Version classiqueVersion mobile
OpenEdition Books

Arguing about justice

Axel Gosseries
Philippe Vanderborght

Multilingual democracy and public sphere: what Belgium and the EU can learn from each other

Dave Sinardet


Volgens veel auteurs kan aan het democratisch deficit van de EU enkel verholpen worden als er zich een Europese publieke sfeer ontwikkelt. Met name het bestaan van louter nationale media, sterk gericht op de nationale context, is hiervoor een obstakel. Zelden wordt echter opgemerkt dat een federaal meertalig land als België met een vergelijkbare kwestie kampt: media zijn georganiseerd op taalbasis en ook de inhoud die ze produceren draagt niet bij tot de ontwikkeling van een federale publieke sfeer. De Belgische case geeft ofwel aan dat de verwachtingen van de pleitbezorgers van een Europese publieke sfeer onredelijk zijn ofwel dat ook België kampt met democratisch deficit. We betogen dat de parallellen tussen België en de EU op het vlak van publieke sfeer deels kunnen verklaard worden door institutionele gelijkenissen, vooral op het vlak van partij- en kiessysteem. Pogingen om de ontwikkeling van publieke sferen te stimuleren op het niveau van België en de EU moeten daarom vooral focussen op institutionele hervormingen.

Texte intégral

Public sphere and transnational democracy

1In democratic theory, the existence of public spheres, a concept developed by Habermas (1962) within the context of the modern Westphalian (nation)- state, is often considered as an essential prerequisite for the democratic legitimacy of these states. As Fraser (2007: 7-11) argues, the public sphere is part of a normative political theory of democracy, according to which ’democracy requires the generation, through territorially bounded processes of public communication, conducted in the national language and relayed through the national media, of a body of national public opinion’. It should allow public information, participation and debate on national policies and enable citizens to influence and hold accountable political decision makers, turning public opinion into political power.

2The main carriers of such public spheres are modern mass media, as they are able to generate a common public debate, where the same issues can be discussed at the same time and under the same premises (Eriksen 2005: 343).

3However, in recent years, socio-economic processes and forms of governance which transcend the traditional realm of national democracy have gained ever more importance and transnational institutions have emerged. While we thus increasingly evolve towards an era of ’post-national democracy’, public spheres have largely remained national and have become increasingly inadequate to enable democratic information, participation and debate on a large number of today’s important policies. This is why it is argued that in order for these transnational political systems to become more democratically legitimate, it is necessary for transnational public spheres to develop.

4This argument has been made most extensively concerning the European Union (EU), which through the continuous process of European integration has gained an enormous influence on the lives of citizens in its member states. As Fossum and Schlessinger (2007: 2) argue: ’Inasmuch as the Union actually might serve as an exemplar for the development of post-national democracy at the supranational level, surely such a process has to be rooted in the reshaping of the EU as an overarching communicative space (or spaces) that might function as a public sphere’. Its absence is therefore often regarded as an element in the broader democratic deficit of the EU. Habermas (2001) even sees the development of a European public sphere as the only way the democratic deficit can be eliminated. Moreover, this diagnosis is also made by the EU itself. The European Commission’s ’White paper on European Communications Policy’ of 2006 notes that ’people feel remote from (…) the decision-making process and EU institutions’ and sees ’the inadequate development of a ’European public sphere’ where the European debate can unfold’ as a main reason for this.

5A European public sphere can be imagined in two ways. The first is a pan- European public sphere, carried by pan-European media, available across the entire EU territory. Some of these exist today (Arte, Euronews, European Voice, EUobserver, ...), but reach a very limited audience and are not considered as primary communication instruments by EU decision-makers. A problem for the rise of such European media is obviously the absence of a common language in the EU as English cannot (yet) be considered the lingua franca of all of its social classes and geographical areas.

6A second way to conceive a European public sphere is through ’Europeanisation’ of national public spheres and accordingly of national media reporting. This is generally operationalised as treatment in national media of the same EU-actors and topics, within a similar framework. However, large scale content analysis shows that only very small proportions of news reporting concern EU-actors and topics (Machill & al., 2006). Moreover, EU-news that does reach national audiences is often nationalised, as journalists select topics or angles that relate to domestic political debates (Grundmann 1999: 136-137).

Public sphere and multilingual democracy

7Although this European public sphere debate is blossoming, it is seldom noticed that it starts from a problematic assumption that national public spheres, and consequently national media systems carrying them, are uncomplicatedly present in any given national context. In multilingual federal polities, this is not necessarily the case. As they are composed of different language regions, they might not so much resemble our classic image of a nation state, but rather face comparable challenges regarding public participation and public opinion formation as the EU. They therefore provide an interesting point of comparison to feed the debate on transnational public spheres, which inversely can also shed a new light on democratic legitimacy of these federations.

8Van Parijs (2000) has already pointed at the similarities between the democratic challenges faced by the EU and Belgium. Let us therefore analyse the Belgian case in light of the assumptions of public sphere theory. Can we speak of a Belgian, federal public sphere, similar to how it is envisaged for the European case? Clearly not in its first operationalisation, as pan-Belgian media of importance do not exist. VRT and RTBF, the Dutch-speaking and French-speaking public broadcasting companies, have no structural ties anymore, with the exception of the common central office building in Brussels. The two main private broadcasters, VTM and RTL-TVI, are also independent from each other. They all specifically address only one of the two large language communities and can therefore rather be considered as ’community media’.

9Can we then speak of a Belgian public sphere according to the second operationalisation, which we can define as ’federalisation of community media reporting’ (analogous to ’Europeanisation of national media reporting’ for the EU)? Dutch-speaking and French-speaking media in Belgium do cover the federal level. The question is whether they feature the same federal political actors and topics and if they generate a federal public debate, with the same issues being discussed at the same time under the same premises. Research on the media’s role in the Belgian federal system (Sinardet 2007) allows us to answer this question.

Community media and federal political actors

10The Belgian federal government takes decisions on important policies affecting the entire country. Moreover, it is composed in linguistic parity (the prime minister possibly excluded), an equilibrium which is usually also achieved in the importance of portfolios. One would therefore expect the number of Dutch-speaking and French-speaking ministers featured in tvnews to roughly turn around the same percentage, approximating 50-50.

11However, a representative survey of a year’s news programs on Belgium’s four main television channels showed that on the Dutch-speaking news programs of VRT and VTM, around 80% of the federal ministers interviewed were Dutch-speaking, while on RTBF’s and RTL-TVI’s French-speaking tvnews, respectively 70 and 80% were French-speaking. The lack of language knowledge of the ministers does not explain these figures, as bilingual ministers of the ’other’ language group were also not often interviewed. Moreover, language is not an obstacle for news programs, which can subtitle or dub interventions. So while they make policy and take governmental responsibility for the entire country, the news value of federal ministers seems to depend on the language group they belong to.

12This is also true for issues that fall under these minister’s competences, which are less covered when the minister belongs to the ’other’ language group. Dutch-speaking as well as French-speaking viewers can therefore not be fully informed on the actions and policies of the federal government.

13This dynamic is even stronger in election time. In the run up to the 2007 federal elections, all electoral tv-debates were held between politicians of the same community. Policies of federal ministers belonging to the 'other' language group were often attacked by opposition parties but generally not defended as the ministers or representatives of their party were not present. For instance, in Dutch-speaking media, debates on justice or finance were conducted without the incumbent French-speaking Ministers. Inversely, the three Flemish candidates to the post of prime minister never presented their national programs on French-speaking television. The only exception was one special debate between the president of the Flemish regional government and his Walloon counterpart, broadcast simultaneously on VTM and RTL. However, rather than as a federal debate, this was framed as a confrontation between ’Flanders’ and ’Wallonia’.

Community media and federal political issues

14Even issues that divide politicians of both communities, do not generate a federal debate. This is shown by content analysis of one year’s coverage of the highly symbolical and conflictuous issue of the split of the electoral district of Brussels-Halle-Vilvoorde (BHV) by the two main weekly political debate programs of VRT and RTBF, De zevende dag and Mise au point.

15Again, debates mostly took place among politicians of the same community (more markedly on VRT where only 1 out of 14 guests was francophone, compared to 1 out of 4 for Dutch-speakers on RTBF). More surprisingly, those politicians that did appear on the screens of the 'other' community, often did not take part in the TV debates within their own community and vice versa: only 2 out of 5 French-speaking politicians on De zevende dag also appeared on Mise au point, while of the 12 Dutch-speakers on Mise au point, only 5 could be seen on De zevende dag. This is because top politicians generally leave it up to less high profile colleagues to cross the invisible borderline separating the VRT-wing from the RTBF-wing at the Brussels broadcasting building.

16This obviously influenced the content of the debates, which took place within the parameters defined by the political consensus of the 'own' community. The heart of the matter – should BHV be split or not? – was almost never debated. Rather, discussions started from a given premise, which was radically different on both programs: the necessity to split BHV – the Flemish position – was never questioned on De zevende dag. In contrast, opposition to a split – the francophone position – was never questioned on Mise au point. Critical questions from journalists always concerned politician's strategies to achieve (De zevende dag) or to block (Mise au point) a split. Therefore, viewers could not witness a debate on the heart of the issue. Factual elements were also emphasized or omitted depending on whether they fitted in the political consensus. On De zevende dag, politicians numerously refered to an 'important ruling' of the Constitutional Court, which they claimed demands a split of BHV. This incorrect statement was never questioned. The much discussed ruling was however not mentioned at all on Mise au point, although it is an important element that prescribes that some form of solution to the BHV problem is needed.

17In some cases, journalists even positioned themselves discursively in the conflict by using deictic references (such as 'we', 'us', 'our', 'theirs', etc). For example, on De zevende dag, the interviewer asked: 'And we Flemish, we will not give anything in return for what the francophones get?', while in Mise au point reluctant French-speaking politicians were asked: 'This means there is a price we could hope for? What would that price be? Come on, we are between ourselves here'. Such references incite viewers to look at the issue from the same viewpoint as the interviewer and further contribute to hinder an open federal public debate as could be expected in a genuine federal public sphere.

The need for electoral reform

18Analysing the media’s role in the Belgian federal system shows that the 'problems' facing transnational polities due to the absence of corresponding transnational public spheres can also be found in federal Belgium: while important political decision-making still takes place at the federal level, public information, debate and participation are largely organised at a community level. The question is whether this is really a problem. And if so, what can be done about it.

19If we resituate the Belgian case in the debate on transnational public spheres, we can draw two quite opposite conclusions. On the one hand, it can be argued that if the prerequisites for a genuine public sphere are not even met in a bilingual country such as Belgium, it is unreasonable to put similar demands on the EU for it the earn the quality label of democratic legitimacy. Because it is stuck in the model of the monolingual Westphalian nation-state, public sphere theory would foster unrealistic expectations. Belgium would then incite us to profoundly rethink this theory for an increasingly multilingual and post-national world. On the other hand, if we do follow the assumptions of numerous authors arguing for the development of a European public sphere as a prerequisite to eliminate the democratic deficit of the EU, we can only conclude that Belgium faces a similar democratic deficit.

20While both conclusions lack nuance, we do believe that both political systems deal with the question how to make multilingual democracy work in a democratically legitimate way. Moreover, as shown by the case of media reporting on BHV in Belgium, the current situation can also fuel political incomprehension and conflicts. Therefore it is necessary to look for ways to stimulate the development of public spheres in both cases. While the development of a genuine pan-Belgian and pan-European public sphere would be the most ideal way to solve the democratic deficit, certainly in the short term this is not very realistic and more can be expected from federalization and Europeanisation of respectively community and national public spheres.

21The question then becomes how. As main carriers of public spheres, media are seen as the instrument through which to stimulate their development. In the past, the EU has argued for and invested in the development of thematic television channels, pan-European media and European media programs. Options suggested for the Belgian case include establishing pan-Belgian media, attributing an additional media competence to the federal level which could then support exchanges and collaborations between community media, the organisation of federal political debates simultaneously held on Dutch and French-speaking channels, broadcasting subtitled versions of news programs of the ’other’ community (the latter was actually done by RTBF during the 2010 federal election campaign). While such proposals have their merit, both for the EU and Belgium they do not stand much chance of realisation in the current political and institutional context. This suggests it is precisely this context we might have to look at. Focusing (only) on media reform does not touch upon the heart of the problem, which can rather be situated within the political system. Indeed, in both Belgium and the EU, we see that media reporting is influenced by political behaviour.

22Political elites of the EU may express their frustration about the lack of Europeanisation of national media. They nevertheless contribute to it by communicating decisions primarily to media of their home country. After meetings of the European Council, 27 simultaneous press conferences are usually held in separate rooms, allowing to spin the joint decisions as 27 separate negotiation victories to the national electorate. Even the communication strategy of the European Commission is nationalised with spokespersons ’adapting’ information to the country they are briefing. This type of strategic behaviour does of course not stimulate but rather hamper construction of a European public sphere, a project nonetheless officially very dear to the Commission.

23A similar dynamic can be seen at work in Belgium, where federal ministers also tend to communicate their decisions primarily to their ’own’ media. Press conferences of the federal government often feature the prime minister, flanked by a Dutch-speaking and French-speaking minister who afterwards both cater for their own media. Politicians anticipate that journalists will want a quote of a politician of the own community, thus also reinforcing this attitude. Program makers often complain about the reluctance of federal politicians to speak to media of the ’other’ community. In the past, even perfectly bilingual prime ministers such as Dehaene or Verhofstadt were at times difficult to convince to address media of both communities. In any case, responsibility for the continued segregation within European and Belgian public opinion cannot be exclusively attributed to journalists, but is clearly also encouraged by the behavior of political elites.

24This parallel behaviour can in turn be explained by institutional similarities between the EU and Belgium. EU politicians’ communication strategies are national because their electorate also is. The EU does not have full blown European parties, nor is there electoral competition organised at the EU level, as the electoral system is organised on a national basis. Because of this, European elections are in fact the sum of 27 national elections.

25Or rather 28, given that European, but also federal elections in Belgium can be seen as simultaneous community elections for largely the same reasons: Belgium has no federal parties and most electoral districts do not cross the language border. This causes federal ministers to only be accountable to the electorate of one of the two communities and explains why they are less inclined to show interest for media of the ’other’ community. The heart of the democratic deficit lies there. It influences behaviour of politicians and journalists in their mutual relations: media need politicians that are relevant to their audience, politicians need media that are relevant to their electorate. This contributes to the lack of a genuine federal public sphere which in turn reinforces the democratic deficit.

26That is why arguments focusing solely on changes in the media system largely miss the point. If there is a key to a further development of genuine European or Belgian public spheres, it is more likely to be institutional reform of the political and electoral system itself. Here again, parallels can be drawn between the EU and Belgium, for which similar reforms have been suggested that would incite politicians to address voters in the entire polity, such as a polity-wide constituency.

27In 2011 a large majority of the Constitutional Affairs Committee of the European parliament proposed that 25 MEP’s would be elected through a European constituency, which would see the introduction of transnational lists, containing candidates from at least one-third of the members states. In Belgium, Dutch-speaking and French-speaking academics, brought together by Philippe Van Parijs, have argued for the introduction of a federal electoral district to elect part of the federal representatives, a proposal also supported by an increasing number of political parties1

28In both cases, politicians up for election in such a constituency would become relevant for voters and consequently media over the entire polity and will also be incited to address them, thus contributing to the development of a European- and Belgian-wide political and public debate.

29Of course, certainly in Europe, linguistic differences will continue to form an obstacle for the emergence of a public sphere such as conceived in the case of monolingual nation-states. But while electoral reform may not be a sufficient condition for the development of genuine public spheres, it is clearly a necessary condition take down the important barriers separating public spheres within Belgium and Europe today.



ERIKSEN, E. O. (2005), 'An Emerging European Public Sphere', European Journal of Social Theory, 8 (3): 341-363

FOSSUM, J.E. & SCHLESSINGER, Ph. (eds.) (2007), The European Union and the public sphere: a communicative space in the making?, London: Routledge

FRASER, N. (2007), 'Transnationalizing the public sphere. On the Legitimacy and Efficacy of Public Opinion in a Post-Westphalian World', Theory, Culture and Society, 24 (3): 7-30.

GRUNDMANN, R. (1999), 'The European public sphere and the deficit of democracy', in D. Smith & S. Wright (eds.), Whose Europe? The turn towards democracy, Oxford: Blackwell, 125-146

HABERMAS, J. (1962), The Structural Transformation of the Public Sphere. An inquiry into a category of Bourgeois Society, Cambridge: Polity Press

HABERMAS, J. (2001), 'Why Europe needs a Constitution', New Left Review, 11: 5-26

MACHILL, M. & al. (2006), 'Europe-Topics in Europe's Media. The Debate about the European Public Sphere: A Meta-Analysis of Media Content Analyses', European Journal of Communication, 21 (1): 57-88.

SINARDET, D. (2007), Wederzijdse mediarepresentaties van de nationale andere. Vlamingen, Franstaligen en het Belgische federale samenlevingsmodel, Antwerp: University of Antwerp (PhD thesis).

VAN PARIJS, P. (2000), 'Must Europe be Belgian? On democratic citizenship in multilingual polities', in Catriona McKinnon & Iain Hampsher-Monk (eds.), The demands of citizenship, London & New York: Continuum, 235-253.


1 See


Dave Sinardet is Professor of political science at the Free University of Brussels (VUB) and the University of Antwerp. He lectures at the Facultés Universitaires Saint-Louis. He is specialised in federalism, nationalism, political communication, multilingual democracy and Belgian politics. His PhD thesis dealt with the role of media in the Belgian federal system and the mutual representation of both language communities. He is also a columnist for De Standaard and Le Soir and an active contributor to the Belgian public debate on federalism and institutional reform.

© Presses universitaires de Louvain, 2011

Conditions d’utilisation :