative democracy. Cambridge: Cambridge University Press.

Fung, A., and E. O. Wright, 2003. Deepening democracy. London and New York: Verso.

Gastil, J., and P. Levine, eds. 2005. The deliberative democracy handbook: Strategies for effective civic engagement in the twenty-first century. San Francisco: Jossey-Bass.

Giddens, A. 1990. The consequences of modernity. Stanford: Stanford University Press.

Gomes, W. 2008. Capital social, democracia e televisão em Robert Putnam. In Comunicação e Democracia - Problemas e Perspectivas, ed. W. Gomes and R. Maia. São Paulo: Paulus.

Gurza Lavalle, A., P. P. Houtzager, and G. Castello. 2006. Representação política e organizações civis: novas instâncias de mediação e os desafios da legitimidade. RBCS 21 (60): 43-66.

Gutmann, A., and D. F. Thompson. 1996. Democracy and disagreement. Cambridge, Mass.: Belknap Press.

______. 2004. Why deliberative democracy? Princeton, NJ: Princeton University Press.

Habermas, J. 1996. Between facts and norms: Contributions to a discourse theory of law and democracy. Cambridge: Polity Press.

______. 2005. Concluding comments on empirical approaches to deliberative politics. Acta Politica, International Journal of Political Science 40 (3): 384-392.

______. 2006. Political communication in media society: Does democracy still enjoy an epistemic dimension? The impact of normative theory on empirical research. Communication Theory 16:411-426.

Hendriks, C. M. 2006. Integrated deliberation: Reconciling civil society's dual role in deliberative democracy. Political Studies 54 (3): 486-508.

Herreros, F. 2000. Social capital, associations and civic republicanism. In Democratic innovation – Deliberation, representation and association, ed. M. Saward, 184-194. London: Routledge.

Keck, M. 2004. Governance regimes and the politics of discursive representation. In Transnational activism in Asia: Problems of power and democracy, ed. A. Uhlin and N. Piper. London: Routledge.

Maia, R. C. M., ed. 2008. Midia e deliberação. Rio de Janeiro: FGV.

Mansbridge, J. 1999. Everyday talk in deliberative system. In Deliberative politics: Essays on democracy and disagreement, ed. S. Macedo, 211-239. Oxford: Oxford University Press.

______. 2003. Rethinking representation. The American Political Science Review 97 (4): 515-528.

Marques, A., R. Mendonça, and R. Maia. 2007. O sistema deliberativo e seus espaços discursivos: a articulação entre diferentes modos de comunicação. Paper presented at the XVI Encontro Anual da Compós, UTP, Curitiba.

Meier, P. 2000. From theory to practice and back again – gender quota and the politics of presence in Belgium. In Democratic innovation – Deliberation, representation and association, ed. M. Saward, 106-116. London: Routledge.

Parkinson, J. R. 2003. The legitimation of deliberative democracy. PhD diss., Canberra, Australian National University.

Pitkin, H. F. 2006. Representação: palavras, instituições e idéias. Lua Nova, no. 67:15-47.

Rehfeld, A. 2006. Towards a general theory of political representation. The Journal of Politics 68 (1): 1-21.

Runciman, D. 2007. The paradox of political representation. The Journal of Political Philosophy 15 (1): 93-114.

Saward, M. 2009. Authorisation and authenticity: Representation and the unelected. The Journal of Political Philosophy 17 (1): 1-22.

Smith, G. 2000. Toward deliberative institutions. In Democratic innovation – Deliberation, representation and association, ed. M. Saward, 29-39. London: Routledge.

Squires, J. 2000. Group representation, deliberation and the displacement of dichotomies. In Democratic innovation – Deliberation, representation and association, ed. M. Saward, 93-105. London: Routledge.

Tatagiba, L. 2002. Os conselhos gestores e a democratização das políticas púbicas no Brasil. In Sociedade civil e espaços públicos no Brasil, ed. E. Dagnino, 47-103. São Paulo: Paz e Terra.

Urbinati, N. 2005a. Continuity and rupture: The power of judgment in democratic representation. Constellations 12 (2): 194-222.

______. 2005b. What makes representation democratic? Paper presented at the Political Theory Colloquium, Department of Government, 17 November, in Harvard University.

______. 2006. Representative democracy: Principles and genealogy. Chicago: University of Chicago Press.

Wampler, B., and L. Avritzer. 2004. Públicos participativos: sociedade civil e novas instituições no Brasil democrático. In Participação e deliberação, ed. V. Coelho and M. Nobre. São Paulo: Editora 34.

Warren, M. 2001. Democracy and association. Princeton, NJ: Princeton University Press.

Young, I. M. 2000. Inclusion and democracy. Oxford, UK: Oxford University Press.

Submitted in September, 2008.
Accepted in December, 2008.

* I have discussed previous versions of this paper with John Dryzek, Leonardo Avritzer, Selen Ayirtman, Bora Kanra, Simon Niemeyer, Melissa Lovell and Penelope Marshall. I am thankful to them for their valuable comments. The paper has also benefited from a discussion in a work-in-progress seminar at the Political Science Program of the Research School of Social Sciences at the Australian National University. I am also indebted to Rousiley Maia and to the anonymous reviewer of BPSR for their valuable contributions. Lastly, I am grateful to Fapemig and to Capes for their support.