WHICH EUROPEAN PUBLIC SPHERE?
NORMATIVE STANDARDS AND EMPIRICAL INSIGHTS FROM MULTILINGUAL SWITZERLAND ANKE TRESCH
Abstract
Since the beginning of the 1990s, the EU has been increasingly criticised for its democratic deficit, which is intrinsically linked to the absence of a public sphere at the European level. Whereas scholars consider the emergence of such a public sphere as a necessary requirement for the democratisation of the EU, they disagree on the conceptualisation and normative requirements for a meaningful public sphere at the European level. This article takes an empirical perspective and draws on the nation-state context of multilingual Switzerland to get insights into what a European public sphere might realistically look like. Based on a content analysis of the leading quality paper from each German- and French-speaking Switzerland by means of political claims analysis, it shows that three of the most often cited criteria for a European public sphere - horizontal openness and interconnectedness, shared meaning structures, and inclusiveness - are hardly met in the Swiss context. On this basis, it concludes that the normative barrier for finding a European public sphere might be unrealistically high and should be reconsidered.
Anke Tresch is Lecturer in the Department of Political Science and International Relations, University of Geneva; e-mail: anke.tresch@unige.ch.
r^. ■
LO si
ro
rN
i—I
o rN
o~>
£
£ 3
a
0 £
1 *
(R 0 C
! LO
Introduction
Despite significant advances in the European integration process since the early 1990s - in the form of a widening from 12 members in 1986 to the current 27 through different rounds of enlargement, and a deepening through the establishment of the monetary union and the introduction of a common currency - the European Union (EU) has been increasingly criticised for its perceived democratic deficit. Part of this deficit allegedly lies in the institutional architecture of the EU - such as the general remoteness and opacity of EU institutions, the lack of accountability of the EU Commission or the weakness of the European Parliament (for a summary, see e.g. Follesdal and Hix 2006) - but many scholars blame the absence of a European public sphere as the main cause. This public sphere deficit materialises in the discrepancy between the continuing transfer of decision-making power from the national to the EU-level and the ongoing predominance of the nation-state as the primary locus of public debate, opinion formation and citizen participation (e.g., Gerhards 1993; 2000). Yet, the emergence of a European public sphere is critical for the democratisation of the EU: on the one hand, it allows European citizens to inform themselves about EU institutions and policies and to hold them accountable and, on the other, it allows EU actors and institutions to observe public opinion and to gain public resonance, support and legitimacy. Such an interface between European citizens and political elites is all the more important in the context of difficult national ratifications of EU treaty reforms and several no-votes in recent national referendums.
Against this background, a vivid academic debate has developed over the normative standards of a European public sphere and the conditions of its emergence. The earlier literature is characterised by much disagreement on how to conceptualise and measure a European public sphere, and this has led to diverging conclusions about the public sphere deficit in Europe (Risse 2002). Based on a public communication approach, many researchers have more recently come to conceive of y a European public sphere as Europeanisation of national public spheres through communicative flows that transcend the boundaries of the nation-state. While it is relatively undisputed that these communicative flows should increasingly reach up vertically to the EU level in the form of growing visibility of EU institutions and policies in the national media, it is still contested whether and in which form they should also reach horizontally across to other European countries. In addition, other disagreements relate to the normative importance and empirical operationalisation of more qualitative discursive aspects, such as shared meaning structures or the inclusiveness of Europeanised public communication. ^	Short of any generally accepted normative standards and requirements, this
article takes a more pragmatic and empirical approach towards the study of a European public sphere. However, the aim is not to offer yet another empirical analysis of the current level or form of Europeanisation of national public spheres, but to draw on the Swiss case to provide useful comparative insights for the assessment of what a European public sphere might realistically look like (for a similar, but
0
c >
historic approach without empirical testing, see Ernst 1998; Neidhardt et al. 2000). My argument is that we should not expect anything more from a European public ^ sphere than what we can find in the nation-state context of one of the oldest and
most stable European democracies, which can in many ways be characterised as "pocket-size Europe" (Kriesi 1992, 576). Therefore, my goal is to determine whether the Swiss public sphere meets the main normative standards that scholars have set for a meaningful European public sphere: Do we find interconnectedness and discursive exchanges between the Swiss language regions? Is there a shared system of meaning? And how inclusive is the Swiss public sphere?
This article proceeds in four steps. First, I review the most influential theoretical and empirical literature on the European public sphere deficit and identify the main indicators used to measure the degree and form of Europeanised public discourse. Next, I introduce my data and measurements, before I present my empirical findings on the degree of openness and interconnectedness, the convergence of meaning structures and the level of inclusiveness of public communication in the media of the Swiss language regions. Finally, the conclusion discusses implications of my findings for the European public sphere deficit and the prospects for democratisation of the EU.
In Search of a European Public Sphere
Research on the European public sphere has flourished since the early 1990s, when the difficult ratification process of the Maastricht Treaty signalled the end of the era of "permissive consensus" and opened up a new period of growing public contestation over Europe. In one of the earliest and most cited articles in the field, Gerhards (1993, 100) has sketched two models for a European public sphere: on the one hand, a unified, pan-European public sphere carried by European-wide media and, on the other, a Europeanisation of the various national public spheres in the form of increasing national media coverage of EU themes and actors from a non-national perspective. Earlier studies disqualify this second model as insufficient and unable to alleviate the democratic deficit of the EU. In this view, national news media are "directed at national publics and remain attached to national viewpoints and communication habits" (Grimm 1995, 295). Therefore, they are likely to "domesticate" European topics rather than to "reorientate an audience towards a common European perspective" (Schlesinger 1995, 25-6). Thus, while these authors claim that a nation-transcending communicative context can only be created by the emergence of a European-wide media system, they reject this scenario as unfeasible due to the absence of a common language and shared structures of perception and understanding (see also Kielmansegg 1996, 27-8).
More recently, this view has been criticised as being deficient because it relies on an idealised conception of a homogeneous national public sphere and, most importantly, mistakenly equates the public sphere with the media system (e.g., Kantner 2003; van de Steeg 2002, 2006). Newer studies acknowledge the crucial importance of the media, but argue that the media only constitute a forum for the representation of the public sphere, not the public sphere itself.1 In this view, the public sphere is defined as a system of communication (e.g., Neidhardt 1994) and whether or not it is a European public sphere does not depend on the geographical boundaries of the media system, but on the spatial reach and characteristics of public communication in the national media. As a consequence, the recent literature has become more empirically-oriented and has focused on media coverage to establish the degree of Europeanisation of public communication over time and/or across
countries. Many of these studies have narrowed down their analysis on public debates about specific European issues at given points in time - cases in point are, for instance, the Haider debate (Berkel 2006; van de Steeg 2006), EU Eastern enlargement in general (Adam 2007; van de Steeg 2002) and Turkish EU accession in particular (Wimmel 2004), the EU constitution (Adam 2007), EU summits (Meyer 2010), the EU Commission's corruption scandal (Trenz 2000), or the BerlusconiSchulz case (Downey and Koenig 2006) - but two recent large-scale collaborative research projects have offered a cross-sectional and longitudinal examination of the overall patterns of public communication in the national media (Wessler et al. 2008; Koopmans and Statham 2010a).2
Despite this impressive accumulation of empirical evidence over the last years, there remains disagreement on how to measure Europeanised communication in the various national public spheres. However, it is possible to subsume existing indicators into three dimensions: first, from a more quantitative perspective, Europea-nization has something to do with the degree of openness and interconnectedness of public communication in national public spheres; second, it refers to a shared system of meaning; third, it deals with the inclusiveness of public debates.
Regarding the first dimension, the degree of openness and interconnectedness of national public spheres, many scholars agree that Europeanisation refers to a process that increasingly enlarges the scope of public communication beyond the boundaries of the nation-state in a vertical and horizontal direction. The vertical dimension was already present in Gerhards' (1993) early conception of Europeanised national public spheres and implies increasing visibility of EU institutions and policies in the national media. Visibility of EU-level politics allows citizens to become aware of Europe, to scrutinise EU decision-making and to form an opinion. In that sense, it is often seen a precondition for anything that could meaningfully resemble a European public sphere and contribute to the democratisation of the EU (e.g., Trenz 2004). Yet, it is not the only possible form of Europeanisation and, arguably, not a sufficient one. Given the strong intergovernmental elements within the EU, national y actors and policies of other EU member countries become increasingly relevant for one's own country. As a consequence, on the horizontal dimension, Europeanisation means that public debates in the national media should gradually open-up to other EU countries and become more entwined or interconnected. According to Eder and collaborators (Eder et al. 2000; Eder and Kantner 2002; Trenz 2004), connectivity of communication can be achieved when the same political issues are discussed at the same time and under a common frame of relevance. Critics have argued that such a synchronisation of public debates does not qualify as Europeanisation. In their view, synchronised public debates appear to be purely national debates from the perspective of the individual citizen if there are no cross-references to other countries. Hence, public communication in the different national public spheres should not only be parallelised, but also interconnected through "discursive interaction" (Risse 2002; van de Steeg 2002, 2006; Wimmel 2004; Siffi et al. 2007) or "communicative linkages" between actors from different countries (Koopmans and Erbe 2004; Koopmans and Statham 2010b). In addition, some treat the appearance
c >
of actors from one country in the national media of another country, without any explicit communicative linkages, as "weak variant" of horizontal Europeanisation (jq (Koopmans and Erbe 2004; Koopmans and Statham 2010b). In this view, the vis-
ibility of foreign actors is an indicator of the openness of national public spheres towards one another, but it only counts as Europeanisation if it increases relative CO to international (i.e., non-European) news coverage.3	^
Although most empirical studies find that EU institutions and topics get rather low attention in the national media, they point to significant increases over time in all countries (for a review, see Meyer 2010, 34-5), especially in those policy fields where the EU has gained strong supranational competences (Koopmans et al. 2010). Horizontal Europeanisation in the form of discursive interaction, in contrast, has overall stagnated at low levels (Siffi et al. 2007; Koopmans et al. 2010) - even though it can occasionally reach high levels, such as in the case of public debates about Turkey's EU accession (Wimmel 2004). Thus, while EU actors and policies play an increasing role in the national media, public debates in different national public spheres are rather disconnected from each other and have so far not integrated into a common, European discourse. Some interpret this situation as "segmented Europeanisation" and evidence for the persistence of the public sphere deficit in Europe (Siffi et al. 2007). This pessimist interpretation grounds on the assumption that discursive integration is an "integrated" form of Europeanisation that would lead to "collective identification" and a sense of belonging to the same community. From this perspective, therefore, horizontal Europeanisation is a "crucial prerequisite for the development of a common European opinion formation" (Siffi et al. 2007, 131), whereas vertical Europeanisation is a weaker variant of Europeanisation and merely generates parallel universes of EU-focused public communication ("EU-isation")(for similar views, see van de Steeg 2002; Wimmel 2004). Other researchers interpret their broadly similar findings in a more positive light and argue that horizontal exchanges are not necessarily a stronger variant of Europeanisation. Quite to the contrary, in those fields where the EU has supranational features, more vertical forms of Europeanisation are needed to alleviate the public sphere deficit (Statham 2010, 287).
Similar contrasting perspectives also persist with respect to other defining features of a Europeanised public sphere, which relate to more qualitative aspects of public communication in the national media and have been less explored empirically. As mentioned above, the second dimension of Europeanisation pertains to the existence of shared meaning structures. When Gerhards (1993) first outlined his model of Europeanised national public spheres, he mentioned two defining criteria: a growing visibility of EU actors and topics (vertical Europeanisation) on the one hand, and the evaluation of these themes and actors from a European perspective that extends beyond the interests of a particular country, on the other. This perspective has been criticised as unnecessarily restrictive. In fact, even within the nationstate context, much communication from special interest groups is not orientated towards a common, national good, but is still considered part of a national public sphere (Eder et al. 2000; Koopmans and Statham 2010b, 36). What matters instead, according to Eder and collaborators (2000), is that the same (European) topics are discussed under a "common frame of relevance." In other words, Europeans should agree on the relevance or importance of any given topic and therefore have a shared understanding of issue priorities. For Risse (2002) and van de Steeg (2002, 2006), a shared system of meaning refers to a common definition or interpretation of a specific issue rather than to its perceived importance, and can empirically be captured
through frame analysis (see also Downey and Koenig 2006). Frames also refer to identity constructions and provide answers to the question what Europe stands for (e.g., a community of values, an economic space, a political union, etc.) (Risse 2002, 8). Empirical evidence on this dimension is scarce, especially as compared to the large number of studies on vertical and horizontal Europeanisation, and highly contradictory. Whereas some found that the framing of the EU in national public spheres is broadly similar across countries, both cross-sectionally (Diez Medrano and Gray 2010) and with respect to the interpretation of the "Haider case" (van de Steeg 2006), others highlighted the continuing predominance of distinct national patterns of interpretation and the absence of distinctly European framings in national public spheres (Trenz 2000; Downey and Koenig 2006).
A third, more qualitative dimension against which the public sphere deficit has been evaluated in the literature deals with the degree of inclusiveness of public debates in the national media. This dimension has been conceptualised in two fundamentally different ways. On the one hand, inclusiveness can refer to "the inclusion of the other in the demarcation of the polity" (van de Steeg, 2002, 511). This aspect relates to the extent to which fellow Europeans are accepted as legitimate speakers in the public sphere and treated as part of the same community, indicating that there is some degree of collective identification (Risse 2000, 8). Similarly, Siffi and coauthors (2007) refer to "collective identification" as a second, qualitative aspect of horizontal Europeanisation. In their view, communicative exchanges beyond national borders should be "acknowledged by its participants "subjectively" as a common discourse" (2007, 131), as revealed by references to a common European public ("we Europeans").5 On the other hand, inclusiveness can relate to the type of actors who have a voice in Europeanised communication in national public spheres (Koopmans 2007) or who can act as agenda-setters or initiators of Europeanised news stories (Trenz 2004).
Conceptualised in either way, inclusiveness is a missing element in European-ised public communication and lies at the heart of the public sphere dimension of y Europe's democratic deficit. In fact, identification with a common European public is virtually non-existent (Siffi et al. 2007) and European public communication in the national media is dominated by powerful government and executive actors, who are systematically overrepresented in Europeanised as compared to purely national public debates (Koopmans 2007) and act as the most powerful agendasetters (Trenz 2004).
Overall, the empirical evidence on the existence and extent of a public sphere deficit in Europe is mixed, depending on the applied criteria and normative standards. Although a Europeanisation of public spheres is under way in the form of increasing visibility of EU institutions and issues in the national media, this process has hardly satisfied the more demanding qualitative requirements for Europeanised public communication, in terms of interconnectedness, framing and inclusiveness. The question is whether these standards can realistically be met in the foreseeable future. In fact, there has been a tendency in the literature to set the normative barrier for finding adequate Europeanisation very high - and often higher than for
c >
——^---——--------
national public spheres (for this criticism, see, e.g., Eder and Kantner 2002). The aim of this contribution is not to decide what would be normatively desirable for (jq a European public sphere, but to draw on the nation-state context of multilingual
Switzerland in order to get empirical insights into what the European public sphere might realistically look like. Although Switzerland has been characterised LO as "pocket-size Europe" (Kriesi 1992, 576) due to the presence of four national ^ languages and a linguistically segmented media system, a strong denominational cleavage, important divisions between urban and rural areas, extensive cantonal autonomy or the collegial executive body, the Swiss case is far less complex than the European multi-level polity. For the purpose of this paper, this is not a disadvantage, however. It can rather be argued that if the normative standards set for a European public sphere are not even met in a similar, but less complex context of one of the oldest and most stable democracies, then we should maybe reconsider existing conceptualisations and requirements for a public sphere at the European level. Before I turn to an empirical assessment of the Swiss case, I present my data and measurements in the next section.
Data Gathering
In line with the dominant approach in the European literature, I focus on the print media as the main forum for the representation of the public sphere. Newspapers have been the primary data source in virtually all recent empirical studies on the European public sphere. On the one hand, newspapers are readily available and can easily be retrieved and coded over a long period of time. On the other hand, and most importantly, newspapers have a broader thematic scope, offer more space, are less event-orientated and allow a greater discursive elaboration and argumentation than television or radio (Jarren and Donges 2002, 195). In Switzerland, newspapers are of "paramount importance" in the media system (Marcinkowski 2006, 398), not least because they are the main source of (political) information for most citizens, especially in the run-up to federal votes (Tresch 2008, 119). Given their crucial role as agenda setters and opinion leaders within the print media market, I concentrate on the leading quality paper from each German- and French-speaking Switzerland - the Neue Zürcher Zeitung (NZZ) and Le Temps (LT).
Thematically, I mainly focus on European integration policy during the period between February 2000 and March 2001. At the time, European integration policy was one of the most salient issues in Switzerland and figured on top of the political agenda, not least because the Swiss were called to the polls twice. In May 2000, they had to vote on a set of bilateral agreements with the EU for a reciprocal opening of the markets in seven specific areas and in March 2001, they had to decide on the popular initiative "Yes to Europe" asking for immediate membership negotiations with the EU. In-between the two popular votes, in addition, the federal parliament debated on "Yes to Europe" during its summer and autumn sessions in June and September 2000. Whereas the political elites in both language regions were largely consensual during the parliamentary debate and the subsequent voting campaign on the bilateral agreements, centre-right parties were deeply divided along the language borders in case of the popular initiative "Yes to Europe" (Tresch 2008, 104-9). Both votes gave rise to very intense campaigns and a higher-than-average electoral participation. At the ballot box, the bilateral agreements were finally approved by a large majority of 62.7 percent of the voters, whereas the popular initiative "Yes to Europe" was massively rejected by 76.3 percent of the voters and all Swiss cantons.5 European integration policy is a well-suited issue to confront the
normative standards for a European public sphere with Swiss reality: it repeatedly opens deep cleavages between the language regions as well as between urban and rural areas and therefore comes close to the situation within the EU where member states often have contrasting positions and interests.
The data gathering process followed a two-step procedure. First, all news articles dealing with Swiss European integration policy and published in the national news sections between February 2, 2000 and March 17, 2001 were retrieved (full sample). Second, the selected articles were coded by means of "political claims analysis" (PCA) (Koopmans and Statham 1999). PCA allows for the identification of political opinions expressed by political actors in the media - regardless of the form this expression takes (verbal statement, demonstration, political decision, etc.) and regardless of the nature or the scope of the actor (supranational/national/regional/local government, parliamentarian, political party, interest group, etc.). Ideal-typical claims can be broken down into seven elements - the location of the claim in time and space (where / when), the claimant (who), the form (how), the addressee (at whom), the substantive position on an issue (what), the actor concerned (for/against whom) and the justification (why) - but many claims are less differentiated and miss one or several elements (Koopmans and Statham 2010b, 54-7). In the case at hand, 491 claims were coded in the NZZ and 594 in LT, but given the fragmentary structure of many claims, the number of cases included in different analyses can vary depending on which claim element is studied.
According to the literature, the public sphere deficit in the EU does not primarily result from insufficient visibility of EU-level politics. It should rather be seen in terms of lacking openness towards and interconnectedness with other European countries, divided meaning structures, and limited inclusiveness of civil society actors. Therefore, these three normative requirements will be applied to the Swiss case. The degree of horizontal openness and interconnectedness is measured in two ways: first, I look at the geographical scope of claimants and, for individual actors from Switzerland, their regional origin. In this way, I assess the degree of openness y of mass-mediated communication on European integration policy towards foreign, national and regional actors as well as towards actors from the different language regions. This operationalisation comes close to the "weak variant" of horizontal Europeanisation (Koopmans and Erbe 2004), which refers to the appearance of actors from a given country in the national media of another country. To put the results into perspective, I additionally rely on PCA of public communication on immigration, pensions and education in the NZZ and LT during the years 2000-2002. Second, I examine the degree of interconnectedness by analysing the presence of addressees. The proportion of claims with a (positively or negatively evaluated) addressee gives an impression about the extent of discursive exchanges in public communication on European integration in general and the share of discursive exchanges between a claimant and an addressee from different language regions informs about the degree of discursive interconnectedness between language regions. The existence of shared meaning structures is analysed based on the justification of the claim (see Diez Medrano and Gray 2010). Justifications were coded with an open-ended list and then grouped into broad, general issue frames (see, Tresch 2008). To assess the degree of inclusiveness of public communication on Swiss ^ European integration policy, I follow Koopmans (2007) and examine which types
of actors appear as claimants in the media. The next section presents empirical findings on these three dimensions.
Empirical Results
Whereas vertical Europeanisation refers to increasing visibility of EU actors and policies, a weak form of horizontal Europeanisation relates to increasing visibility of actors from other EU member countries, indicating how open various national public spheres are towards one another. In a similar way, Table 1 shows the geographical scope of actors appearing as speakers in public communication on European integration policy in Switzerland as compared to debates on immigration, pensions, and education.
Table 1: Geographical Scope of Speakers (in percent)
	Europe		Immigration		Pensions		Education	
	NZZ	LT	NZZ	LT	NZZ	LT	NZZ	LT
EU / Foreign	10.4	11.8	3.0	5.3	1.5	0.0	0.0	0.0
National	83.1	73.4	86.6	92.1	97.0	96.7	79.7	76.7
Regional	6.5	14.8	10.4	2.6	1.5	3.3	20.3	23.3
Total	100.0	100.0	100.0	100.0	100.0	100.0	100.0	100.0
N	491	594	67	38	67	60	79	43
Note: NZZ=Neue Zürcher Zeitung, LT=Le Temps
In general, public debates in the Swiss quality press are quite closed and nationally-oriented. Even in the field of European integration, foreign and EU-level actors are quite invisible and clearly dominated by Swiss actors from the national level, which account for at least three quarters of all claims in both papers and all policy domains. Regional actors (from the cantonal or local level) get only a significant share in public debates on education, a policy field that falls mainly in the responsibility of Swiss cantons. But even in this policy field, public debates are strongly dominated by national-level actors and institutions. This result confirms the idea that national politics serve as a common focal point, able to integrate regional public spheres in a vertical way (Kriesi 1992). Going one step further, Table 2 focuses on Swiss actors and, wherever possible, looks into their regional origins.6
Table 2: Regional Origin of Swiss Spokespersons (in percent)
	Europe		Immigration		Pensions		Education	
	NZZ	LT	NZZ	LT	NZZ	LT	NZZ	LT
German-sp.	67.5	44.0	60.9	50.0	81.8	44.4	69.8	20.8
French-sp.	25.8	51.0	39.1	50.0	18.2	44.4	27.9	79.2
Italian-sp.	6.7	5.0	0.0	0.0	0.0	11.2	2.3	0.0
Total	100.0	100.0	100.0	100.0	100.0	100.0	100.0	100.0
N	225	302	23	6	22	9	43	24
Note: NZZ=Neue Zürcher Zeitung, LT=Le Temps
It clearly appears that named actors from a newspaper's language region are dominant: in the NZZ, German-speaking actors appear more than twice as often
as speakers than French- or Italian-speaking actors, whereas French-speaking actors have the most prominent position in LT. Admittedly, the regional composition of claimants is much more balanced in LT than in the NZZ. Except for public debates on education, French-speaking actors are only slightly more visible in LT than German-speaking actors. In terms of horizontal openness towards the other language regions, LT can therefore be characterised as more open than the NZZ. Yet, it could be argued that LT is not sufficiently open towards German-speaking actors who represent a clear majority in the country. In fact, given that less than a quarter of the Swiss population is French-speaking and that only about 23 percent of all seats in the national parliament are occupied by French-speakers, they seem to enjoy a disproportionately high visibility in public debates reported by LT. This relative overrepresentation of French-speaking actors can be understood from the perspective of news value research (e.g., Galtung and Ruge 1965): the "cultural proximity" between French-speaking actors and LT contributes to their newswor-thiness and increases their chance to get a voice in this newspaper.
Table 3 analyses the degree of interconnectedness in the form of discursive exchanges between a claimant and an addressee.
Table 3: Discursive Exchanges in Mass-mediated Communication about Swiss-EU Relations
u
S
i
0
£ *
I *
(R 0 C
(6
00
	NZZ		LT	
	%	N	%	N
Proportion of claims with discursive exchanges	36.2	178	43.8	260
- critical exchanges	65.2	116	72.3	188
- supportive exchanges	34.8	62	27.7	72
- addressed at national institutional actors	73.0	130	74.2	193
- intra-region exchanges	4.5	8	3.1	8
- inter-region exchanges	0.6	1	1.2	3
Total number of claims		491		594
Note: The denominator for each percentage is the number of claims with discursive exchanges; NZZ=Neue Zürcher Zeitung, LT=Le Temps
First, only a minority of all claims on European integration policy in Switzerland contain any discursive elements at all (about a third in the NZZ and 44 percent in LT). Second, even in a consensus democracy like Switzerland, discursive exchanges are for the most part negative in tone; claimants mostly address other actors to express their criticism, not their support. This finding substantiates the theoretical expectations of Gerhards and Neidhardt (1991, 66) and underlines that the public sphere mostly is a "critical public sphere" (Neidhardt et al. 2004, 27). Third, and most importantly for this paper, discursive references are almost always directed at national institutional actors (more than 70 percent), especially at (a member of) the national government. Discursive exchanges between actors from the same language region are very exceptional, discursive interaction between actors from different language regions virtually inexistent. Overall, thus, public communication on European integration policy appears to be a series of monologues rather than a dialogue (see also Neidhardt 1994, 20). Interconnectedness is the exception rather than the rule, and goes mostly in a vertical, not a horizontal direction.
en
KO
Table 4: Framing of Public Communication on Swiss-EU Relations (in percent)
	Bilateral agreements		"Yes to Europe!"	
	NZZ	LT	NZZ	LT
In favour				
Codetermination	22.0	13.9	9.6	11.1
Economic advantages	26.9	30.6	3.5	6.6
Generally pro-EU	2.9	9.0	3.5	13.2
Good moment, time is ripe	n.a.	n.a.	6.1	8.3
Tactical "Yes"	5.0	2.1	6.1	6.6
Legal / procedural reasons	12.1	12.5	1.7	2.8
General	12.8	7.6	3.5	3.9
Against				
Self-determination	2.1	2.1	5.3	7.2
Economic drawbacks	11.3	11.8	0.0	5.0
Generally anti-EU	2.8	4.9	11.4	5.0
Bad moment	n.a.	n.a.	21.9	14.3
BA have priority	n.a.	n.a.	18.4	7.7
Tactical "No"	0.0	0.7	1.8	6.1
Legal / procedural reasons	n.a.	n.a.	7.0	1.7
General	2.1	4.9	0.0	0.6
Total	100.0	100.0	100.0	100.0
N	141	144	114	181
Note: n.a. = not applicable; NZZ=Neue Zürcher Zeitung, LT=Le Temps; BA=bilateral agreements
In both newspapers, the bilateral agreements are framed in a highly similar way and citizens from both language regions were exposed to a common discourse. Most often, the agreements with the EU were interpreted (by supporters and opponents alike) from an economic perspective, underlining the advantages and costs of market liberalisation with the EU for the Swiss economy. With regard to "Yes to Europe," in contrast, some notable differences in issue framing appear between the two newspapers. In the NZZ, the three most visible frames are directed against the popular initiative, but only one of them ("bad moment") is also prominent in LT (although at a lower level). Conversely, the second-most important frame in LT ("generally pro-European") is hardly ever used in the NZZ. The reason for these differences is that political actors were divided on "Yes to Europe" along linguistic lines: French-speakers were much more supportive to the popular initiative than German-speaking actors. Given that political actors get more media attention in their home region (see Table 2), dominant issue frames can diverge. However, even French- and German-speakers campaigning on the same side tended to use differ-
Thus, with regard to their degree of horizontal openness and interconnected-ness, the regional public spheres in Switzerland appear to be as much disconnected from one another as the various national public spheres in the EU. Table 4 below shows to what extent the framing of public communication in the run-up to the votes on the bilateral agreements and the popular initiative "Yes to Europe" point to the existence of shared meaning structures.
ent frames (results not shown); for instance, French-speaking opponents framed their rejection of the popular initiative much more often in tactic terms whereas German-speaking opponents more frequently justified their position with generally anti-EU arguments.
Last, Table 5 investigates the degree of inclusiveness of public communication in Switzerland and shows which types of actors appear as claimants in different phases of the policy cycle.
Table 5: Types of Speakers in Public Communication on Swiss-EU Relations according to Policy Phase (in percent)
	NZZ			LT		
	Parlam. Phase	Voting campaign	Routine politics	Parlam. Phase	Voting campaign	Routine politics
State actors	95.4	41.0	60.8	72.0	41.9	57.9
Executive	14.8	19.5	29.2	15.9	20.2	33.1
Administration	0.0	1.2	2.3	0.0	5.1	2.2
Legislative	80.6	20.3	29.2	56.1	16.6	22.5
Intermediary actors	4.6	48.2	392	17.1	53.3	41.0
Political parties	4.6	21.5	21.5	9.8	17.2	11.8
Economic interest groups	0.0	15.9	8.5	0.0	12.3	8.4
Other civil society actors	0.0	10.8	9.2	7.3	23.8	20.8
Media	0.0	10.8	0.0	11.0	4.8	1.1
Total	100.0	100.0	100.0	100.0	100.0	100.0
N	108	251	130	82	332	178
Note: NZZ=Neue Zürcher Zeitung, LT=Le Temps
A
Although state actors clearly dominate public communication on European integration policy in Switzerland, this dominance is less pronounced than in other ^ European countries (see Koopmans 2007). At least part of the explanation lies in the Swiss system of direct democracy, which is not only a mechanism of vertical integration between the language regions (Kriesi 1992), but also reinforces the position of non-state actors in the mass-mediated public sphere (Hoglinger 2008). Overall, intermediary actors make about a third (NZZ) respectively 44 percent (LT) of all claims on Swiss-EU relations. In both newspapers, this proportion drastically Q declines during parliamentary sessions, but significantly increases during voting ^ campaigns, when intermediary actors have an even higher share of claims-making than state actors. Thus, direct democracy is a clear opportunity structure for intermediary actors and may help alleviate the public sphere deficit in terms of inclusiveness - at least occasionally for the duration of a voting campaign.
Discussion
c >
II
o
Since the beginning of the 1990s, the EU has been increasingly criticised for its democratic deficit, which is intrinsically linked to the absence of a public sphere
at the European level. Whereas scholars consider the emergence of such a public sphere as a necessary requirement for the democratisation of the EU, they tend to set the normative barrier for finding adequate Europeanisation very high.
Against this background, the aim of this contribution was to draw on the nation-state context of multilingual Switzerland in order to get empirical insights into what a European public sphere might realistically look like. Based on a content analysis of the leading quality paper from German- and French-speaking Switzerland by means of political claims analysis, this paper tested to what extent three of the most often cited criteria for a European public sphere - horizontal openness and interconnectedness, shared meaning structures, and inclusiveness - are met in the Swiss context. First, the openness of public debates towards actors from other language regions is relatively limited and discursive exchanges virtually never reach across the language borders, but only go up to national decision-makers. Horizontal integration, in other words, is largely missing - within Switzerland as much as between European countries. In light of the relative absence of discursive exchanges in the Swiss case, it seems questionable whether such linkages will eventually emerge on a European level. While some scholars portray them as a superior form of Europeanisation and some sort of final stage towards which the EU should gradually evolve (Siffi et al. 2007), others expect an inversed trend and suggest that increased supranationalisation of policy-making in the EU "transforms the communicative structure from horizontal, transnational network structure into a hierarchical, vertical structure, in which actors in national polities are linked indirectly through common references to European actors and policy contexts" (Koopmans et al. 2010, 94). Both perspectives suggest that national public spheres pass through different stages as the European integration process advances, but they anticipate this process to go in reversed directions. The findings presented here allow no conclusions to be drawn on these opposite perspectives. Historically, however, a "Swissification of regional public spheres" was only possible through (rejecting and supporting) references to the project of a federal state (Ernst 1998, 230) and thus the emergence of vertical communicative linkages.
Second, the frame analysis showed that citizens in the two language regions are not always exposed to the same discourse. In case of the bilateral agreements, political actors from all over the country framed this issue in terms of economic advantages and disadvantages. In case of the popular initiative "Yes to Europe," in contrast, different interpretive perspectives dominated on the two sides of the language border. On the one hand, general pro-European frames were more visible in the French-speaking region given that support for the initiative mainly came from this region. On the other hand, even within a political camp, French- and German-speaking actors tended to use different frames (i.e., tactic "no" of French-speakers vs. generally anti-EU feelings of German-speakers). Despite such different attitudes towards the European integration process, it has to be acknowledged that public discourses on Swiss-EU relations are not regionally-oriented. In the same ways as EU member countries often interpret EU politics in terms of domestic consequences (Siffi et al. 2007), Swiss actors also analyse the implications of Swiss-EU relations for the national economy, national political institutions or, sometimes, for cantonal prerogatives, but not for the language region. In this sense, public discourse in Switzerland is less segmented than at the European level.
Third, although public discourse in the Swiss quality press is dominated by state actors, it is more inclusive than Europeanised (and even nationally-confined) public communication in EU member countries. In Switzerland, direct democracy strengthens the position of intermediary actors in the mass-mediated public sphere,
especially during voting campaigns. Thus, direct democracy may help alleviate the public sphere deficit in terms of inclusiveness. In addition, direct democracy has been shown to synchronise public discourses in the language regions and to draw attention in a vertical way on common, national issues. Whether or not (and under what conditions) direct democracy could contribute to the democratisation of the EU and the emergence of a more integrated Europeanised public sphere is the subject of ongoing academic discussions, but recent experiences with no-votes on EU treaty reforms in several member countries clearly show that direct democracy can also slow down the EU integration process. In fact, depending on constitutional provisions for referendum votes (national vs. European-wide, binding vs. non-binding, required vs. optional, etc.), direct democracy might have an integrative and democratising potential or, alternatively, reinforce existing legitimacy deficits of the EU (e.g., Biaggini 2005; Hug 2005).
Overall, if one evaluates public communication on European integration policy in Switzerland based on the normative criteria applied to the European public sphere, the existence of an all-Swiss public sphere must be called into question. However, given that Switzerland is an old and stable democracy, I'd rather argue that these criteria, which implicitly seem to be derived from a deliberative public sphere model, set the barrier for finding a Europeanised public sphere unrealisti-cally high and should be reconsidered.
Notes:
1.	Encounters or assemblies are other public sphere arenas, but the mass-mediated public sphere is generally considered as the key forum for public communication and opinion formation in modern democracies because of its wider reach and greater impact (e.g., Gerhards and Neidhardt 1991).
2.	Note that many more studies look at European public debates in the media, but not from a public sphere perspective (e.g., de Vreese et al. 2001; Kevin 2003).
3.	At first sight, the "weak variant" of horizontal Europeanisation might look similar to Eder et al.'s "same time, same topic" criteria. It's not, however, because simultaneous debates can be purely national and do not necessarily imply references to other countries.
4.	Note that Sifft et al. (2007) and van de Steeg (2002) treat inclusiveness in the form of "we-references" as second aspect of discursive interaction.
5.	For official results of all referendum votes, see http://www.admin.ch/ch/d7pore/va/index.html.
6.	Of course, this was not possible for collective political actors. Disregarded are named spokespersons of national institutions, such as individual members of the Swiss government (Federal Council). In case of European integration policy, for instance, both Federal Councillors in charge of this policy field happened to be from French-speaking cantons at the time of study and this fact would have biased the results. Included, however, are named national parliamentarians who are elected in their home cantons.
£
(R 0 C
CN
References:
Adam, Silke. 2007. Symbolische Netzwerke in Europa. Der Einfluss der nationalen Ebene auf europäische
Öffentlichkeit. Deutschland und Frankreich im Vergleich. Köln, Halem. Berkel, Barbara. 2006. Konflikt als Motor europäischer Öffentlichkeit. Eine Inhaltsanalyse von
Tageszeitungen in Deutschland, Frankreich, Großbritannien und Österreich. Wiesbaden: VS Verlag für Sozialwissenschaften.
Biaggini, Giovanni. 2005. Direktdemokratische Legitimation der EU-Verfassung? In F. Cheneval
(Hrsg.), Legitimationsgrundlagen der Europäischen Union, 349-374. Münster: LIT Verlag. de Vreese, Claes H., Jochen Peter and Holli A. Semetko. 2001. Framing Politics at the Launch of the
Euro: A Cross-National Comparative Study of Frames in the News. Political Communication 18, 2, 107-122.	cn
Diez Medrano, Juan Gray and Emily Gray. 2010. Framing the European Union in National Public Spheres. In R. Koopmans and P. Statham (eds.), The Making of a European Public Sphere. Media Discourse and Political Contention, 195-219. Cambridge: Cambridge University Press.
Downey, John and Thomas Koenig. 2006. Is There a European Public Sphere? The Berlusconi-Schulz Case. European Journal of Communication 21, 2, 165-187.
Eder, Klaus und Cathleen Kantner. 2002. Interdiskursivität in der europäischen Öffentlichkeit. Berliner Debatte Initial 13, 5/6, 79-88.
Eder, Klaus, Cathleen Kantner und Hans-Jörg Trenz. 2000. Transnationale Öffentlichkeit und die Strukturierung politischer Kommunikation in Europa. Antrag auf Förderung eines Forschungsvorhabens im Rahmen des DFG-Schwerpunkts "Regieren in Europa." <http://www. mzes.uni-mannheim.de/projekte/reg_europ/komplett3.html#eder>
Ernst, Andreas. 1998. Vielsprachigkeit, Öffentlichkeit und politische Integration: schweiserische Erfahrungen und europäische Perspektiven. Swiss Political Science Review4, 4, 225-240.
Follesdal, Andreas and Simon Hix. 2006. Why There is a Democratic Deficit in the EU: A Response to Majone and Moravcsik. Journal of Common Market Studies 44, 3, 533-562.
Galtung, Johan and Marie Holmboe Ruge. 1965. The Structure of Foreign News. Journal of Peace Research 2, 1, 64-91.
Gerhards, Jürgen. 2000. Europäisierung von Ökonomie und Politik und die Trägheit der Entstehung einer europäischen Öffentlichkeit In M. Bach (Hrsg.), Die Europäisierung nationaler Gesellschaften, 277-305. Sonderheft 40 der Kölner Zeitschrift für Soziologie und Sozialpsychologie. Opladen: Westdeutscher Verlag.
Gerhards, Jürgen. 1993. Westeuropäische Integration und die Schwierigkeiten der Entstehung einer europäischen Öffentlichkeit. Zeitschrift für Soziologie 22, 2, 96-110.
Gerhards, Jürgen and Friedhelm Neidhardt. 1991. Strukturen und Funktionen moderner Öffentlichkeit: Fragestellungen und Ansätze. In S. Müller-Doohm und K. Neumann-Braun (Hrsg.), Öffentlichkeit, Kultur, Massenkommunikation, 31-89. Oldenburg: Bibliotheks- und Informationssystem der Universität Oldenburg.
Grimm, Dieter. 1995. Does Europe Need a Constitution? European Law Journal 1, 3, 282-302.
Höglinger, Dominik. 2008. Verschafft die direkte Demokratie den Benachteiligten mehr Gehör? Der Einfluss institutioneller Rahmenbedingungen auf die mediale Präsenz politischer Akteure. Swiss Political Science Review 14, 2, 207-243.
Hug, Simon. 2005. Möglichkeiten und Grenzen der direkten Demokratie in der EU. In F. Cheneval (Hrsg.), Legitimationsgrundlagen der Europäischen Union, 411-431. Münster: LIT Verlag.
Kantner, Cathleen. 2003. Öffentliche politische Kommunikation in der EU. Eine hermeneutisch-pragmatische Perspektive. In A. Klein, R. Koopmans, H.-J. Trenz, L. Klein, Ch. Lahusen und D. Rucht (Hrsg.), Bürgerschaft, Öffentlichkeit und Demokratie in Europa, 213-229. Opladen: Leske + Budrich.
Kevin, Deirdre. 2003. Europe in the Media. A Comparison of Reporting, Representation, and Rhetoric in National Media Systems in Europe. Mahwah/London: Lawrence Erlbaum Associates.
Kielmansegg, Peter G. 1996. Integration und Demokratie. In M. Jachtenfuchs und B. Kohler-Koch (Hrsg.), Europäische Integration, 9-34. Opladen: Leske + Budrich.
Koopmans, Ruud and Paul Statham. 1999. Political Claims Analysis: Integrating Protest Event and Political Discourse Approaches. Mobilisation: An International Journal 4, 2, 203-221.
Koopmans, Ruud and Jessica Erbe. 2004. Towards a European Public Sphere? Innovation: The European Journal of Social Sciences 17, 2, 97-118.
Koopmans, Ruud. 2007. Who Inhabits the European Public Sphere? Winners and Losers, Supporters and Opponents in Europeanised Political Debates. European Journal of Political Research 46, 2, 183-210.
Koopmans, Ruud, Jessica Erbe and Martin F. Meyer. 2010. The Europeanisation of Public Spheres. Comparisons across Issues, Time, and Countries. In R. Koopmans and P. Statham (eds.), The Making of a European Public Sphere. Media Discourse and Political Contention, 63-96. Cambridge: Cambridge University Press.
Koopmans, Ruud and Paul Statham, Eds. 2010a. The Making of a European Public Sphere. Media Discourse and Political Contention. Cambridge: Cambridge University Press.
Koopmans, Ruud and Paul Statham. 2010b. Theoretical Framework, Research Design, and Methods. In R. Koopmans and P. Statham (eds.), The Making of a European Public Sphere. Media Discourse and Political Contention, 34-59. Cambridge: Cambridge University Press.
Kriesi, Hanspeter. 1992. Öffentlichkeit und soziale Bewegungen in der Schweiz - ein Musterfall? In B. Schäfers (Hrsg.), Lebensverhältnisse und soziale Konflikte im neuen Europa, 576-585. Frankfurt / New York: Campus Verlag.
Marcinkowski, Frank. 2006. Mediensystem und politische Kommunikation. In U. Klöti, P. Knöpfel, H. Kriesi, W. Linder, Y. Papadopoulos und P. Sciarini (Hrsg.), Handbuch der Schweiser Politik, 393-424. Zürich: NZZ Verlag.
Meyer, Jan-Henrik. 2010. The European Public Sphere. Media and Transnational Communication in European Integration 1969-1991. Stuttgart: Franz Steiner Verlag.
Neidhardt, Friedhelm. 1994. Öffentlichkeit, öffentliche Meinung, soziale Bewegungen. In F. Neidhardt (Hrsg.), Öffentlichkeit, öffentliche Meinung, soziale Bewegungen, 7-41. Opladen: Westdeutscher Verlag.
Neidhardt, Friedhelm, Christiane Eilders und Barbara Pfetsch. 2004. Einleitung: Die "Stimme der Medien" - Pressekommentare als Gegenstand der Öffentlichkeitsforschung. In C. Eilders, F. Neidhardt and B. Pfetsch (Hrsg.), Die Stimme der Medien. Pressekommentare und politische Öffentlichkeit in der Bundesrepublik, 11-36. Wiesbaden: VS Verlag für Sozialwissenschaften.
Neidhardt, Friedhelm, Ruud Koopmans und Barbara Pfetsch. 2000. Konstitutionsbedingungen politischer Öffentlichkeit: Der Fall Europa. In H.-D. Klingemann und F. Neidhardt (Hrsg.), Zur Zukunft der Demokratie: Herausforderungen im Zeitalter der Globalisierung, 263-293. Berlin: Sigma.
Risse, Thomas. 2002. Zur Debatte um die (Nicht-)Existenz einer europäischen Öffentlichkeit. Was wir wissen, und wie es zu interpretieren ist. Berliner Debatte Initial 13, 5/6, 15-23.
Schlesinger, Philip. 1995. Europeanisation and the Media: National Identity and the Public Sphere. ARENA Working Paper 7-1995.
Sifft, Stefanie, Michael Brüggemann, Katharina Kleinen-von Königslöw, Bernhard Peters and Andreas Wimmel. 2007. Segmented Europeanisation. Exploring the Legitimacy of the European Union from a Public Discourse Perspective. Journal of Common Market Studies 45, 1, 127-155.
Statham, Paul. 2010. What Kind of Europeanised Public Politics? In R. Koopmans and P. Statham (eds.), The Making of a European Public Sphere. Media Discourse and Political Contention, 277-306. Cambridge: Cambridge University Press.
Tresch, Anke. 2008. Öffentlichkeit und Sprachenvielfalt. Medienvermittelte Kommunikation zur Europapolitik in der Deutsch- und Westschweiz. Baden-Baden: Nomos.
£	Trenz, Hans-Jörg. 2004. Media Coverage on European Governance. Exploring the European Public
Sphere in National Quality Newspapers. European Journal of Communication 19, 3, 291-319.
Trenz, Hans-Jörg. 2000. Korruption und politischer Skandal in der EU. Auf dem Weg zu einer europäischen politischen Öffentlichkeit. In M. Bach (Hrsg.), Die Europäisierung nationaler Gesellschaften, 332-359. Sonderheft 40 der Kölner Zeitschrift für Soziologie und Sozialpsychologie. Opladen: Westdeutscher Verlag.
Van de Steeg, Marianne. 2002. Rethinking the Conditions for a Public Sphere in the European Union. European Journal of Social Theory 5, 4, 499-519.
Van de Steeg, Marianne. 2006. Does a Public Sphere Exist in the European Union? An Analysis of the Content of the Debate on the Haider Case. European Journal of Political Research 45, 4, 609-634.
Wessler, Hartmut, Bernhard Peters, Michael Brüggemann, Katharina Kleinen-von Königslöw and Stefanie Sifft. 2008. Transnationalisation of Public Spheres. Basingstoke: Palgrave Macmillan.
>	Wimmel, Andreas. 2004. Transnationale Diskurse. Zur Analyse politischer Kommunikation in der
europäischen Medienöffentlichkeit. Zeitschrift für Internationale Beziehungen 11, 1, 7-25.
((B
rv