Introduction
What can we learn from the failed referenda on European Union (EU) treaties? Do they symbolise the failure of ‘EU constitutionalisation’? Or do they, on the contrary, prove that constitutionalisation is not viable without further politicisation, going along with growing public deliberation about the telos of the integration process? The recent events fostered a debate (Bartolini and Hix, Reference Bartolini and Hix2006; Liebert and Trenz, Reference Liebert and Trenz2008) characterised at the one end by sceptical voices against any further ‘politicisation’ of the EU, favouring, instead, the established model of expert policy making (Majone, Reference Majone2005; Moravcsik, Reference Moravcsik2006: 221–222). At the other end, we find tenants of a ‘deliberative approach’ (Dryzek, Reference Dryzek2000; Eriksen, Reference Eriksen2005). They point to the potential for popular appropriation – and thus legitimisation – of the European project via politicisation and deliberation (Fossum and Trenz, Reference Fossum and Trenz2006).
With the notable exception of one comparative research project (Liebert, Reference Liebert2007), little empirical knowledge substantiates the different positions. This chapter therefore proposes an in-depth qualitative approach that takes up and enhances the mainly quantitative studies on mass media and their role in EU referendum campaigns (Perrineau, Reference Perrineau2005; Piar and Gerstlé, Reference Piar and Gerstlé2005; Liebert, Reference Liebert2007; Maatsch, Reference Maatsch2007). It focuses on the debate that developed during the referendum campaigns in France (2005) and Ireland (2008). Instead of privileging either actors’ utility maximising strategies or the influence of alleged cultural variables (‘national identity’) to explain the outcome of the referenda, this chapter proposes as third way to analyse the dynamics of the discursive process that preceded the outcome. Compared to existing quantitative research on EU referenda, the proposed approach allows grasping sequences of the internal dynamics and the argumentative development of debates in France and Ireland. This answers the question of how constellations of actors and issues were constituted in certain contexts.
The added value of this approach is illustrated via a puzzling empirical case that came up in two referendum campaigns, in France (2005) and Ireland (2008). Whereas two different treaties were proposed for ratification (France: Constitutional Treaty, CT; Ireland: Lisbon Treaty, LT), both contained the same articles of a ‘Charter of Fundamental Rights of the European Union’ (Charter):
Art. 1: Human dignity
§ 1: Human dignity is inviolable. It must be respected and protected.
Art. 2: Right to life
§ 1: Everyone has the right to life.
§ 2: No one shall be condemned to the death penalty, or executed.
(European Union, 2007)
By this, the Charter took up norms that were already institutionalised in a number of national constitutions and European legal documents. However, in both referendum campaigns, arguments arose that these articles would transform the respective national legislations on abortion. Whereas in the French case, it was argued that the ‘right to life’ might endanger the French liberal legislation, forcing France to tighten its rules, the Irish debate saw, quite to the contrary, a menace for the strict Irish ban on abortion.
A comparison of the two cases allows for insights into two different constellations. In both referenda, the argument on abortion comes from the No-side and spreads from a rather marginal to a more fundamental point, at once linked to questions of sovereignty and legitimacy. Whereas the Irish case turns into a constellation where a ‘sovereign people’ opposes ‘international organisations’, the French case constructs a more open constellation that addresses ‘European citizens’ and their ‘future rights’ under the proposed ‘constitution’. We can thus empirically investigate two ideal-type constellations of European constitutionalisation (Liebert and Trenz, Reference Liebert and Trenz2008) and of the ensuing different political cleavage lines (national vs. European and politically left vs. right; Cohen and Vauchez, Reference Cohen and Vauchez2008). I will start out with some theoretical and methodological reflections about (1) how to investigate this case. This will be followed by the methodological part on the operationalisation of the (2) proposed framework, before I turn to (3) the empirical analysis. (4) A final section will discuss results in the light of the introduced overarching questions.
Conceptualising a framework
From causal to constitutional logic
In critically assessing recent evolutions of social constructivism, notably in its International Relations version, several authors (Kratochwil and Ruggie, Reference Kratochwil and Ruggie1986; Guzzini, Reference Guzzini2000; Cederman and Daase, Reference Cederman and Daase2003; Rytövuori-Apunen, Reference Rytövuori-Apunen2005) propose a change of perspective. Instead of ‘occupying the middle ground’ by trying to bridge the abyss between positivist ‘science’ and post-positivist ‘reflectivism’, constructivism should ask how actors and social facts are constituted. Instead of turning complex social processes into variables of a causal relationship, we should analyse precisely the processes of their social construction. This should happen by asking how (exactly) identities are constructed, how norms evolve, and how social actors are constituted. To address these questions, we should integrate language and argument into our theoretical frameworks (Checkel, Reference Checkel2007: 60–66).
In the case at hand, different opinions on abortion and family planning are empirically present in both countries. These cleavages are not so much representing ‘national identity’, but social, political and, not least, generational differences. In both cases, France and Ireland, the actors using the abortion argument are rather marginal in the political spectrum; they do not represent ‘the nation’ or a representative sample of it. Much more, they pretend to defend a national consensus on abortion. Turning this claim into an analytical element (as ‘national identity’ or ‘national norms’) would come close to reifying the actors’ strategy to influence the campaign. This article instead, suggests analysing how these actors successfully realise this strategy. That implies the analysis of the discursive dynamics that developed during the two referendum campaigns. By understanding discourse as social practice (Titscher et al., Reference Titscher, Wodak, Meyer and Vetter2000: 147), this contribution asks how the ‘abortion argument’ could be established as a social fact that ultimately engaged all actors in the public debate. The research interest then consists of the analysis of a constitutive logic explaining how actors and social facts are constituted/constituting in a spatial, temporal, and social context (Smith, Reference Smith2000: 156). This, in turn, allows saying why certain discursive practices were possible in this context.
Discourse analysis as a theoretical framework
Turning to a research framework, I propose to investigate the two referendum campaigns via a ‘Foucaultian’ discourse analysis. ‘Taking language seriously’ as social practice (Neumann, Reference Neumann2008) means that ‘language’ is more than the mere individual act of speaking or writing. It evokes different meanings for different readers and is available as a resource in social constellations where it is used to foster understanding. Language is not a neutral vehicle, but constitutes and confirms meaning in an inter-subjective social process. Accordingly, it develops a power of its own, and accepting a discursive act (as ‘true’ or ‘valid’ or ‘legitimate’) is a two-way process. A speaker (an ‘actor’) has to keep in mind the discursive context (most of the time implicitly). On the macro level, this ‘dialectic relationship’ constitutes and confirms social hierarchy via language, in discourse, allowing us to investigate power relationships (Checkel, Reference Checkel2007: 65).
This understanding of language and discourse is taken up by the Vienna school of ‘critical discourse analysis’ (Fairclough and Wodak, Reference Fairclough and Wodak1997; Wodak and Meyer, Reference Wodak and Meyer2005), used, for example, to study the constitution of national identity in media discourse (Wodak et al., Reference Wodak, de Cillia and Reisigl1998; Krzyzanowski and Oberhuber, Reference Krzyzanowski and Oberhuber2007) or struggles over European integration (Diez, Reference Diez2001; Morgan, Reference Morgan2005). It starts from two preliminary reflections: (1) Discourse can refer to the discursive action of actors as well as to its quality as social structure. Both are mutually constitutive and constituting; there is a dialectical relationship between actor and structure (Wendt, Reference Wendt1987; contrast with Wodak et al., Reference Wodak, de Cillia and Reisigl1998: 42). The role and place of the actors in social hierarchies are constituted in and through the existing discourse. This discourse (social structure) enables and constrains an actor, yet it does not determine an actor: it can only be effective by and through the discursive action of the respective actors (Wodak et al., Reference Wodak, de Cillia and Reisigl1998: 42). For reasons of analytical clarity, I distinguish between ‘discourse’ when referring to its quality as social structure and ‘discursive action’ when referring to actors’ discursive action to alter, foster, or transform the existing structure. Both aspects, discourse and discursive action, are mutually constitutive and have to be analysed together. (2) In doing so, several conditions that govern the discourse can be observed. Michel Foucault (Foucault, Reference Foucault1997 [1969]: 61–103; Reference Foucault1998: 39–41; Kerchner, Reference Kerchner2006: 48–49) has called them the ‘order of the discourse’. Successful communication depends on (1) the spatial and temporal context of a statement and (2) the social position of its author (Diez, Reference Diez1995: 18; Larsen, Reference Larsen1997: 19–21). Besides these two conditions that relate to the ‘social’ dimension of discourse, two further conditions relate to the dialectic character of language: (3) a successful discursive act creates resonance with the existing discourse. Resonance depends on (a) the knowledge of and reference to the existing genealogy of a discourse (Foucault, Reference Foucault1997 [1969]: 61–103; Reference Foucault1998: 39–41) and (b) on the coherency and inherent logic of a statement. Finally (4), the frequency and stability of discursive elements show their ability to structure the discourse (Foucault, Reference Foucault1998: 26).
Subsuming this ‘discursive logic’, a discourse analysis then has to analyse how and why certain discursive actors are constituted (accepted as ‘valid’ speakers) within a discourse and why their discursive action is successful (or not) (influencing the discourse). This depends on the following conditions: (1) the context of discursive action and (2) the social position of the discursive actor; (3) resonance between discursive act and discursive genealogy via coherence and (associative) logic, and (4) frequency and stability of the elements.
Methodological reflections and operationalisation
For the time being, there may just not be one authoritative ‘right’ or ‘true’ interpretation of the incriminated treaty provisions. Accordingly, we have to assess the ‘validity’ of claims made during the referendum campaigns within the respective discursive contexts.
Method
Together, the respective discursive action and the existing discourse decide how and why a certain argument becomes true/valid or not. Contrary to ‘standard’ constructivist thinking, the resonance of discursive action is not assessed with regard to externally given variables (‘national identity’, ‘fundamental norms of a society’…), but by assessing social practice in a certain situation, or, more precisely, by analysing the interaction (dialectics) between discursive action and discursive structure.
While assessing discursive acts ultimately requires an interpretative method (for the reasons stated above: language is not neutral, it can mean different things to different people), the proposed conceptualisation of discourse and discursive action allows maximum transparency. Discursive acts are analysed on the level of newspaper articles (unit of analysis). To evaluate their ability to change or confirm the discourse, they are assessed on the four established conditions, resonance (coherence and logic), social position (hierarchy) of the actors, frequency and stability (structure of the discourse), and context of the statement.
Furthermore, a sequence of three steps of analysis is established: (1) constitution of actors and discourse, (2) dynamics of the debate, and (3) (un-)successful discursive action. The claims made with respect to these elements are falsifiable: if further research explores the same or other sources and challenges my interpretation, or if the contextual factors that underlie my assessment are proven to be incomplete or wrong, my analysis is false. In making explicit the criteria upon which my interpretation relies, this analysis can be intersubjectively reproduced.
Operationalisation
Based on these preliminaries, I analysed the two referendum campaigns on the ‘CT’ 2005 (France) and on the ‘LT’ 2008 (Ireland) by drawing on the classical print media.Footnote 1 They were accessed via Lexis-Nexis, a database that allows for digital access to the newspapers’ print editions. For France, the research included five daily national newspapers (Le Monde (LM), Libération (LIB), La Croix (LC), Le Figaro (LF), and L’Humanité (LH))Footnote 2 and took into account the period from 01/09/2004 to 30/06/2005.Footnote 3 Keywords were avortement OR IVG OR droits des femmes AND traité constitutionnel OR référendum. In the Irish case, the newspapers equally comprised national ‘quality press’ (The Irish Times (TIT), The Irish Independent (TII), The Sunday Independent, The Sunday Business Post (SBP), and The Irish Examiner (IEX)).Footnote 4 The analysis stretched from 01/01/2008 to 31/07/2008. Keywords were abortion OR gay marriage AND Lisbon Treaty OR Referendum. Footnote 5
This framework can be criticised: It seems commonly accepted that in both referendum campaigns, ‘new media’ (internet, blogs, and feeds…) played an important role. The proposed framework does not allow to access these ‘new media’ directly. Furthermore, the retrieved ‘national quality press’ could have a tendency to express a ‘semi-official’ discourse. This may discriminate alternative, marginal discourses, which would constitute a selection bias.
As this contribution will show, context knowledge and background information are needed, as in any qualitative analysis, to allow for plausible interpretations. Yet by respecting this caveat and applying the introduced conditions of the ‘order of the discourse’, an analysis of the following question becomes possible: How could the arguments at the outset of a marginal discourse enter the mainstream press and develop into a discursive constellation that was open for these marginal arguments? The discursive dynamics behind this evolution are the article’s main focus. They can be analysed within the proposed framework.
Furthermore, concerning the comparability of the two cases, ‘common wisdom’ holds that the abortion argument was more ‘fundamental’ in the Irish case – yet the reason for this is not clear. Looking at the number of articles that this analysis retrieved (Ireland: 103, France: 81), the difference seems less obvious. In the eyes of 39% of all Irish voters (58% of No voters), a Yes vote would have changed Ireland’s legislation on abortion (Red C poll, 22/6/08). In the French case, while we lack comparable survey data (the ‘abortion question’ was not asked), a large-scale (4071 articles) quantitative analysis of French print media finds the Charter to be the most important of ‘substantive constitutional topics’ (Maatsch, Reference Maatsch2007: 268) treated during the campaign. Accordingly, we should systematically reconstruct the importance of the issue for the two debates within the respective discourses. The goal of this contribution is precisely to understand how the ‘abortion argument’ could become valid in two different constellations and what consequences these constructions implied.
Comparing the French and Irish referendum campaigns
To ensure comparability, the analysis will, in both cases, start out with (a) the constitution of the respective actors and their discourse, followed by (b) a focus on the dynamics of the debate, and finishing (c) with the different (un-)successful attempts to influence the debate.
The French case
Figure 1 gives an overview of the French debate with all articles that we retrieved via Lexis-Nexis.Footnote 6 Articles are represented according to their date of publication, number of words, and the respective newspaper. Numbers (1–12) represent discursive events; the following analysis will refer to them.
Sequencing the French debate: 81 Articles, number of words (51–3.001) and date of publication (19/09/2004–01/06/2005).
Constituting an actor, constituting a discourse
The first question to be answered is how the referendum campaign constituted rather marginal political groups into legitimate actors. Linked to that, I analyse the kind of debate that is constituted by these actors.
In the French case, the only forces affirmative of the ‘abortion argument’ within the retrieved 81 articles, are the leftist no-groups, especially feminist groups.Footnote 7 Their ‘institutionalised’ voice is the Parti Communiste Français (PCF) and its daily press organ LH. Both are historically influential and important actors in the French context, yet both are continuously losing influence and importance (Hudemann and Soutou, Reference Hudemann and Soutou1994: 115; Mischi and Streith, Reference Mischi and Streith2004). The No campaign constitutes an important moment for this movement to regain some of its former influence. While no case was found where the PCF and LH bluntly claimed ‘CT means abortion becomes illegal’, they use the argument ‘CT does not enhance women’s rights’ and ‘CT does endanger women’s rights’ to raise doubts on the treaty clause.
Already on 22/09/04 and in four more articles LH sets the terms of the debate. (1) Seven participants at the European Social Forum that took place in London address the European citizenry (‘À la citoyenneté européenne’). Next to this normative framing, the text starts out: ‘C’est à Rome que se tiendra la signature de la constitution, le 29 octobre’. This first article clearly frames the referendum campaign as a constitutional debate and calls upon European citizens to take part in it. The normative might of this framing of the referendum campaign is difficult to counter, especially as it was the governments in favour of the treaty that came up with the term ‘constitution’. In putting the emphasis on the ‘constitutional character’ of the proposed text and the (lack of) ‘democratic quality’ and ‘democratic and social rights’ contained in it, LH tries to frame a ‘democratic constitutional debate’. If this strategy succeeds (if it is accepted and taken up by other discursive actors), the debate will be enlarged to all ‘democratic forces’ and the range of ‘acceptable’ arguments will be broadened. LH fosters this evolution by creating resonance with ‘great’ French constitutional moments, starting with 1789 and the Declaration of Human and Citizen Rights. Further references are to the separation of state and church (1905), the social constitution established in the Programme du Conseil national de la Résistance (1944), and the legislation that allowed for abortion (1975). All these reference points are still valid parts of the French constitutional order, as well as of the country’s political self-understanding. In taking them up, LH creates resonance with France’s constitutional genealogy and – in inscribing its discursive acts into it – constitutes itself into an accepted actor in the discourse.
(2) On 12/10/04, LF is the first newspaper that refers to the leftist framing – and accepts it as the terms for a larger debate. The long article reflects on the necessary changes in the French constitution under the CT. Concerning the role of the Charter (that contains the ‘right to life’), the article refers to (anonymous) lawyers who think it will change France’s constitution.Footnote 8 Accordingly, LF confirms the framing introduced by LH. A ‘constitutional debate’ takes place, and a newspaper that clearly supports the Yes, president Chirac and his party, recognises a fundamental point (universal validity of the Charter) that the Yes-campaign resolutely contests during the whole duration of the debate. In accepting the framing and the genealogy proposed by the left, Le Figaro shows that as early as October a discourse and its actors are constituted. A ‘constitutional debate’ negotiates key themes of the French constitutional genealogy; it includes all those actors that refer to and accept the established meanings of this genealogy.
LH further enhances this argument in 19 articles between 22/09/04 and 20/03/05, when polls show a negative majority for the first time. Next to LF’s article that confirms this framing, one LM article (11/12/04) (2) fosters the same perspective. In an interview with the leader of the trotskyist Lutte Ouvrière Arlette Laguiller, the article spreads the left framing further. Laguiller fears ‘des dérives réactionnaires’, especially concerning women’s rights, abortion and divorce, and the newspaper does not question this claim. The Yes-side only starts to counter this leftist framing by the end of March 2005. LH, on the other hand, continues to broaden its legitimacy and to constitute a discourse. For this, it refers to ‘the people’, to ‘European civil society actors’ with their ‘concerns’ (instead of PCF party officials). Starting with the first, already mentioned article (22/09), the newspaper sticks to this pattern: It links ‘European citizenry’, fears for women’s rights and ‘abortion’. More ‘critical civil society’ is presented on 7/01/05 and 11/02/05, and the ‘constitutional’ framing is taken up: ‘European civil society’, representing ‘citizens’, not a party (and defending themselves against representing a ‘nationalist’ French vision) are debating a ‘constitution’ – and not a further European treaty. Accordingly, this legitimises a debate around values, democratic quality and the social and democratic rights of the text. Via these quotations, the chapter does not directly claim that the right to abortion would be endangered in France. Instead, (3) the feeling of a lack or imbalance of social, and especially women’s, rights is fostered. By turning the debate into a ‘constitutional’ debate and fostering negative feelings about the CT’s social consequences, citizens’ choice is transformed: They do not have to decide between the existing and a new European treaty, but between the existing French constitution and an incomplete European constitution. The same argument is advanced on 12/02/05, commenting on the internal debate of the French Greens (Les Verts), and on 26 /02/05, when LH presents ‘Un “non” féministe’(3).
Dynamics of the debate
With 19 articles in LH and the two articles in LF (12/10/04) and LM (12/12/04) that accept and foster the left framing, the No-campaign has, end of March, successfully constituted its campaign. It is already recognized by two mainstream newspapers (LF and LM) when the Yes-side only starts to campaign. Until the end of March, only one article (07/03/05), again in LF, postulated that in case of a narrow outcome women’s vote will be decisive (52% of the electorate) and therefore (4) ‘Un effort de pédagogie’ in their direction should be taken by the Yes-campaign. Of course, as we have seen, this ‘effort’ had so far been made only by the leftist No-campaign and no comparable discussion of single treaty provisions was initiated by the Yes-side. Instead, LH responds directly to LF’s (07/03) article – in a ‘pedagogical effort’ (11/03/05). The communist daily ‘explains’ the treaty and gives women ‘seven reasons to vote No.’ (4). Again, no bluntly ‘false’ arguments can be found; instead – working with insinuations and associations – the article diffuses a climate of fear and posits the No-campaign as the only ‘valid’ source of information, whereas the Yes-side seems to ‘hide’ something.Footnote 9 The communist daily, exploiting the lack of precise communication of the Yes-campaign, ‘explains’ contraception and abortion:
Les groupes anti-IVG qui ont fait du ‘droit à la vie’ leur slogan ne peuvent qu’être satisfaits de constater que celui-ci est devenu l’intitulé de l’article II-62 de la charte des droits fondamentaux alors même que le droit à la contraception et à l’avortement n’est reconnu nulle part. Or, ce droit des femmes à disposer librement de leur corps a été obtenu, en France en particulier, à l’issue d’une lutte âpre des femmes et des associations féministes. Dans plusieurs pays européens (Pologne, Portugal, Irlande, Chypre, Malte) ce droit n’est pas reconnu. Or, c’est sur la situation dans ces pays que le projet de traité constitutionnel s’est aligné.Footnote 10
LH is not bluntly wrong, but ignores the scope of the Charter. Instead, the emotional appeal to the battle for abortion in France is a case of associative logic that in the French context creates resonance, again with one of the landmarks of the left’s Republican genealogy. ‘Focussing on the text’ and ‘explaining article by article’, the No-side argumentatively outmanoeuvres the Yes-side. The qualitative analysis provided here allows addressing these internal dynamics and inherent logic of the discourse. By mid-March, (5) this correlates with the first negative majority in the intentions of vote (18/03/05, Piar and Gerstlé, Reference Piar and Gerstlé2005: 43).
The Yes-side now understands the danger. On 24/03/05, the government officially decides to enter the debate and President Chirac announces his personal participation in the campaign on TV (Piar and Gerstlé, Reference Piar and Gerstlé2005: 50). Yet as we have seen, the terms of the debate and the genealogy are already in place. Successful discursive action now, between the end of March and the end of May 2005, depends on the resonance with the established (left) discourse.
(Un-)successful discursive action
The first attempt (6) to critically engage with the No-side’s arguments can be found on 29/03/05 in Libération. Olivier Duhamel, member of the European Convention for the Party of European Socialists, and Jack Lang, the spokesperson of the French Socialists (the yes-side of them…) for the referendum campaign, challenge the left No-interpretation of the CT. They start by creating resonance with the Republican genealogy: ‘Défendre le non au référendum du 29 mai sur la Constitution européenne, c’est évidemment le droit de tout un chacun’.Footnote 11 The next sentence tries to disqualify the No-campaign: ‘Accumuler les mensonges sur le contenu du texte soumis à référendum, c’est totalement indigne de tout républicain’.Footnote 12 This directly challenges the legitimacy of the left No. When raising the particular point of abortion, the authors are equally affirmative: ‘2. La Constitution supprimerait le droit à l’avortement en consacrant le droit à la vie (art. II-62). Affabulation grossière’.Footnote 13 Those who fear the suppression of liberal abortion rights in France are treated as liars – and thus disqualified in the debate. According to the analytical framework, this is not enough to influence the existing discourse. After creating resonance and disqualifying the existing debate, arguments are now needed that reconstruct the discourse in order to confirm the own and disconfirm the existing reading. The authors, both experts of public law, only need one paragraph to contextualise the treaty provision within the Council of Europe’s Human Rights Convention, and to establish the priority of national legislation in the case at hand.
If we assess this discursive act with the established analytical framework, three arguments speak, nevertheless, for the low success of this article: (a) settling the CT within the framework of international judicial norms, while at the same time creating resonance with the established Republican genealogy, is not really coherent. More effort would be needed to establish this alternative understanding of the CT as international, European treaty. (b) Concerning frequency and stability, the authors’ complex argumentation could be found only three more times within the 81 analysed articlesFootnote 14 – not enough to counter the frequent contributions of the No-side concerning ‘women’s rights’. (c) Depending on one’s social position, treating those that defended an alternative vision as ‘liars’ can be counter-productive (if the own position is in a minority) and offending (Morgan, Reference Morgan2005: 33).
Focusing on discursive dynamics and discursive action allows seeing that LH again directly reacts to Liberation’s article (6). For the communist daily (29/03/05), it counts among ‘ces sorties plus insultantes les unes que les autres’,Footnote 15 it accuses the authors of ‘arrogance’. The No-side, according to LH, is using a more ‘honest’ strategy, explaining the ‘constitution’s’ content article by article. In the (discursively) open situation at the end of March, ‘arguing’ is a more successful strategy than ‘disqualifying’. Resonance to a well-known Republican genealogy on the one hand and lacking context knowledge of the electorate concerning the role of international legal norms on the other turns the No-campaign into the central actor of the campaign. The Yes-side must accept to argue with them in order to get access to the discourse.Footnote 16
Focusing on discursive dynamics allows identifying another sequence that sets the path for further development. The president’s party (UMP) is ‘muscling its arguments’ (LF 09/04), as do the socialists (their Yes-side, LF 09/04). The No-side has to react, and this time it is LH that accuses the Yes-side of spreading ‘lies’ (11/04/05). In trying to disqualify the Yes, proponents of the No equally have to adapt their arguments – leading to a situation that is open for argumentation, with both sides now agreeing to direct confrontation (7). PCF leader (secrétaire nationale) Marie-George Buffet engages minister of defence Michèle Alliot-Marie (LM 20/04/05) in a debate. Replying to the question if ‘Le droit à l’avortement peut-il être remis en cause du fait de la Constitution?’,Footnote 17 Buffet answers ‘Non, je n’emploie pas cette formule. Ce que je constate, c’est que dans la Charte des droits fondamentaux, dès le premier article est réaffirmé le droit à la vie mais dans aucun article de cette charte n’est affirmé le droit des femmes à la maîtrise de leur corps’.Footnote 18 This is the, by now well-known, argumentation but Buffet then goes on: ‘Ça aurait pu être un point d’appui pour les femmes qui, aujourd’hui, n’ont pas encore ce droit, comme en Irlande ou en Pologne, or la charte ne leur donne pas ce point d’appui puisque ça n’y figure pas’.Footnote 19 Again, no false claims are made, but a ‘regret’ that not all Europeans can profit from the ‘progressive’ French legislation. Here and in further contributionsFootnote 20 we find resonance with a very old, well-established genealogy of the French Republic. It appeals to France’s ‘universal mission’ that urges her to bring ‘progress’ to the world. The terms of this ‘progress’ are of course developed within the French context; they comprise the establishment of a ‘more social’ and ‘more humane’ society. The struggle for their realisation started with the revolution of 1789 and if one accepts this genealogy, one could in all good faith ‘save’ the other Europeans from a bad constitution by voting ‘No’. That the usually marginalised PCF leader directly engages in a debate the Minister of Defence shows how the extreme Left and their positions is accepted in this discourse. Accordingly, the PCF also changes its social position.
In the following month (April), an open debate develops (8). The Yes-side now intervenes with enhanced frequency against the No’s established position. Between 30th March and 3rd May, 21 articles supporting the Yes-sideFootnote 21 debate the abortion argument, against nine articles of the No-side.Footnote 22 The qualitative analysis notably shows the existence of a ‘feminist Yes’ (9) at the end of this period. Whereas the polls shortly saw the Yes back in majority (02/05/05), the next week already brings more nuanced polls that document the split of the electorate into two camps of comparable size (10).
To overcome this situation, both sides use the last 2 weeks of the campaign to rely on actors who can claim an overarching, influent social position due to their moral integrity (11). The Yes-side goes ahead in winning Simone Veil for their campaign.Footnote 23 Her judgment concerning the alleged danger for the French legislation – ‘Il est rare d’en arriver à des contrevérités aussi flagrantes’Footnote 24 (LIB 02/05/05) – is of a very precious nature for the Yes-side. To counter this weighty contribution, the No-side presents Gisèle HalimiFootnote 25 (LM 19/05/05): ‘Le traité constitutionnel, une menace pour les femmes’Footnote 26 (11). She, like Marie-George Buffet, now propagates a more precise argument: ‘S“il est vrai que, aujourd’hui et en l’état, rien n’interdit explicitement dans ce texte le droit de choisir ses maternités, il serait honnête de préciser que rien ne le garantit.” ’Footnote 27 By arguing the point, the Yes-side has forced the No-side to be clear on the lacking link between abortion and the CT. In order to keep their legitimacy in the debate, the No-side admits that ‘abortion’ is not an issue in the CT. Similarly to Buffet, Halimi transforms her statement into a question of ‘progress’: ‘Nul ne peut donc prétendre que cette Constitution apporte aux femmes, sur ce point fondamental, une avancée. Au contraire, à cet égard, elle ouvre la porte à un dangereux laisser-faire, laissez-passer’.Footnote 28 Again a lack of progrès compared to the existing constitution is insinuated, fostering a climate of insecurity.
A final intervention by philosopher Sylviane AgacinskiFootnote 29 in favour of the Yes-campaign does not really change the discourse; both sides and their arguments are well in place and can claim ‘credibility’ and ‘validity’ (11). One day before the referendum, LM (28/05/05) draws a conclusion of the campaign’s key debates and starts out with ‘fundamental rights’, containing ‘abortion’ (12). This text, the longest article (3008 words) I considered, opposes both argumentations, but cannot decide which one is right, leaving the reader to judge (Maatsch, Reference Maatsch2007: 270).
This analysis of the French case allows for three conclusions. (a) The first-mover advantage of the left No-side allows them to frame the debate in a twofold manner: In creating resonance with the French republican genealogy, they can constitute a ‘democratic constitutional debate’. This comprises access to the debate for all those forces (and their arguments) that are ‘struggling for democracy’ – irrespective of the CT’s paragraphs. The French extreme Left turns into a legitimate discursive actor, and their arguments enter the debate via a focus on the Charter and the ‘fundamental rights’ of the CT. The Yes-side is at pains to regain a place in the debate when it starts campaigning 3 months later. To create resonance, it has to accept the genealogy and the actors that are already in place. (b) The genealogy that is established by the No-side appeals to all the symbols that have a fundamental constitutional value within the republican discourse of the French left, from the universal Declaration of the Rights of Men and Citizens in 1789 to the struggle around the legislation allowing for abortion in the 1970s. It allows the No-side to easily adjust its argumentation once it is challenged for its ‘nationalism’ by pointing to the ‘universal mission’ of France and the ‘democratic struggle’ of the Left, elements of the republican genealogy. (c) Yet the Yes-side comes back into the debate and, once taking up the argumentative challenge of the left No-side, succeeds in pointing out the larger context of European and international norms, of which the CT is only one part. Suffering from the difficulties of communicating a complex argumentation, one successful step is to rely on accepted ‘moral authorities’ within the discourse. The Yes-side relies on Simone Veil to bring an advantage, yet the No-side is quick to counter with Gisèle Halimi. At the end, we cannot clearly figure out a ‘winner’ in the debate surrounding abortion. To create resonance with the republican genealogy, the Yes-side, however, would have to convince voters of the CT’s democratic progrès.
The Irish Case
Once again, the analysis starts by asking how the specific discursive dialectics of the Irish referendum campaign on the ratification of the Treaty of Lisbon (2008) constituted the (usually marginal) groups that supported a ‘pro-life’ No-campaign into actors in the discourse. Linked to this, how could those actors constitute a discourse that allowed them to legitimately advance their argument and gain credibility for it? A second step focuses on the discursive dynamics and the resonance with genealogies. Finally, discursive action will be analysed.
Constituting an actor, constituting a discourse
For the Irish case, 103 articles were retrieved in the different newspapers. Figure 2 gives an overview of the articles, again with number of words, date, and newspaper of publication. Key discursive events (1–12) are again referred to in the text. On 27/01/08, there is still no date for the referendum (12/6/08), yet the No-side is already in place and has started its campaign, as the Sunday Business Post reveals in an overview of different No-groups (1). The article sees a rather marginal role for the ‘pro-life’ campaign, especially as their argument seems neither very new, nor credible.Footnote 30 In the first 2 months reviewed here, the existing social hierarchy marginalises the abortion argument, yet this does not mean that it is absent from the debate.Footnote 31 The debate really starts off early in March, and as in France, the No-campaign sets the terms of the debate. On 07/03/08, the Irish Independent reveals ‘A “yes” may bring abortion changes, says MEP’ (2). The well-known Euro-sceptic, Danish MEP (Member of European Parliament) Jens-Peter Bonde, raised the subject in giving a keynote address to a ‘Forum on Europe’. He carefully weighed his words, as his letter (18/03/08) to the Irish Independent shows where he complains about having been falsely quoted:
The other day, I was quoted as saying that an Irish ‘Yes’ to the Lisbon Treaty would outlaw the Irish rules on abortion. I did NOT say that. I would never interfere in an Irish debate for or against abortion. I talked about small nations and the Lisbon Treaty and I proposed to safeguard all kinds of specialties in authoritative interpretations from the council or legally binding protocols. […] Shame on those MEPs who reacted against my very well-founded advice and simply spun it to be for or against abortion. Don’t they listen?
– What had happened? Between 07/03/08, Bonde’s interruption in the debate, and his letter to the editor, five articles dealt with his allegations (3).Footnote 32 The central focus of all these articles was the LT’s possible impact on abortion. Yet, as Bonde correctly claimed in his justification on 18/03, he had not bluntly stated (on 07/03/08) that the LT would allow abortion in Ireland. If we go back to the Irish Independent’s article of 07/03/08, it becomes clear that the resonance of Bonde’s claim with an already existing discourse brought up the abortion argument. After raising fear in the header, only the fourth sentence of the article brings Bonde’s full quotation: ‘the Danish MEP claimed the “culture of consensus” would disappear under the LT if half of the member states could outvote the other half. Sensitive policy issues such as abortion, corporation tax and direct foreign investment could be affected, he claimed’. With a look on Flash Eurobarometer (FEB) 245 (post-referendum poll), we can state that Bonde mentioned all those issues apparently ‘sensitive’ to Irish voters in two sentences. By this amalgam, he, of course, appealed to fears already present in the discourse – but that had not yet been raised within the referendum campaign of the LT. The Yes-side, equally assisting the Forum on Europe-debate, is quick to launch a counter-attack, as the Irish Independent transmits.Footnote 33
Sequencing the Irish debate: 103 Articles, number of words (112–2.863) and date of publication (27/01/2008–12/06/2008).
The follow-up of this discussion is revealing as it shows that a discourse existed before the beginning of the debate. It dates back to the previous referenda on EU treaties held in Ireland. Contrary to France, the arguments for EU treaties had already been constituted. Bonde just had to insinuate the ‘abortion topic’, and the discussion started again. The same was true for the counter-attack already lined up. Bonde’s actual argument (power structures in the EU) was not addressed by other contributors, it lacked resonance. Therefore, if the discourse was constituted, we would have to ask which actors it constituted. A first point confirms the initial theoretical considerations: A ‘legitimate’ actor is an actor that knows the genealogy of the discourse and creates resonance with it. In this case, even a ‘foreigner’ (MEP Bonde, but also the ‘German legal expert’, see footnote 30) can play a legitimate role. The arguments that Bonde introduced in the discourse can then ‘constitute’ those as actors who were at first marginal forces within the ‘semi-official’ discourse.
In the following weeks, this discourse continues to deny the relevance of the abortion argument. Yet, the subject is always present, even if actors supportive of this claim are marginal and difficult to identify via the media discourse.Footnote 34 Two main groups acted in this direction. One, a publication of conservative catholic pressure groups called ‘Alive!’, appears (in this analysis) for the first time via a letter to the editor of TIT (published on the same day as Bonde’s letter). Treating a subject in this way (‘letter to the editor’) allows ascribing its author a low social position, while still admitting that an issue becomes relevant. Accordingly, the letter complains about the allegations of Alive!: ‘Of course these statements [on abortion] are quite simply untrue and intended to mislead the readers of Alive!’. In the analysed sample, ‘Alive!’ was never presented in a more credible or serious manner. Discursive context knowledge allows participants of the debate to clearly identify the role and social position of this group within the Irish society. Only after the discourse constituted the particular issue (abortion), via the contribution of a ‘foreigner’, the groups supporting this issue are constituted (by the discourse) into accepted actors.
In this analysis of mainstream print media, the second ‘pro-life’ group ‘Cóir’ becomes visible only on 15/05/08, yet in a spectacular way (8).Footnote 35 No other actor in the Irish referendum gained the same attention for his campaign, yet the NGO is quickly characterised as marginal, ‘extreme right’, ‘right-wing’ and ‘fundamentalist’ (TIT 15/05/08). They are related to the ‘Mother and Child’ campaign of the Nice Treaty referendum, to the ‘Youth Defence’ and the ‘Pro Life Alliance’ – a well-known genealogy in the Irish discursive context. The question than is why could the abortion argument continue to influence the campaign, even if its main proponents (Alive! and Cóir) were disqualified and occupied a marginal space in the semi-official discourse. To answer this question, we have to turn to the dynamics and the resonance of the debate.Footnote 36
Dynamics of the debate
As in the French case, the No-campaign started earlier than the ‘official’ Yes-campaign. In this analysis of the Irish debate, Libertas and Jens-Peter Bonde intervened early in March, around the same time as ‘Alive!’. To appear in the analysed media, their campaigning must have already been going on for some time. Against these agitations, the Yes-side decides to start its campaign at the end of April and begins campaigning in mid May (6).Footnote 37
Concerning the framing of the argument, most of the articles deny the validity of the No-claim and the abortion argument predominantly appears through its denial: After an actor had claimed that abortion would be allowed under the LT, the newspapers, spokespersons, party chairmen etc. stand up to denounce this (in their eyes false) claim. We thus have to understand how the argument becomes possible against this discursive stability. An example can be found on 21/03/08 in the Irish Independent (4), in the aftermath of the ‘Bonde debate’. The opinion article ‘Let’s tell Europe where to get off on abortion issue’ is labelled as ‘Analysis’. The article comments on a decision of the European Court of Human Rights that ‘has told Ireland to decriminalise abortion’. It frames the subject in terms of ‘normal Irish people’ defending ‘their sovereignty’ against ‘all sort of international organisations’:
The Council of Europe is not the only international organisation attempting to force Ireland into line with the international liberal/left consensus on ‘human rights’. The United Nations is continually at it through myriad conventions which are always, but always, interpreted in a radical manner by social radicals used by the UN. More worryingly, but, alas, predictably, the European Union is at the same thing.
From this mixture of factual and emotional language, challenging the EU and the LT is easy:
Needless to say, this is hardly a good ad for the Lisbon Treaty. If the EU is already abusing the powers it has, who in their right mind would want to give it even more powers? It turns out that every major international organization of which we are part, whether it be the EU, the UN or the Council of Europe, has a social agenda that is completely out of line with the wishes of probably a majority or Irish people.
Contrary to the French case, the EU is framed as an ‘international organization’, opposed to ‘the Irish people’. The article further elaborates:
In ways that are radically and deeply undemocratic, these organizations and their various off-shoots are using every means at their disposal to force us into line with their social agenda, an agenda that includes gay adoption, abortion, the limiting of religious freedom, and the overall imposition through State power of political correctness.
This is the main framing of the No-argument as it appears throughout the analysis of the 103 articles. Based on ‘normative’ arguments, an emotional appeal that constructs an ‘us/them’ divide is fostered. It seems obvious that explaining the judicial context with differing norms, opt-outs and protocols will not suffice to invalidate the fundamental normative argument, especially if it is referred to on emotional, not argumentative grounds:
From now on, whenever the UN, the EU, or the Council of Europe attacks our rightful sovereignty, whenever they attempt to abuse their power, whenever they try to force their politically correct agenda down our proverbial throats we need to tell them, in the politest way possible, to get stuffed.
With this discursive constellation, it cannot be astonishing that on 27/04/08, the SBP announces a swing in opinion polls towards a negative outcome and accuses the Yes-side of ‘arrogance’ (5). As in France, finding an alternative approach takes time for the Yes-campaign. In a sequence comparable to the French debate’s turn to an open argumentative situation, articles in TIT (13/05/08) and IEX (13/05/08) now analyse the treaty and ask voters not to give in to arguments based on fear (7).Footnote 38TIT exposes the argumentation of both sides concerning the question ‘Does the treaty express new values?’ and concludes:
Legal experts are divided on the charter’s possible impact. […] The final impact of making the charter legally binding may not become fully clear until the ECJ builds up its case law over the next decade if Lisbon is ratified and enters into force. But, overall, Lisbon confers few new competencies on the Union and probably goes further than any previous treaty in defining exactly when the EU has the right to act and when competence lies with states.
Based on this judgment, it is not possible to decide which side is ‘right’. However, at the very end of the article, a paragraph explicitly deals with abortion:
Claims that a treaty change would pave the way towards abortion have also surfaced, with Libertas claiming in leaflets that abortion could fall under ‘the free movement of services directive’. Yet Lisbon does not change the protocol inserted into the Maastricht Treaty in 1992, which confirms that nothing in EU treaties shall affect the Irish Constitution’s article 40.3.3 (right to life of the unborn).
This example shows the similarities of the French and Irish campaigns. In both cases, the referendum was seen as the potential move from the ‘status quo’ to an alternative unclear in most of its consequences. As in the French case, the No-side used the early time of the campaign to create resonance between this argument and a normative genealogy. However, whereas in France this normative genealogy was the old and established Republican constitutional discourse, the Irish No-side referred to the established discourse on abortion and EU referenda in Ireland. In the absence of a powerful Yes-campaign that would explain the difficult relationship between national and international rights and norms, the climate of fear and stereotyping that was underlying the Irish No-campaign’s argument was enough to make more than one voter doubtful.
As in France, after a first period where both sides had tried to disqualify each other (‘liars’, ‘conspiracy’…) and a second period characterised by argumentation, a third period now sees ‘discursive action’ to convince voters.
(Un-)successful discursive action
If we look at the frequency and stability of arguments, the now following month (15/05–15/06/08) shows (again as in France) a clear bias in favour of the Yes-side (9). This bias comes as (a) arguments by politicians and campaigners of the Yes-side, (b) a huge number of letters to the editor that allow a journal to denigrate the No-side without being held (directly) responsible for it, and (c) analysis and opinion directly issued by the newspapers. The semi-official newspapers analysed here all more or less clearly favoured a Yes-vote. Against these 45 contributions, I found nine articles in favour of the No-campaign’s arguments on abortion. To understand why a majority of voters nevertheless believed that a Yes-vote would change Ireland’s legislation on abortion (Red C poll, as related in SBP, 22/06/08), the discursive action of both sides must be analysed.
Both sides presented ‘independent’ and ‘neutral’ actors in order to give legitimacy to their claims. The most remarkable interventions came from the Irish Bishops’ Conference (30/05/08) and from the Referendum Commission (05/06/08).Footnote 39 Both instances were pressured to declare that abortion is not an issue within the LT, and both complied with these pressures. All neutral or ‘pro-Yes’ articles dealing with the subject after their interventions referred to them. Yet the articles dealing with the No-side and their supporters were less convinced.
Most of the newspaper reactionsFootnote 40 referred to the statement by the archbishop of Dublin and head of the Irish Bishops’ Conference, Dr Diarmuid Martin, during the press conference that accompanied the presentation of the bishops’ pastoral letter (11). The bishop said: ‘I do not believe that this treaty changes the current position with regard to Ireland’s position on abortion within the EU.’ Yet TIT on 30/05/08 is the only paper to give full coverage to the press conference, and it is somewhat more nuanced. In fact, the Bishops’ Conference did not explicitly urge a Yes-vote. And the pastoral reflection did not state that abortion is not an issue in the LT, this is only the orally given opinion of the archbishop, quoted in all newspapers. Yet a ‘pastoral reflection’ is normally read in all catholic churches during the dominical mass. The title of the document – ‘Fostering a Community of Values’ – gives a first hint of the Bishops’ perspective. While TIT correctly quotes the bishops’ condemnation of ‘misleading or even patently incorrect advice’ within the campaign, the bishops’ letter refrains from identifying who or what would fall under such advice. From a sceptical view, this could also be the ‘hidden agenda’ of the government. The fourth section of the pastoral letter finishes with a paragraph that could even confirm the suspicions of the No-side:
In a climate of legal positivism attempts may well be made to use traditional language concerning human dignity in ways which are contrary to traditional sense. Court decisions on a national or EU level tend to interpret language. Administrative decisions may well tend to opt for particular interpretations of norms. It is to be hoped that our public representatives will make a public commitment to engage actively with a broad coalition across the EU to resist interventions of this kind and to combat attempts to weaken the sense of subsidiarity which is an essential dimension of the Treaty of Lisbon. We also appeal to the Courts to respect the ethical values of human dignity that have been the sustaining pillar of European Culture (Irish Bishops’ Conference, 2008).
This is an ambiguous statement that seems to confirm both sides in their argumentations. It can be understood to express the fears of the pro-life side in a subliminal way, that there is an alternative understanding of the fundamental norms of social life in Europe ready for use by European judges against the ‘Irish’ understanding of these norms.
The Irish Independent, the least pro-Lisbon of the analysed papers, published an ‘opinion’ on ‘I’ll be voting ‘No’ for the sake of democracy’ on 30/05/08. Literally in the centre of the now open debate (10), it finds the Bishop’s statement
[…] would have been more credible had it also named some of the initiatives it does not agree with, such as the EU funding of abortion in the developing world, EU support for abortion at the UN and other international fora, and EU funding of embryo research. It might also have mentioned EU funding of far-left groups dedicated to advancing an agenda entirely inimical to Europe’s Christian heritage, for example, to radically defining the family in ways that are indubitably ‘contrary to common sense.
In a nutshell, we have the sceptical position that leans, in the absence of clear answers to complicated questions, towards a negative vote. The No-side enforces its arguments through emotional appeals and us/them stereotyping:
It doesn’t help, of course, that there is a constant, low, anti-religious drumbeat emanating from the EU, something that is especially strong among the often rabidly anti-religious socialists in the European parliament. […] Third, there is the Charter of Fundamental Rights. This sounds innocuous enough but will serve to override our own time-tested Constitution whenever our national laws clash with EU law. In other words, the Charter’s interpretation of rights will override our own interpretation of those same rights. That is not democratic.
The second ‘moral authority’ that the Yes-side is relying on is the Referendum Commission and its Chairman, the Irish constitutional judge, Justice Iarfhlaith O’Neill (12). In a press briefing on 04/06/08, the Referendum Commission announced intervention in the debate, as it ‘believe[s] there may be some confusion on a number of issues’. Among them, we find ‘The effect of the Treaty on Ireland’s constitutional law on abortion’. With regard to this point, the Commission’s statement is clear: ‘[…] this means Ireland’s constitutional position on abortion would not be affected by the ratification of the Lisbon Treaty’.
What should come as a boost to the Yes-side transformed into ambiguity in the Irish Independent’s coverage (05/06/08): ‘Referendum questions stump the experts’ translated the hesitations and unclear answers that the Commission Chairman gave to questions by the media.
Their chairman, Judge Iarfhlaith O’Neill, later insisted they had merely ‘hesitated’ in a bid to give accurate answers’ […] two minutes and thirty-three seconds later […] ‘Well, it’s eh… It’s quite difficult to be precise about what that means’ said Mr O’Neill. Thirty seconds later: ‘There certainly isn’t a precision about it whereby we could say it applies to a, b, c, or d.’ Pressed further, he said they would return to it later. Except there was no later. This was the last test before the climax of next Thursday’s polling day. Two minutes into the long dazed silences, one helpful broadcaster suggested an answer. But by then, the test was over and the Referendum Commission had raised more questions than answers.
Lacking coherence and consistency, this discursive act could not be convincing and it became understandable (from within the Irish discourse) why a No-vote might be the ‘rational’ choice.
So far, this analysis allows for three conclusions on the Irish case. (1) Ireland with its tradition of referenda on European treaties has an established genealogy on European issues. This makes it easier for the marginal actors of the No-side to create resonance and to turn into accepted discursive actors. It suffices that a foreign MEP raises three keywords (corporate tax, direct foreign investment, abortion) and the debate starts off, with the different genealogies and discursive elements ready for use. The semi-official discourse analysed here predominantly shows the attempt of established forces to prevent the No-side from entering the debate. The No-argument is accessible through its denial. (2) Again, as in the French case, frequency and stability of the arguments give an advantage to the No-side. Starting their campaign early, the No-campaign can frame the debate and create resonance with the existing EU referenda genealogy. Under pressure, it adapts its argumentation. However, the Irish No-side differs in this respect from France – ‘universal mission’ and ‘progress’ are not part of the Irish genealogy. Instead, actors and voters are concerned with protecting Ireland from the rest of the EU, and the discourse is that of a democracy that is challenged by an external power. (3) As in the French case, the Yes-side, at pains in countering the easy arguments of the No-side, draws on ‘neutral’ moral authorities in the discourse. What could seem as an intelligent move has not the same positive effects as Simone Veil’s appearance in France. The first actor, the Bishops’ conference, is ambivalent in its message. A sceptical electorate could even be confirmed in its doubts. The second actor, the Referendum Commission, gives an incoherent appearance. In the concluding section, we can compare the results of this analysis and link them to the debate introduced in the introductory section.
The future of Europe: civil society/ies and constitutionalisation
Drawing together and comparing the results of the two case studies, two constellations of political conflict in national referendum campaigns on Europe emerge. Briefly reviewing and contrasting them, I will discuss their possible consequences for EU’s constitutionalisation.
Both campaigns show patterns comparable under the framework of discourse analysis. Whereas resonance seems to be the most important condition for constituting an actor or an issue, the discourse’s external dimension (social hierarchy) is less important. This may be an effect of internet campaigning able to challenge discursively established social hierarchies (Fouetillou, Reference Fouetillou2007). Yet, it is also a consequence of marginal actors that constitute a ‘constitutional debate’ and thus gain access for their normative arguments. In liberal democratic discursive environments, there is a principal possibility for access by using and referring to the underlying social and political norms. These are mainly shared European norms, yet they can develop in different constellations. In the observed cases, two different patterns emerged. In Ireland, with an already existing discourse, actors could refer to an established genealogy to become legitimate voices in the debate. This was successfully done by ‘civil society’ movements that fostered the No campaign. In the French case, the left No’s interpretation of the Charter turned ‘abortion’ for the first time into an issue of political conflict over Europe. To constitute into a legitimate discourse, the (at first marginal) argument had to be framed within the French republican genealogy. This and the reference to ‘European civil society’ gave the PCF not only access to the broader debate, the frequency and stability of the argument turned it into an actor that could set the terms of the debate. The other discursive actors had to react to their arguments.
This initial setting subsequently structured the dynamics of the campaigns, yet we found different paths here. Whereas in the Irish case, the actors invoked ‘national sovereignty’, in France, the internal dynamics forced the left No-side to adapt its argumentation during the debate and to integrate ‘social progress and human rights’ for all Europeans. Table 1 gives an overview of the constitution of actors and discourses, the dynamics of the campaign and attempts at discursive action.
Comparing actors and discourse, dynamics, and discursive action
Finally, based on these two patterns of structuring political conflict in national arenas, some claims can be made regarding the initially evoked debate on ‘politicising’ the constitutionalisation of the EU. Two different patterns of ‘political conflict over European integration’ (Marks and Steenbergen, Reference Marks and Steenbergen2004: 4–10; Liebert, Reference Liebert2007: 243) become apparent. Whereas (1) in the Irish case, the Euro-scepticism of the political (extreme) right fosters a contestation based on ‘traditionalism, authoritarianism, nationalism’ (Liebert, Reference Liebert2007: 243; Marks and Steenbergen, Reference Marks and Steenbergen2004: 4–10), (2) the French case shows an alternative outcome. Initially similar – the political opposition of the (extreme) left is fused with its Euro-scepticism – this constellation breaks up due to the argumentative processes in the referendum campaign. The left can politically not afford to be enclosed in a constellation of ‘nationalism’ vs. ‘europeanism’. Instead, once the Yes-side starts arguing, the left No-side has to adapt its strategy, appealing to a ‘European’ citizenry and ‘another Europe’.
It then seems obvious that the differences around a question of fundamental norms are not ‘enshrined’ in national identities that are inaccessible to deliberative processes. Rather, ‘deliberation’ in a political debate on Europe leads to specific paths that allow for different outcomes. Context and social structures, if important for any analysis, do not determine the results. It is at least as necessary to integrate the developing social (political) practices within the current relationship between the European Union and its member states.
Acknowledgements
This is the revised version of a paper presented at conferences in Florence (EUI), Delmenhorst (University of Bremen) and Mannheim (Chair of Political science and Contemporary history). I would like to thank the discussants and participants of these conferences, as well as the editors and anonymous reviewers of EPSR for their very helpful comments and critique. The usual disclaimer applies.
