European civil society: Between participation, representation and discourse

Core ideas of this article were first elaborated as a comment to Beate Kohler-Koch's intervention on “Political Representation and Civil Society in the EU” at the Connex Thematic Conference on Political Representation. European University Institute Florence, May 25–26, 2007. The author would like to thank Beate Kohler-Koch for the very fruitful discussion on this topic.

Abstract

Institutional designs of governance in the European Union frequently underlie a dichotomy between participation as the realm of civil society and representation as the realm of national governments and parliaments. The aim of this paper is to consider organised civil society not as distinct from but as part of the multi-level representative field that is emerging in the EU. This is done by distinguishing two distinct mechanisms of political representation in aggregating individual preferences or in integrating the political community of the EU. In order to spell out this latter integrative function of political representation as a creative practice, the notion of representative claims-making will be introduced. The practice of representative claims-making can then be analysed, first of all, as a way of distributing the social capital of the actors and institutions that populate the European field of civil society activism. Secondly, the practice of representative claims-making can be analysed as a way of building new forms of cultural and symbolic capital of civil society that are needed to occupy the new transnational positions that are made available by European integration.

1 Introduction

Institutional designs of governance in the European Union frequently underlie a dichotomy between participation as the realm of civil society and representation as the realm of national governments and parliaments. The aim of this paper is to consider organised civil society not as distinct from but as part of the multi-level representative field that is emerging in the EU. This is done by distinguishing: (a) three conceptual frames of conceiving of the representativeness of organised civil society in relation to EU-governance, and (b) two distinct mechanisms of political representation in aggregating individual preferences or in integrating the political community of the EU. In order to spell out this latter integrative function of political representation, the notion of representative claims-making will be introduced. Representative claims-making can be analysed, first of all, as a way of distributing the social capital of the actors and institutions that populate the European field of civil society activism. As such, it points at the contentious logics of the field, in which social positions are unequally distributed, thus creating legitimatory constraints for some to defend their hegemonic position against their alleged constituents and for others to challenge the hegemony of the self-acclaimed representatives. Secondly, representative claims-making can be analysed as a way of building new forms of cultural and symbolic capital of civil society that are needed to occupy the new transnational positions that are made available by European integration. As such, it points at the cultural logics of the contentious social field, in which collective actors always perform in front of a larger audience.

Opening the black box of representation turns away from “civil society” as a domesticated and organised space for participation and brings in a different notion of “(civil) society” as a discursive field for making claims of representation and legitimacy. Representation thus shifts the attention from “civil society” as the intermediary realm of activated citizenship, voice and participation to the “social constituency” as latent structure, image and identity (Fossum & Trenz, 2006).

2 The promise of participatory governance

Civil society has been traditionally conceptualised as the realm of voluntary action and participation activating the citizens and channelling their voice into the system of political representation. This is in line with mainstream political thinking which has discussed representative-parliamentary and direct-participatory government as alternatives referring to mutually exclusive modes of democratic legitimacy (Held, 1987, p. 4; Plotke, 1997) and insisting on a democratic division of labour between institutional and non-institutional politics (Habermas, 1996a,b, pp. 329ff.). Also in empirical terms, civil society is frequently perceived to act as the opponent of the elected representatives by mobilising the direct voice of the citizens from outside or from below (Tilly, 2004).

Following this main line of political thinking, European civil society has been mainly analysed as a substitute for representative democracy (Finke, 2007). It has been argued that representation in a polycentric non-state polity would be problematic in principle and counter-effective since there is no common ground shaped by cleavages, coalitions and identities that could be used for building representative relationships (Abromeit, 1998). The allegedly unstable majorities of the European Parliament would be only very indirectly representative of a European electorate and there would be no stable public opinion on which representative government could rest. Such structural limitations in models of representative democracy have thus enhanced the search for participatory models of legitimate governance, where civil society is the most significant player (Kohler-Koch & Finke, 2007; Nanz & Steffek, 2007).

Such a trust in participatory governance is also shared by European Union institutions. In its more recent strategic papers, the Commission has made its preference for more flexible and dynamic arrangements of participatory democracy explicit, relying on the direct inputs of citizens expressed through stakeholder networks and forums. Such proposals are carried by the conviction that the channels of parliamentary representation can, in principle, be sidestepped by consulting as wide a range of stakeholders as possible before proposing new legislation or new policy initiatives.¹ The principle of good governance is thus expressed in the guarantee of participation of all affected parties in consultation, which is put into practice by the Commission as the only instance with the mandate to pursue the European “common interest” (Kohler-Koch, 2007, p. 18). In this way, a supranational decision-making body would be enabled to build direct interactions with society. It would be the instant receiver of societal inputs and demands, and would no longer depend on the allegedly “erratic” outputs of the deficient direct and indirect representative channels – European Parliament and national governments – that mediate between the EU and the member states.

Empirical accounts of the practices of consultation indicate that the Commission has indeed expanded its participatory regime (Kohler-Koch & Finke, 2007; Smismans, 2007a) inviting differently affected groups for participation, thus channelling citizens’ voices. Civil society advocates profit from the Commission's new responsiveness in a number of ways. Through its funding practice, the EU keeps the infrastructure of European networking alive and most transnational campaigns and initiatives would not run without financial, technical and often even ideational support from Brussels. Civil society advocates increasingly find themselves in a partnership with supranational governance and, eventually, re-directing their critical voices from the EU to the governments of the Member States (Ruzza, 2004; Trenz, 2007).

3 The conspicuous absence of political representation in civil society discourse

While acknowledging its innovative and progressive character, participatory governance in the EU has been also confronted with a rather harsh normative critique. Lord (2007) has noted that the current practice of “participation by invitation” is only a disguise for the problem of representation. Although some principles of “good governance” are formulated to guarantee the representative quality of stakeholder networks involved in European governance, the question remains unsettled as to who guards the guardians of representativeness in the EU (Lord, 2007, p. 149). Greven (2007, pp. 244–5) has formulated a categorical critique of participatory governance, which, in his view, is a manifestation of a technocratic practice of authoritative problem solving. Participatory governance would stand for the “structural inequality of memberships” in organised civil society against the principled equality among citizens: “Citizenship in a democratic system is not voluntary or optional in the same manner as in civil society” (ibid.). Participatory governance is therefore nothing more than a “private contract” between government, voluntary associations and its respective members: “But in the very moment when non-members are affected by these private contracts a different form of legitimacy is required, which (…) can only be located in representative offices” (ibid.).

Participatory governance has been further criticised for its technocratic character. By assuming a substitute role with regard to the representative channels provided by national governments and parliaments, organised civil society would risk becoming merely an auxiliary of depoliticised EU-governance. In the new institutional and constitutional setting of the EU, this auxiliary function of participatory forms of government is enshrined in rules for civil society consultations and expertise through stakeholders, specialists and professional activists. Nevertheless, the Commission perceives the inclusion of civil society as an instrument to strengthen the representative elements of the EU. The auxiliary function of civil society as a tool to enhance the representativeness of EU-governance is laid down in the Lisbon Treaty, stipulating the participation of “representative associations” in the democratic life of the EU. With regard to the enforcement of this principle, the European Commission's role as a gatekeeper of civil society participation and dialogue is reconfirmed (Art. 11(3)). This would enshrine a technocratic understanding of civil society as a tool of “good governance”, which would empower the Commission to identify (and subsequently represent) common European interests. The function of representation would thus ultimately rest with the Commission as the guarantor of procedural fairness and inclusion of relevant interests and perspectives (Kohler-Koch, 2008, p. 17).

This self-description of the role of the Commission as the guardian of representation by European civil society has remained surprisingly unchallenged within the academic community and even within civil society at large. Kohler-Koch (2007) makes this point when speaking of a conspicuous absence of representation discourse in the official talk on EU-governance. In other words, representation in relation to civil society is a non-issue.

On the critical side, one might expect that there are systematic reasons for this failure to reflect more thoroughly the questions of political representation in European civil society discourse. Kohler-Koch (2007) attributes a political rationality to the fact that representation is not mentioned in relevant documents and strategy papers issued by the Commission. When the Commission launched the White Paper on European governance, it had no interest in overburdening the agenda of institutional reform or challenging the intra-institutional equilibrium (especially its relationships with the governments of the Member States and the European Parliament). A commitment to participation was the most plausible and also the least controversial approach to improving European governance. More critically, the insistence on the representation–participation dichotomy can be also read as a form of social control. By separating civil society participation and issues of representation, European institutions can either locate social interest actors outside the realm of representative government or legitimately claim to subordinate and control them as an auxiliary of governance.

This political rationale would explain path-dependent or strategic institutional choices. It does not explain, however, why representation is also a missing category in theoretical and normative accounts of EU-governance and civil society.² At this point, there is a need to open up the conceptual discussion, which has perceived civil society participation thus far as a substitute for representative democracy. My argument, to be developed in the following, is that beyond the political rationale, there is also a theoretical rationale for blending out issues of representation in EU-governance research. By focusing on flexible participatory arrangements, the theoretical and normative debate could more easily detach civil society from its supposedly organic link to the nation state and focus on conceptions of the emerging global or cosmopolitan civil society. The new agenda of participation was thus part of the programme to overcome social sciences’ “methodological nationalism”, which consists precisely in avoiding the kind of substantialist references to territory, state or collectivity, which are conventionally used for constructing political representation (Beck, 2003).

The problem then lies in the ways to link – conceptually and pragmatically – the unbound civil society back to the newly emerging institutional and procedural settings of transnational governance. Against the insufficiencies of institutional accounts of partnership governance, which seek to internalise the problem of political representation by appropriating representative claims and affirming that only the Commission can formulate the common European interests, my account proposes to externalise again the problem of political representation as constitutive to civil society practice. I thus assume that there is a hidden agenda of representation underlying the very architecture of European civil society, which is built on a multi-level system of interest representation in which associational actors are seen as intermediaries called upon to aggregate individuals’ preferences into collective demands and thus bridge the gap between the EU and its citizens. It is then necessary to observe how civil society actors comply with this double role ascription as participants in EU-governance and as representatives of their own – frequently undefined – constituents.

The theoretical argument to be developed in the following is that civil society activism can only be linked to formal political structures and institutions if it is incorporated into the framework of representative politics. Political representation is therefore not secondary or inadequate but it is constitutive of civil society. This thesis of the intrinsic relationship between civil society and political representation can be corroborated to the extent to which European integration is developed as a political project. Representation needs to be reconstructed from a hidden agenda to an explicit one, by deconstructing the justificatory discourse of participatory governance. Only by disentangling the participatory conundrum that has developed in relation to EU-governance over the last two decades can we arrive at a normatively adequate and practically satisfactory clarification of the role of political representation in relation to organised civil society and EU-governance. Representation is then no longer seen as a form that lies outside participatory governance and to which civil society should relate in one way or another, but, as a key mechanism that shapes civil society from inside, and accounts for its dynamic unfolding.

4 Organised civil society and representative governance: From partner to constituent

Civil society involvement in the official discourse on EU-governance is meant to complement the elements of representative democracy on which the EU is founded (Kohler-Koch, 2008, p. 12). The particular answer provided by the Commission for strengthening its own outputs in terms of efficient and democratic decision-making lies in promoting the concept and the agenda of good governance. It is the way of gradually incorporating European civil society as an auxiliary of governance, which adds to the efficiency, functionality and overall legitimacy of EU policy-making (Heinelt, 2005). In this view, the conditions for “good governance” depend on what Kohler-Koch (2007) aptly calls “participatory engineering”. This practice results in a fusion of governance tasks in policy-making and implementation, where institutional actors and social actors are increasingly relying on each other.

Over the last two decades, the role of civil society in relation to EU-governance has been defined in an expansive way, from expertise and close lobbyism to open consultation and citizens’ dialogue (Bignami, 2003). If the EU is primarily conceived as a regulatory political system which relies on expertise and information inputs at various levels, representation is only relevant at the highest level of political aggregation. It falls back upon the Commission to articulate the common European interest out of the plural inputs and contributions, and to translate it into “adequate” policy choices. If the EU is conceived as a system of participatory governance, which instrumentalises so-called stakeholders as co-producers of efficient and effective policy regulations, representation also becomes important at the individual level. The representativeness of the invited partners determines the quality of the aggregated interest. If the EU is ultimately conceived as an emergent polity embracing a political community formed through active citizenship, representation is needed to demarcate the political community. We would thus observe a social constituency in the making, which is comprised by a process of collective will formation.³

On the one hand, in the regulatory mode of governance, civil society is clearly subordinated to governance tasks and produces inputs “on demand” in the form of knowledge and information that is needed by EU-institutions to improve governance performance. In the partnership mode, civil society becomes the co-author of governance. It is ultimately defined as an intermediary arena of participation, which is incorporated by governance but still maintains sufficient independence to follow its own agenda. On the other hand, the polity-constituency mode invokes an indeterminate state-society relationship that still needs to be given meaning and defined in inclusive terms. It allocates the representative role to the whole of the political community, which is engaged in a process of collective will formation. Civil society is still perceived as the sphere of interest intermediation but, as such, it is grounded in a general public sphere, which has the potential to mobilise a collective will in support of or in opposition to power holders. In order to fulfil this function, civil society needs to be discursively (re-)embedded; in other words, it needs to generate mass, or at least sector-specific, public debates and discourses that include the whole of the political community.

5 Representation as an aggregative and as an integrative mechanism

Following the expansive logic of modelling EU-society relations, representation is brought back in two different but still largely unspecified ways. The first of the two modes measures the representativeness of individual civil society organisations that are (or claim to be) involved as partners in EU-governance. This brings in representation as an aggregative mechanism of preference formation. The latter mode measures the performance of civil society as a whole in representing the common good by integrating (or creating) a common interest. This brings in representation as an integrative mechanism of civil society in relation to the citizenry as a whole. In the first case, representation depends on the capacity of organised civil society to aggregate citizens’ preferences. Representation is thus grounded in a linear power relationship between the principal (the constituent) and the agent (the delegate), which runs forward through an explicit mandate of the constituency, ideally through electing the delegates. In the second case, representation depends on the systemic performance of government and organised interests to ensure the integrity of (civil) society. Representation thus moves backward from the representative who is trying to anticipate the beliefs of its potential constituents.

Table 1 spells out two mechanisms that account for the active making of representation in EU-society relations. In both variants, representation is understood as a creative process that constructs the representative (the traditional aggregative mode) but that also has repercussions on the self of political representation and in this sense can be said to depict the represented or the underlying social constituent (the integrative-identitarian mode). To argue in this way is to stress the performative function of political representation as a twofold constitutive relationship. The aggregative and the integrative mode of representation are relying on a two-directional process of building the representativeness of civil society in relation to government and its social constituents.

Table 1 Mechanisms of political representation in EU-society relations.

Representativeness	Type
	Representation as an aggregative mechanism	Representation as an integrative mechanism
Criteria for measurement	Equal consideration	Political justification
Relationship	linear (principal–agent)	Systemic
Focus	Capacity of civil society to aggregate citizens’ preferences and to hold government accountable	Capacity of government and organised interests to guarantee the integrity of (civil) society
Means	Participatory claim Societal self-organisation	Representative claim Social engineering
Direction	Represented constitute the representative	Representative constitute the represented

According to the aggregative mechanism, the patterns of representativeness of civil society in relation to EU-governance are identified in a forward-looking move from the principal (the represented) to the agent (the representative). The principle of “equal consideration” applies in so far as equal weight should be given to the interests of each. This corresponds to the imaginary of society made up by individuals (i.e., the single members of the constituency) that constitute the whole (i.e., the polity). The represented constitute the representative. According to the integrative mechanism, the patterns of representativeness of civil society in relation to EU-governance are identified in a backward-looking move from the representative to the represented. The principle of political justification applies, arguing that collective choices need to be defended beyond the question of what is the common concern and what serves the public good. This corresponds to the imaginary of the polity as a hegemonic entity which constitutes the single parts (a constituency made up of citizens imbued with rights and duties). The representative constitutes the represented.

The first mode of conceptualising representation as an aggregative mechanism delivers a rationale for the self-organisation of civil society. The classical assumption is that civil society needs to sustain its autonomy against the state and against the market. As such it can only be self-organised, emerging from the spontaneous action of free and equal citizens (Calhoun, 2001). The representativeness of European civil society is based on its self-image and identity as a collective actor that is different from the state and the market.

The second mode of approaching the question of representation as an integrative mechanism delivers the rationale for civil society engineering from above. The classical assumption is that civil society as the world of independent associations is embedded in a moral and normative (legal) order, which is protected by the state. As such, the organised forms taken by civil society have to take account of the integrity of the political community or the polity to which it is bound by the imaginary of self-rule and popular sovereignty (Taylor, 2002). The representativeness of civil society is based on the image of the polity as a whole. It rests with the citizens’ state and the various forms of virtual representation invented by it. Or, in the case of the EU, according to the tentative conclusion of Kohler-Koch (2007), it ultimately rests with the European Commission, which defines itself through participatory governance.

6 Beyond delegation and interest aggregation

Why has the traditional perspective of conceiving “representation” as an aggregative mechanism proven insufficient for determining the representativeness of organised civil society in relation to EU-governance? First of all, civil society representatives are normally not directly and equally elected by their constituencies. The problem of delegation through the aggregation of individual preferences is aggravated due to the multinational set-up of the EU. Studies of the EU-organisational world have repeatedly shown that EU social actors tend to be free riders rather than delegates. Associations or NGOs which are active at the transnational level need to refer to their constituents in an acclamative way. They are also only rarely relying on formal membership as a criterion that could be used to determine their representativeness. Their special knowledge and expertise turns them into general interest entrepreneurs who are generally trusted but not made publicly accountable (Imig & Tarrow, 2000; Ruzza, 2004; Trenz, 2007). The good governance agenda is one answer to this problem in the sense of allocating the task of selecting and aggregating societal interests within the Commission as a guarantor of equal consideration of societal interests and exempting social stakeholders from representative tasks (Smismans, 2007b).

Such limitations for applying the aggregative-electoral mechanism of representation to the case of EU partnership governance speak for the urgency of conceiving alternative, non-electoral modes of political representation and assessing their legitimatory potential. The literature on civil society has emphasised the impact of the integrative mechanism in terms of trust in expertise, reflexive capacities, moral integrity or simply advocacy and advertisement by mostly self-appointed civil society representatives.

Rethinking representation in the framework of deliberative democratic theory and for the context of parliamentary representation, Jane Mansbridge has conceived of the represented–representative relationship as one that is not necessarily based on a linear power relation but on “reciprocal power and continuing mutual influence” (Liebert, 2007; Mansbridge, 2003, p. 518). Replacing the dyadic principal–agent model, representation becomes a systemic phenomenon (Saward, 2007, p. 6). As such, it would require EU-governance to be grounded in a discursive relationship between EU-institutions and citizens, mediated by organised actors who perform in front of a larger audience.

Representation of civil society is thus specified as a form of cultural creativity that unfolds as a particular semantic within the public sphere (see Eder in this issue). It then needs to be understood how public discourse, instead of reading off civil society's objective interests or expressing its inherent identity, constitutes civil society by representing it as a particular form of culture and solidarity (Calhoun, 2002, p. 159). I will use this insight in the creative and constitutive force of public discourse to propose a different notion of civil society as a discursive formation within the public sphere (see Trenz (2005) for details on this proposal). This allows me to conceive the emergence of a European civil society not as a matter of new structures of collective action and organisation but as a new way of imagining identity, interests and solidarity (Calhoun, 2002, p. 171).

In the following, this integrative mechanism of political representation will be further specified by adapting Mansbridge's (2003) helpful distinction of four modes of representation to the case of civil society. Mansbridge shares our concern in the insufficiencies of the traditional mode of perceiving representation as an aggregative mechanism (of what she calls promissory representation) and classifies three additional modes of representation, each of which, as I will try to demonstrate, has the potential of fulfilling an integrative (systemic) function with regard to the dynamic and reciprocal polity-constituency relationship, in which civil society is embedded:

•	Anticipatory representation transcends the principal–agent model by focusing on the continuous communicative exchanges between the represented and the representatives that are facilitated through the mass media. In modern media democracies, political rulers not only address the voters at the occasion of an election but they are put under constant constraints to justify their political choices in front of the wider audience. In the case of the EU, organised civil society would facilitate such permanent communicative exchanges between the EU and its citizens. It would give EU power holders the chance to observe reflexively and to anticipate the preferences of their constituents. By establishing a reciprocal relationship of power and of continuing mutual influence, power holders as well as constituents become educable and malleable (Mansbridge, 2003, p. 519). The legitimacy of the representative-constituent relationship can then be measured in the deliberative quality of communication, which is driven by the merit of argument and justification.
•	Gyroscopic representation relies on forms of generalised trust towards the representatives whose alleged attitudes and characteristics make their future behaviour for the voters predictable. In the case of the EU, the constituents would need to share a particular preference or identity with their representatives (either EU-institutions or societal stakeholders) but would not try to induce the representative's behaviour (Mansbridge, 2003, p. 521). Opinion polls show, for instance, that civil society actors whose commitment to the public good is beyond question are generally much more trusted than elected representatives. Also the permissive consensus towards the EU can be interpreted as a case of gyroscopic representation. It allows the EU system of governance to be operated with considerable discretion. Its legitimacy is measured by generally accepted outputs without necessarily having to rely on regular inputs or on traditional forms of responsiveness and accountability (Majone, 1998).
•	Finally, the case of civil society points to the relevance of surrogate representation, in which the represented are chosen by their representatives. The surrogate representative typically chooses an issue (e.g., minority rights), which she persistently defends within the political arena even though members of the particular community (e.g., minorities) have no direct relationship with the person in office. In the case of the EU partnership governance, civil society activists are found to frequently act as surrogate representatives of such groups without a direct voice in the political process; environmental groups claiming to represent future generations, human rights groups claiming to represent political refugees or a local pressure group claiming to represent a particular neighbourhood. Also in this case, the representative is not accountable or even responsive to the represented. Crucial to this process is rather the idea of a representative's claims, which call into existence the represented as an addressable entity (Saward, 2006, 2007). The legitimacy of surrogate representation therefore lies in the performative act of representation itself, through which a particular person claims to speak for somebody else and in the general resonance which this performative act creates within a wider audience (which is not identical with the constituent).

Anticipatory, gyroscopic and surrogate modes of representation differ in how they conceive of the integrative (systemic) function of political representation beyond the traditional principal–agent model of a linear aggregation of individual preferences from the constituents to the representatives. The three modes are relying on discursive acts linking different actors’ positions through justificatory logics and intermediary performances to create resonance within a wider public. The presence of a third (the audience or the general public) is seen as an integral element of political representation. Last but not least, it is the character of this dynamic discursive relationship that constitutes the representative and the represented. Table 2 summarises some of the characteristics of these different modes of building a systemic representative relationship. Differences lie in the ways the discursive-representative act is enacted, what kind of normative criteria are evoked by it, what kind of systemic operation is performed and who is potentially included/addressed by it.

Table 2 Representation as an integrative mechanism.

	Anticipatory representation	Gyroscopic representation	Surrogate representation
Integrative mode	Communicative exchange	Trust	Advocacy
Normative criteria	Quality of deliberation between represented–representative	Authenticity	Credibility
Performance	Education	Identification	Acclamation
Target	Reciprocity between represented and representatives	Relative autonomy of the representative	General public

The conceptual link between civil society and political representation postulates a two-directional process, in which the linear principals–agent relationship is replaced by a non-linear dialogical relationship. Accordingly, the representativeness of an actor is not determined as a zero-sum relationship through elections, which establish who is elected (and therefore representative) and who is not (Saward, 2006, p. 299), but as a positive sum relationship, which involves represented and representatives in a continuous process of collective will formation.

Representation by organised civil society in the EU can thus be said to work along two different directions. The first direction still follows the path of traditional aggregative “promissory representation”. EU associations enter as brokers of aggregated citizens’ preferences into the political arena. In this function, they receive signals from the citizens, select and transform them into positions which are directed at the European Commission and other institutional bodies of the EU. Interest politics in the EU have unfolded for several decades now, but a convergent system of European interest representation is still out of sight. The EU has opened a plurality of channels for lobbying and consultations, but national interest profiles and national channels of interest intermediation prevail. European umbrella organisations have notorious difficulty in finding the lowest common denominator of their members’ interests and are therefore frequently handicapped in expressing their positions in EU negotiations (Heinelt, Getimis, Kafkalas, Smith, & Swyngedouw, 2002).

The second direction opens civil society participation up to deliberative “anticipatory representation”. Beyond the function of interest mediation, EU associations also perform as partners of EU-governance. In this role, they improve the governmental outputs by giving expertise and applying the norms of deliberation and, at the same time, by enhancing the quality of information between the EU and its citizens. To the extent that the deliberative mode of collective decision-making has gained prominence in the EU setting, forms of anticipatory representation can develop through which civil society delegates progressively defend the interests of their constituents. While the responsiveness of political institutions in a deliberative setting can be measured by their degrees of participation, inclusiveness and plurality of procedural designs, representation as such is not quantifiable. It is a quality that is acquired through discussion (Eriksen & Fossum, 2007, p. 8, based on Manin, 1997) and it is linked to the search of common understanding, which unfolds through free, equal and open debates in an all-inclusive public forum.

Political representation also comes in, however, when it needs to be decided who is entitled to sit in the forum and to actively participate in the debate and who may attend the debate as a passive observer. This problem has often been sidelined by reference to the abstract principle of publicness, which – as empirical research has repeatedly shown – is only an insufficient guarantee of publicity and attention as long as an encompassing resonance body in the form of a European public and media sphere is not available. Others recur again to statistical representativeness to resolve the problem of selecting voice in participation. Deliberative polls, for example, rely on random samples of citizens’ representatives, but are so far rather understood as experimental designs without enforcing capacities (Fishkin & Lushkin, 2000). One possible solution is to bring in additional forms of gyroscopic and surrogate representation to explain the integrative mechanism of political representation in relation to the general legitimacy of EU civil society in relation to EU-governance.

7 Assessing European civil society as a gyroscopic and as a surrogate representative

The impact of gyroscopic representation can be measured through generalised trust in the political system in general and NGOs and civil society actors in particular. The proliferation of civil society is often seen as a symptom of the crisis of trust in the institutions of representative democracy. Opinion polls like Eurobarometer show that trust in representative institutions among the populations of the old and – even more so – the new Member States tends to be low (with the noticeable exception of Scandinavian countries). NGOs and international organisations like the UN, but also the European Union to some extent, are more trusted than national governments, political parties and even parliament. Such patterns of generalised trust indicate the relevance of gyroscopic representation in which the represented identify with particular principles or share a moral conscience that is defended by particular organisations (a church, or Greenpeace) or by charismatic personalities. The representatives in turn (the particular organisations or prominent persons) can be largely autonomous and unconstrained by public opinion changes. They are trusted not because they follow the majority vote but because they are independent from it and rotate like gyroscopes on their own axes (Mansbridge, 2003, p. 520).

Gyroscopic representation has the potential to release patterns of trust and political loyalty from the bonds of the nation states. Non-elected, gyroscopic trustees are often transnationally active (the pope, the Dalai Lama or even pop stars). They stand for the moral conscience of the world and not of a particular community. As such, they become integrating figures of global civil society but are only loosely related to the technocratic setting of EU-governance, which is rather characterised by the lack of charismatic figures and prominence. Eurobarometer, which measures trust in political institutions, delivers only rough data on civil society. The expression of a high level of trust in single international NGOs and general interest representatives, such as Greenpeace or Amnesty International, cannot be turned into conclusions about the representativeness of European civil society as a whole.

The analysis of the impact of surrogate representation requires a more qualitative research strategy and the application of ethnographic and discourse analytical methods. Instead of quantifying representation or applying standardised procedures and norms, attention is turned towards the practice of representative claims-making. Rather than asking who constitutes civil society, one can instead ask how civil society goes about the task of being representative. What can then be answered through interview techniques, questionnaires or text analysis is what kind of representative claims are seen as “socially approved” and “appropriate” in a transnational representative field as constituted by the EU.⁴ The research focus is thus on “the active making (creating, offering) of symbols or images of what is to be represented” (Saward, 2006, p. 301). The representatives are observed in their role of “choosing their constituents” and “portraying them or framing them in particular, contestable ways” (ibid., pp. 301–2).

When asked about their own perceived constituencies, civil society activists have rather diffuse groups in mind and refer to general and undetermined categories like citizens, the global community and sectoral or territorial groups of various size and scope. As advocates of global concerns, internationally active NGOs tend to be free-floating and detached from territorial constituencies or the interests of particular memberships (Keck & Sikking, 1998). Their strength lies precisely in de-coupling and re-embedding local concerns in global discourse and world polity (Meyer, Boli, & Ramirez, 1997). They thus reject the idea of membership accountability and question the legitimacy of any particularistic expression of popular sovereignty. Cosmopolitan civil society makes a case for surrogate representation precisely to denounce the artificiality and arbitrariness of traditional representative settings. Surrogate representation thus sets the grounds for cosmopolitan claims-making.

Representative claims remain highly contested from within civil society and from the outside. By focusing on the performative side of representative claims-making, this justificatory practice of public discourse is put centre-stage. In this sense, representation is replaced by public discourse. The reasons for this replacement are twofold: Firstly, because political representation remains parasitic on deliberation, “as no person can consider herself to be legitimately represented unless the mandate and accountability terms are spelled out, and the represented are offered acceptable justifications for decisions taken on their behalf” (Eriksen & Fossum, 2007, p. 9). Secondly, the good reasons and justifications through which the representative quality can be acquired in public debates are themselves in need of representation to be conveyed from deliberating bodies to larger audiences. In this sense, research needs to focus on the performative and aesthetic component of representative claims-making in shaping its own discursive referents. Surrogate representation is therefore not distinctive from other more “authentic” modes of political representation. More accurately, it is all what political representation is about as the process of defining and redefining the relationship between the political rulers and the citizens.

The notion of surrogate representation as manifested in representative claims-making thus has an important impact on a theory of political representation and a theory of civil society. Ultimately, it is the ongoing discourse within the public sphere that represents political actors, opinions and reasons and makes them significant. Civil society is then no longer perceived as a field of participation but as a projection of representative discourse. We thus arrive at a different notion of civil society, as a discursive formation and not as a collective actor or a social structural entity, and we arrive at a different notion of political representation not as a formal, one-to-one relationship between well-defined social units but as a signifying practice of a shifting and mutually constitutive relationship.

8 Conclusion

European civil society seems possible but it is less to be described in socio-structural terms and it is also only insufficiently grasped as an organisational feature of collective action. European civil society seems mainly possible as a collective act of representation. Following this main line of argumentation, the article has emphasised that political representation is a significantly broader topic than suggested by traditional political theory. In reconstructing the different modes in which the representativeness of European civil society can be conceived and assessed, it has been argued that representation is only insufficiently described as the aggregation of individual preferences and the empowerment of political delegates by the vote of the citizens. Beyond the principal–agent model, there is a second, systemic function of political representation, which is grounded in a dynamic and reciprocal polity-constituency relationship, based on deliberation (or political discourse).

In approaching this systemic function of political representation as the safeguard of the integrity of the political community, this article relies on a classification proposed by Mansbridge (2003) to identify three additional modes of reconstructing the representativeness of civil society. Anticipatory, gyroscopic and surrogate representation are grounded in a triadic discursive relationship between the political rulers, organised civil society and the general audience. Political representation is thus embedded within a plural and dynamic social field, in which delegates not only need to relate formally to their constituents but also constantly justify their choices in front of the larger audience (the anticipatory mode), engender trust and loyalty (the gyroscopic mode), or respond to the generalised expectation of being representative (the surrogate mode). In analytical terms, this implies the need to measure the creative force of political representation in promoting justificatory logics, engendering collective identities and, last but not least, also in constituting particular actors’ positions as signifying and as addressable units.

In order to spell out how political representation unfolds as a creative practice, the notion of representative claims-making has become central. Representative claims-making does not simply refer to the technical aspect of political representation in drawing the distinction between the representee and the representative. Representative claims-making is rather embedded in a double constitutive relationship of what Saward (2006) has called the “active constitution of constituencies” and, as we can add now, the simultaneous and interlinked constitution of the polity. As such, it refers to a double claim about the “aptitude or capacity of a would be representative and about relevant characteristics of a would be audience” (Saward, 2006, p. 303). There is a performative and an aesthetic moment in political representation as a process of constituting the representee as well as the representative.

In this sense, political representation can be said to introduce a distinct and novel notion of European civil society as a collective act of imagination that is needed for it to enter into democratic practice and to apply the discourse of democracy to the European political setting. Representation is at the heart of civil society activism. It is not second best because deviating from the participatory ideal, but it is what democratic politics are all about (Plotke, 1997; Saward, 2007, p. 5). Civil society as a unitary construct of representation is what is comprised in anthropological studies of ritual and magic as a fetish: a non-casual by-product of a collective act of imagination. The research attention then shifts from “civil society” as collective agency to “civil society” as discourse, which generates particular ideas and images about the constitution of collective agency.

In this last sense, the notion of representative claims-making is meant to pave the way for an integrated analysis of the multi-level field of representation that new and old modes of EU-governance have opened up. As stressed by Crum and Fossum (2009, pp. 14ff.), such a field is integrated by a set of basic values and procedural norms that constitute institutionalised patterns of interactions and that tell us how democratic representation ought to look like in the EU. As has become clear by now, the contours of this field stretch beyond the channels of parliamentary representation to include a wider range of actors committed to the practices of “claiming” for rights and democracy in the EU. The practice of representative claims-making can then be analysed, first of all, as a way of distributing social capital among the actors and institutions that populate the European field. As such, it points to a contentious logic operating within the field in which social positions are unequally distributed and creating thus legitimatory constraints for some to impose their hegemonic position against their alleged constituents and for others to challenge the hegemony of the self-acclaimed representatives (Eder & Trenz, 2003). Secondly, the practices of representative claims-making can be analysed as a way of building new forms of cultural and symbolic capital that is needed for taking advantage of the new transnational positions that are made available by European governance. As such, it points at the cultural logic of a field where collective actors always perform in front of larger audiences. It is through this creative performance of political representation within the public sphere that the symbols and norms of European integration are enacted and transmitted. Political representation is then indeed what democracy is all about: giving meaning to the EU polity and its social constituencies and connecting them through the narratives of a double constitutive relationship.

Notes

Core ideas of this article were first elaborated as a comment to Beate Kohler-Koch's intervention on “Political Representation and Civil Society in the EU” at the Connex Thematic Conference on Political Representation. European University Institute Florence, May 25–26, 2007. The author would like to thank Beate Kohler-Koch for the very fruitful discussion on this topic.

1 The Commission's guide to Eurojargon defines a stakeholder as: “Any person or organisation with an interest in or affected by EU legislation and policymaking” (http://europa.eu/abc/eurojargon/index_en.htm, last seen on 31 May 2007).

2 Especially when taking into account that conceptual history has always emphasized this constitutive link between civil society and the imagination of the citizenry (Calhoun, 2001).

3 See Kohler-Koch (chapter 3 in this special issue) for the further explication of these modes of governance with civil society.

4 ‘Representative claims’ as units of analysis can be singled out from different data sources (texts or interviews) as public speech acts which articulate political demands, calls to action, proposals, and criticisms, which, actually or potentially, affect the interests or integrity of claimants and/or other collective actors in a specific issue-field (Statham and Trenz, 2008). The exercise of “mapping representative claims-making” would then imply the task to compare data sources from different actors who claim for or who dispute public legitimacy.