Successful Joint Venture or out of Control?
Framing Europe on French and Dutch Websites

Renée Van Os
Radboud University Nijmegen

Baldwin Van Gorp
Radboud University Nijmegen

Fred Wester
Radboud University Nijmegen

The Netherlands

Abstract: This paper focuses on online political communication about Europe produced by French and Dutch political parties and NGOs in the context of the 2005 referendum on the European constitution, and compares this Internet-based communication with news in French and Dutch newspapers in the context of the same event. The aim of this study is to disclose the (sometimes hidden) frames within these various types of political communication, and determine whether these frames, conceptualized as common understandings of what constitutes “Europe,” are cross-nationally shared among political actors. In the inductive phase of the study, three frames have been reconstructed. In each frame Europe is portrayed in a different manner: (1) in the Donor Frame, Europe is portrayed as a successful joint venture; (2) in the David vs. Goliath Frame, as an oppressive superstate; and (3) in the Invention Frame, as out of control. In the deductive phase of the study, the three reconstructed frames were subsequently examined for their presence in a larger set of texts (n = 268). The most commonly shared understanding appeared to be the Donor Frame, which was employed by 81% of the political actors in the two countries.

Introduction

This study is executed within the context of the referendum on the European constitution held in spring 2005 in France and the Netherlands. We focus on Europe-related lines of reasoning provided by French and Dutch political actors (i.e., political parties, NGOs, and journalists) in their communications about Europe, European integration, and/or the European constitution (henceforth: Europe). The theoretical starting point is the notion of European public sphere, which takes shape through the sharing of common understandings of what constitutes “Europe.” This sharing of common understandings is manifest in the cross-national employment of similar frames. In earlier studies we have shown that political actors with the same political position (left/right-wing) employ a similar European or national focus and express a similar attitude towards Europe (Van Os, 2005; Van Os, Wester & Jankowski, 2007), but we do not yet know whether various types of political actors apply the same frames in their external communication about Europe. Therefore, the aim of the study is to disclose the (sometimes hidden) frames within political communication about Europe present in newspapers and on websites produced by political parties and NGOs, and to determine the extent to which these frames are cross-nationally shared. The general research question is:

Can cross-national similarities be observed in the manner in which various political actors in France and the Netherlands employ frames in their communication about Europe during the 2005 referendum campaign on the European constitution?

Online political communication from political parties and NGO’s will be compared with offline communication about Europe in newspapers. In contrast to journalists, who create news based on professional criteria of newsworthiness, political parties and NGOs produce content from their own perspective on issues and events in society. These extensions of the mediated public sphere potentially help citizens to become better or more broadly informed about European political affairs, something which is considered an asset to (European) democracy (e.g., Norris, 2001, 2003; Tsagarousianou, 1999).

In the next section, the theoretical notions of a (European) public sphere and framing are discussed. Then, the two phases of this study are methodologically outlined, and subsequently the results of these two phases are presented. In the inductive phase of the framing analysis a repertoire of potential frames is reconstructed by defining the framing and reasoning devices by which the frames may manifest themselves in texts. It is followed by a second deductive phase, consisting of a systematic investigation of the extent to which the frames are shared among a larger set of diverse texts – produced by the online actors political parties and NGOs as well as by journalists of offline newspapers.

Theory

The public sphere and the Internet

The notion of public sphere, first elaborated by Jürgen Habermas(1964/1974, 1989), places emphasis on the deliberative and discursive aspects of democracy. It could be summarized as being the intermediary communicative space between politics and society, or as the “forum in which private people come together to form a public” (Habermas, 1989: p. 25). In the academic tradition that has emerged around Habermas’ initial contribution, the public sphere has been granted the normative status of being “the space within which the affairs of the state can be subjected to public scrutiny” (Kunelius & Sparks, 2001: p. 11).

During the last two decades, scholars have agreed that this ideal type of public sphere, if it ever existed, does not exist any longer. Mass media such as newspapers and television seem to have largely taken over the role as communicative space. Nonetheless, several scholars have also argued that this mediated public sphere is in fact less open in terms of equal access for private individuals, since it is created by media organizations that have their own selection criteria (for a review of this hegemonic perspective see, e.g., Bennett et al., 2004; Savigny, 2002).

In this context, scholars have been discussing the role the Internet could play in democracy and the public sphere. Early “cyber optimists” like Rheingold (1993), who claimed that the Internet could fuel the process of democratisation through opportunities for deliberation and direct decision-making, have been succeeded by “cyber pessimists” like Margolis and Resnick (2000), who warned that the Internet can even widen the gap between the engaged and the apathetic. Scholars like Norris (2001, 2003), Foot and Schneider (2002); Schneider & Foot (2002) and Ward, Gibson, and Lusoli (2003) take a more middle ground position, and suggest a balance between these two extremes, pointing out specific positive developments or aspects. Foot and Schneider (2002) stress the importance of independent political websites developed by national and state advocacy groups, civic organizations, and mainstream and alternative press. Norris (2001, 2003) mentions the existence of websites produced by minor and fringe parties, enabling citizens to learn more about the range of electoral choices than was previously possible.

Taking the moderate position of these scholars into account, the added value of the Internet should particularly be seen as potentially facilitating citizens broadening their knowledge about political opinions and the views existing in society. In comparison to a public sphere that manifests itself in the traditional mass media, more diverse actors participate in an online public sphere. In modern society, a broad range of websites is available to citizens for information or communication about political affairs, thereby serving as platform for a public sphere. These websites are created by diverse political actors, such as politicians and political parties, lobbyists, and advocates who represent the interests of a specific group, moral entrepreneurs, and intellectuals who try to generate public attention for particular issues, and finally media actors.

The Notion of European Public Sphere

In the last two decades the concept of public sphere has begun to play a central role in academic discussions about European integration. Our interpretation of what constitutes the European public sphere places emphasis on political actors, including citizens, communicating about Europe, either directly (face-to-face) or indirectly through media or Internet-based representations. This Europeanized political communication can, in the present situation, mainly be found within national media systems, and is often produced by national political actors (e.g., Koopmans et al., 2004; Meyer, 1999; Ward, 2001).

Roughly two approaches can be distinguished that measure elements of Europeanization of political communication, and subsequently a (mediated) European public sphere. The first approach essentially measures how often Europe, European institutions, or European issues are mentioned in the mass media (e.g., Gerhards, 2000; Trenz, 2004). A second approach concentrates on the occurrence of cross-national similarities in the manner inwhich the mass media communicate about Europe. Scholars refer to this approach as framing (Risse, 2003; Risse & Van de Steeg, 2003; Semetko, De Vreese, & Peter, 2000; Van de Steeg, Rauer, Rivet & Risse, 2003; Trenz, 2004). An extensive overview of empirical studies conducted within both approaches is provided elsewhere (Van Os, Jankowski & Wester, in press). The research project presented in this article can be placed within the second approach: it investigates whether similar frames – common understandings of what constitutes “Europe” – are employed by political actors from various EU member states in their communication about Europe.

Framing in the European Public Sphere

Simon and Xenos (2000) discuss the dynamics of a public sphere in which various political actors participate: political actors all try to define political and social issues by using certain frames, an act which results in an often implicit indication of the underlying problems of the issues, the designation of causal and treatment responsibility, and the passing of moral judgments (see also Nelson, Clawson, & Oxley, 1997; Entman, 1993). Cappella and Jamieson (1997) stress that frames select particular aspects of reality, organizing those aspects around a central idea and, thus, emphasize how to look at and interpret issues and events (see also: Gitlin, 1980: p. 11; Pan & Kosicki, 1993). Reese (2001: p. 11) defines frames as “organizing principles that are socially shared and persistent over time, that work symbolically to meaningfully structure the social world.” Finally, Trenz (2004: p. 308) speaks of a thematic field, that is, “the specific meanings, expectations and world views that are channelled through and conveyed by debates in the public sphere.”

Framing analysis is concerned with the investigation of frames as competitive ways of making sense of reality that manifest themselves in the public sphere. Each of these competing frames makes a validity claim to define the situation at hand. Many scholars identify generic frames in texts: frames that trace journalists’ habits in reporting the news (cf. Semetko & Valkenburg, 2000). Examples of generic frames that have been investigated within media communication about Europe are the Conflict Frame and the Human-Interest Frame (De Vreese, 2005). However, this approach does not provide a straightforward way to define or give meaning to the reported news event or issue under debate, which can be seen as the main function of framing (Entman, 1993).

A second type of frame does involve different definitions of an issue or event: issue specific frames. For example, Risse and Van de Steeg (2003) have defined two frames in news about the rise of the controversial Austrian politician Jörg Haider: the Europe as Moral Community Frame and the Europe’s Legal Standards Frame (Risse & Van de Steeg, 2003: pp. 6-7). Schuck and De Vreese (2006) have investigated the presence of two frames evaluating the EU enlargement either as a risk or as an opportunity for Europe in the news. These valenced frames evaluate political issues or situations in either positive or negative terms, whereas we want to focus on frames that go beyond notions of negative and positive, or favorable and unfavorable (see also Tankard, 2001: p. 96). Issue frames are closely linked to the particular issue they are developed for, whereas we prefer to focus on frames that are not exclusively connected to Europe but that make an appeal to already existing cultural knowledge. As such, the process of social construction underlying the incorporation of these frames by message receivers remains largely invisible and unnoticed; as a result these frames are probably more credible and convincing than other kinds of frames (cf. Gamson, Croteau, Hoynes, & Sasson, 1992).

In this study we look for certain ways of thinking about Europe that are promoted and legitimized by political actors and which assure an association between Europe and a broader cultural phenomenon. It is exactly this collective character and the frames’ cultural embeddings that make them suitable for giving meaning to various issues, events, and to a new reality, that is, to Europe and the European constitution. Framing analysis should therefore first of all be aimed at disclosing the broader cultural phenomenon to which the issue or event under investigation appears to be associated with by frame sponsors, in our case by political actors communicating about it. The analysis starts with an inductive phase, in which a repertoire of frames and the reasoning devices by which the frames manifest themselves in texts is reconstructed from a small sample of texts. Next, in a deductive phase, the extent to which the frames are employed in the larger set of relevant texts is determined (see also Van Gorp, 2005).

The Inductive Phase

Method

As argued by Van Gorp (2007), frames can be reconstructed and presented as interpretative frame packages (cf. Gamson & Lasch, 1983). Each package includes two types of indicators by which the frame can be identified: framing devices and reasoning devices. Framing devices include metaphors, lexical choices, and catchphrases (see also Gamson & Modigliani, 1989). Further, a frame package consists of reasoning devices, which are causal statements that are connected to the four functions of framing in the communication process, as identified by Entman (1993: p. 52): defining a problem, assigning responsibility, passing a moral judgement, and reaching possible solutions. Reasoning devices can be latently present in the text, whereas framing devices are always manifestly present. The true weight of the package is the central idea that connects the various framing and reasoning devices in a particular text with culturally shared phenomena such as values, myths, or archetypes.

The reconstruction of the frame packages in the inductive phase can most adequately be performed through an in-depth qualitative and interpretative analysis of a small sample of texts that are representative for the material to be included in the study. The goal of this interpretative analysis is to identify logical chains of framing and reasoning devices across the separate texts (Van Gorp, 2005: p. 487; cf. Tankard, 2001). In the study presented in this article, websites produced by three Dutch political parties were closely studied on statements made by these actors in the 2005 referendum campaign. The selected parties hold various political positions in the political system: one left-wing party (Socialist Party, SP), one right-wing party (Volkspartij voor Vrijheid en Democratie, VVD) and one centre party (Christen Democratisch Appèl, CDA). By including websites produced by parties with different political positions, we expected to find a broad variety of statements on all kind of aspects of the European constitution that would show clusters of related statements.

Per frame package, the reasoning devices were formulated by answering the following questions for each text: (1) How does the author define the problem or situation? (2) Does the author give an indication of the source creating the problem or situation? (3) To whom/what does the author assign responsibility for the problem or situation (cause)? (4) Does the author suggest solutions/remedies to solve the problem or ameliorate the situation? (5) What moral judgements are made? Across the three websites and the various texts, statements logically “belonging” to each other were combined together in a frame package. An extensive list of feasible statements was thus collapsed into three mutually exclusive frames, each with its particular central idea and structure of framing and reasoning devices. Also, for all three frames a counter-frame was defined, that is, a frame which challenges the legitimacy of the opposite frame by resisting it, or by re-framing the issue in terms of an alternative cultural idea. A limited number of frames and their respective counter-frames were reconstructed, because we wanted to reach acceptable levels of coder agreement in the deductive phase (see Tankard, 2001: pp. 102-103).

The overall packages and reasoning devices composing these packages were finally checked for their applicability to websites created by three French political parties upholding similar political positions (the left-wing party Ligue Communiste Révolutionaire, the extreme right-wing party Front National, and the centre-right party Union pour un Mouvement Populaire, UMP), and five Dutch and five French NGO websites; this check concerned an examination of the presence of the identified reasoning devices for each frame package within this small selection of these websites. No new reasoning devices were formulated during this check. Simultaneously, a list of framing devices was composed for each frame: Combinations of words, catchphrases, and metaphors reflecting aspects of the frame were taken (literally) from both the French and Dutch texts.

Results: Three Frames Giving Meaning to “Europe”

The signature matrix presented in Table 1 presents the result of the inductive phase. Such a matrix presents the coherent structure of framing and reasoning devices for each frame package (see also Gamson & Lasch, 1983).

In the process of reconstruction of the frames we first recognized the cultural archetype of the helper or donor, addressed by Propp (1928/1958) as central idea. Propp identified this archetype as one of the key narratives in folk tales: a donor, as magical agent, takes care of the hero, providing him/her with a magic potion. We found that political actors also apply implicitly this archetype to the European context: they refer to Europe as an agent taking care of its citizens (through cooperation, legislation, policy, etc.). Accordingly, the Donor Frame defines Europe by emphasizing the positive consequences of European cooperation for citizens, for example, in terms of peace and safety, but also in economic and social terms. Here, the definition of the situation is: “Europe delivers a distinguished surplus-value.” In the Donor Frame, Europe is presented as successful joint venture providing European citizens with all kinds of goods and services. Globalization is the main reason why individual member states must continue and even expand their cooperation in order to cope with issues such as migration, external trade, and terrorism. For example, the Dutch party Democraten ’66 argued that: “This constitution brings the Netherlands economic strength; together it will become easier to solve large-scale problems such as terrorism, international crime and environmental issues.” The French Parti Socialiste makes a similar argument: “This constitution will bring citizens more security, facing international crime together (terrorism, money laundering, and violations of human rights).”

Within the Donor Counter-Frame this surplus value is resisted. For example, the left-wing party Ligue Communiste Révolutionaire argued that: “This constitution is dangerous, it consecrates the absolute superiority of the ‘free market.’” The extreme right-wing party Front National also employed the Donor Counter-Frame: “We oppose this antisocial Europe that destroys our jobs.”

Second, we identified the cultural motive addressed in the Bible story of David and Goliath (cf. Dahinden, 2006) as central idea. In this story, David is a shepherd boy who opposes a suppressing force represented by the giant Goliath. At the end of the story, David defeats Goliath. Political actors applied this cultural motive to the European context: the individual nation states are challenged by the EU, or large countries within the EU, that has become too powerful. The policy solution they put forward is that nation states should conquer the “giant” Europe by restricting or even reversing European integration. Accordingly, the David vs. Goliath Frame addresses the loss of national sovereignty and identity to a dominant Europe. In this frame, Europe is presented as superstate. For example, the Dutch Socialist Party argued that “The European constitution means handling over even more authority and power to Brussels. It will become a superstate. The Netherlands, as small country, will have no say over its own foreign, defence and justice policies anymore.” Second, the French Front National argued that “The European constitution handles over total power to the European institutions. It is a death threat to our nation. We must remain in charge of our own destiny.”

In the counter-version of the David vs. Goliath Frame, the idea of Europe as a superstate is resisted by arguing that the unique character and sovereignty of the individual nation state is not threatened by European integration but preserved. For example the Dutch Partij van de Arbeid (PvdA) argued that “We do not have to fear Europe’s interference in everything. Issues like the Dutch drugs policy, gay marriage and euthanasia – there is no need to fear that The Netherlands has to give up anything in any of these areas.”

Third, we recognized the cultural myth of an invention that ends up turning against its inventor – with the monster of Frankenstein as extreme example – as central idea. This iconic character from the novel written by Mark Shelley, intended to be beautiful, turns out to be a horrid creature. More realistic examples of such inventions are dynamite and nuclear power. We discovered that political actors applied this cultural myth to the European context: they claim European citizens have lost control over the European Union – an invention they created themselves, and that was initially intended to serve them. Accordingly, the “Invention that turns against its inventor” frame (henceforth: Invention Frame) deals with the current state of the European Union: how it has deteriorated over time into an elephantine, ungovernable and undemocratic body. In this frame, Europe is presented as out of control: European citizens have lost control because European legislation is too complex and is the result of one-sided decisions taken by the European political elite. Policy solutions include increasing the democratic character of the EU with more transparency and involvement of European citizens. The moral basis appealed to by the frame is the idea that a democratic EU should serve its citizens and not vice versa. For example, the Dutch Socialist Party stated that: “If you say YES to this European constitution, you say YES to an undemocratic and non-transparent Europe in which citizens have no say.”

Within the Invention counter-frame the idea of Europe being or becoming undemocratic and ungovernable is resisted. This idea is for example expressed by the Dutch Democraten ’66: “The European constitution makes the European Union more democratic and resolute through more effective decision making, better division of powers and more control on European policy.” Resistance to the Invention Frame also becomes manifest in statements like “Citizens should take responsibility by becoming more actively involved in European issues, instead of only complaining they were not informed” (Parti Socialiste).

The Deductive Phase

Research Questions

In the deductive phase of our analysis we surveyed a large collection of French and Dutch websites and newspaper texts on the use of the three frames reconstructed in the inductive phase. Comparisons of the use of the frames are drawn between the two countries included in the study, between the various political actors, between the two different types of communication, and between advocates and opponents of the European constitution. This has lead to the following specific research questions:

RQ1: To what extent do French and Dutch political actors employ the three frames in their communications about Europe?

RQ2: Do differences exist concerning the “richness” of the three frames, in terms of numbers of reasoning devices explicitly elaborated on within each frame?

RQ3: Can differences in frame use be observed between opponents and advocates of the European constitution in the Netherlands and France?

RQ4: Can differences in frame use be observed between the various actor types, that is, political parties, NGOs, and newspapers, in France and the Netherlands?

Method

Selection of texts. Texts were selected based on the criterion of containing Europe-related lines of reasoning: those not (only) focussing on national political affairs and/or not (only) focussing on campaign-related affairs, but rather on particular aspects of Europe, European integration, the EU, the European constitution, or consequences on a social, economic, political policy and/or legislative level.

As concerns the collection of websites, in the six weeks prior to the referendum in both countries search engines [1] were used on a weekly basis to search for sites produced by NGOs [2] and political parties expected to be involved in the 2005 referendum on the European constitution. In the week prior to, and in the week after the referendum, all collected sites were archived with the tool Teleport Ultra [3]. Only eight French and nine Dutch political parties turned out to provide Europe-related lines of reasoning on their websites in the 2005 campaign on the European constitution. These were mainly the political parties with seats in the European Parliament. Other political parties were, as a result, not included in this study. In both countries quite a few NGOs turned out to provide Europe-related lines of reasoning on their websites. Yet, more French (n = 65) than Dutch (n = 49) NGOs did so and therefore were included. This can be considered a sign of a higher level of political activity concerning European political issues outside the institutionalised political order in France. For each website, one key document was selected in which the actor put forward its “Europe-related” lines of reasoning – used to advocate or oppose the European constitution.

Further, a collection of newspaper articles focusing on the 2005 referendum was composed for both countries. In total, 69 Dutch and 72 French newspaper articles were included. These articles were selected in the two weeks prior to the referendum and in the week after the referendum, based on the criterion that they contain Europe-related lines of reasoning [4]. There was significant attention paid to the referendum on the European constitution in the newspapers included, but only a small part of this news stories truly dealt with the aspects that interested us.

Coding instrument. On the basis of the matrix resulting from the inductive phase, a coding instrument for the deductive analysis was developed that measures the actual presence of the framing and reasoning devices in the entire collection of texts. The coding instruction and the coding sheet contained three separate parts: (1) actor’s opinion on the European constitution; (2) presence of framing devices; and (3) presence of reasoning devices.

First, the actor’s opinion on the European constitution was measured. Coders could choose between (a) advocate; (b) neutral/unclear/mix; and (c) opponent. For each actor, its opinion towards the European constitution was measured on a scale from advocate (score 1) to opponent (score 3). In a general sense, French actors appeared to have a slightly more negative opinion (M = 2.09) on the European constitution than Dutch actors (M = 1.90), although this difference was not significant: t(266) = 1.827, p = .069. More variation can be observed when the two countries are compared. In France, for political parties (M = 2.38) and NGOs (M = 2.40) quite negative means were calculated, in contrast to the relatively positive mean for French newspapers (M = 1.77), F(2, 141) = 9.122, p < .001. In the Netherlands, a similar but slightly less contrasting pattern could be observed: Newspapers (M = 1.72) were more positive than political parties (M = 1.89) and NGOs (M = 2.15), F(2, 121) = 3.884, p < .05.

Second, coders were instructed to highlight the framing devices in the text with a felt tip pen. For each frame, a list of lexical choices (framing devices) plus possible synonyms was included in the coding instruction, both in French and Dutch. Coders also needed to determine which of the identified frames was dominant in the text by weighting the number of sentences in which the framing devices of the frames are present. Coders could choose between (a) frame not present in the text; (b) frame present but not dominant; and (c) frame present and dominant in the text. Third, at the more interpretative level, coders needed to determine whether the reasoning devices systematically put together in the matrix were present in the text. The coding instruction elaborated on the statements in detail – explaining the core components of each statement. In some cases, the statement was manifestly present, but in many cases the statement needed to be extracted from the text and was only latently present. Training sessions were aimed at reaching a sufficient level of inter-coder agreement among the six coders about the level of interpretation: it was important to know when to stop interpreting. This was especially important for the reasoning devices and their (potentially) latent presence in a text. This knowledge was obtained by practicing together using several example texts.

Reliability. A stratified sample of 55 texts (about 20% of the total number of texts) was double-coded in order to determine the exact level of inter-coder agreement. For each variable, two reliability coefficients were computed: Holsti and Krippendorf’s α [5]. Whereas Holsti only calculates the percentage of agreement, α corrects for chance agreement in computing a reliability assessment (Riffe, Lacy, & Fico, 2005: p. 149). As lowest level of reliability we considered the following figures acceptable: .60 for α and .80 for Holsti. In doing so we follow the lead of scholars who consider these figures acceptable for studies with a qualitative, interpretative nature (cf. Riffe et al., 2005: p. 151). Those variables that met both criteria were included for further analysis [6]. Other variables appeared to be too complex or difficult to identify in the text, especially those variables that were mainly latently present in the texts. These variables are not reported on in this article.

It was within the following two areas that the inter-coder agreement was insufficient. First, the extent of domination of all three counter-frames had not been coded reliably enough. Compressing the two measures present but not dominant, and present and dominant into one measure present resulted in one of the reconstructed counter-frames meeting the criteria. As a result, this article only reports on one counter-frame, that is, the Donor Counter-Frame; for the other two frames, the normal and counter versions of the frame are jointly reported on (as one frame). Second, some of the elements within the logical chain of reasoning devices in the three frame packages had been coded insufficiently reliable, especially those that were not very often present in the texts. By only looking at statements made within the dominant frame in the text, however, most statements were coded reliably.

Results: Cross-National Comparisons in Employment of Frames

In Table 2 the overall employment of the three frames by the political actors is presented. In both countries political actors employed the Donor Frame the most often: In 81% of the 268 texts the Donor Frame could be identified. The other two frames were less often employed – in about 50% of the 268 texts. This means that in most texts two frames were employed. French actors employed the Donor Frame somewhat more often than Dutch actors: (86% compared to 75%). Yet this difference is not significant. In contrast, Dutch actors employed the David vs. Goliath Frame and the Invention Frame significantly more often than French actors: 65% versus 38% for the David vs. Goliath Frame, and 67% versus 36% for the Invention Frame. The Donor Frame is more dominant in France than it is in the Netherlands, where each frame was used to about the same extent.

Our next question concerns the extent to which political actors elaborated on the various elements of the frame packages outlined in Table 1. The extent in which political actors explicitly include one or more of these elements in their texts is an indicator of the ‘richness’ of their use of these frames. However, as mentioned earlier, the coding of the reasoning devices turned out to be quite difficult. Still, if we concentrate only on the dominant frames in texts, the following can be concluded for all the three frames: nearly all political actors elaborate on the definition of the situation/ problem. For example, in 107 of the 110 texts (97%) in which the Donor Frame was dominant, the reasoning device “European cooperation has surplus value” could be identified. For the David vs. Goliath Frame this percentage was 88%, for the Invention Frame 95%. Political actors elaborated less on the other three elements of the frames.  For example, in the texts in which the Donor Frame was dominant, actors elaborated only in 23% of these texts on the source and in 40% on the causal responsibility for the problem.

As most of the texts contained lines of reasoning for voting “YES” or “NO,” we then investigated the frame use of advocates and opponents of the European constitution. Table 3 elaborates on the observed similarities and differences between advocates and opponents of the European constitution as concerns their employment of the three frames.

Table 3 shows that advocates, neutrals and opponents of the European constitution employed the Donor Frame to a similar extent. 83% of the advocates and 82% of the opponents employed the Donor Frame. When looking at the differences in employment of the normal and counter versions of this frame, it becomes clear that advocates particularly employed the normal version of this frame (82%, in contrast to 13% of the opponents). In contrast, opponents employed the Donor Counter-Frame the most often (76%, in contrast to 6% of the advocates). Political actors with a neutral opinion on the European constitution employed the two versions of the frame to about the same extent.

The David vs. Goliath Frame was employed somewhat more often by advocates than by opponents; 60% versus 50%. This pattern is visible in both countries, yet differences are not significant. The neutrals employed this frame much less often: 33%. Because of unreliable coding, no difference could be made between normal and counter version of the David vs. Goliath Frame, as was also the case for the Intervention frame.

As concerns the employment of the Invention Frame, variation was mainly observed among French actors; here, advocates of the European constitution employed this frame less often in comparison to the opponents: 22%, in comparison to 49%. Among Dutch actors, this variation was not observed – all employed the Invention Frame to a similar extent, around 67%.

Next we come to the question whether frame use is related to the actor type producing the text: relatively extensive texts from political parties elaborating on all kinds of subjects related to Europe, compared to small texts from NGO’s discussing some subjects relevant to their perspective, and small texts from newspapers related to European news events.

In Table 4 the three actor types are compared with each other on their employment of the three frames; also this table compares the actors along the lines of their national basis. The overall variation in the employment of the Donor Frame between French and Dutch actors (as outlined in Table 2) can be observed among all three actor types. Variation was especially observed among French and Dutch news actors: in 70% of the Dutch newspaper articles the Donor Frame could be identified – in contrast to 85% of the French newspaper articles.

For the Donor Frame it was measured separately whether the frame was employed in its normal or counter version. Results show that almost all Dutch political parties employed the normal Donor Frame (eight out of nine of the Dutch parties included in the study), whereas the French political parties were more mixed in terms of either employing the normal or counter version of this frame. For the other two types of actors, French and Dutch counterparts showed a similar pattern: (1) NGOs particularly employed the counter version of the frame: 54%, in contrast to 36% normal version; (2) newspapers employed the normal version of the frame most often: 59%, in contrast to 34% counter version.

With regard to the other two frames, the David vs. Goliath Frame and the Invention Frame, as mentioned before it was the Dutch actors that employed these frames significantly more often. Only one French actor type employed these frames to a similar extent than its Dutch counterpart: the political parties. As concerns the other two actor types, the results show that Dutch NGOs and newspapers employed these frames significantly more often than their French counterparts. For example, 48% of the Dutch NGOs employed the David vs. Goliath Frame, in comparison to 28% of the French NGOs. Additionally, in 67% of the Dutch newspaper articles the Invention Frame could be identified, in comparison to 28% of the French newspaper articles.

Conclusions

The type of frames investigated in this study differs from other types of frames often being discussed in framing literature. Generic frames seem to lack the quality or suitability to define issues and to identify causal and treatment responsibility. Issue-specific frames do not correspond to the rule of thumb that the same frame should be applicable to define multiple issues. Therefore, in this study we have opted for reconstructing frame packages in which a culturally shared idea functions as a central organizing idea. Yet, this does not mean that a focus on culturally embedded frames needs to be seen as a totally new approach to framing: there are clear linkages with both generic and issue-specific frames. The David vs. Goliath Frame falls under the umbrella of the Conflict Frame, which is dominantly used in news stories and has a much stronger capacity to construct meaning. If this familiar story is applied as a core idea, it becomes clear that we should sympathize with the weakest side in a battle in which unequal weapons are used. In contrast to the Social Responsibility and the Economic Consequences Frames, two generic frames which are often perceived as mutually exclusive frames, we perceive aspects of responsibility and consequences of importance for each frame. For example, when the Donor Frame is applied to define the European Union it becomes clear that this supranational structure is not driven by self-interest but that it places itself in the service of the economy and social well-being of the European member states and their citizens. Finally, with the use of the Invention Frame the opposite idea is expressed, that is, the European Union has become a purpose in itself; moreover, it turned against its creators, the individual member states.

The limitation in the inductive phase to only three frames seemed to be a requirement for achieving sufficient inter-coder reliability coefficients. Although the identification of frames can be rather subjective, we expanded the inductive study with a deductive analysis in which the three frames were systematically coded for their presence in newspapers and on websites. The deductive phase has demonstrated that within one media text multiple frames can be applied; the use of one frame does not necessarily exclude the use of another, even an opposing one. The same is true for websites and newspapers. Existing research shows that within a text usually only one frame suffices to turn a story into a comprehensible whole (see, e.g., Van Gorp, 2005). The complexity of the issue at stake in our study – Europe, European integration, and the European constitution – likely means that the use of only one frame is not sufficient to make the issue understandable and comprehensible for the general public.

A comparison of the extent to which all constituting parts of the pre-defined frame-packages are explicitly included within a single newspaper article or website shows that the definition of the situation is touched upon in more than 9 out of 10 texts. By contrast, the ‘richness’ of the applied frames is rather limited, because the causal and treatment responsibility, the possible solutions, and the moral basis on which one relies are less frequently included, especially by French newspapers and NGOs. With regard to the inclusion of several reasoning devices, only the political parties, both in France and in the Netherlands, adequately refer to the background of the issue at hand. This may partially be related to the fact that the texts produced by these actors turned out to be relatively long, in comparison to texts produced by NGOs and in newspapers. In these rather extended texts, political parties usually put forward various arguments. One limitation of our study is the fact that we did not take the length of the texts into account. In addition, it was not always possible to make adequate comparisons between several types of actors. The infrequent presence of part of the reasoning devices in the material is the main reason why this aspect turned to be difficult to code in a sufficiently reliable way. In attempting to reveal the latent meaning structures of texts, and to make a cross-cultural comparison of common understandings, we have inevitably run into difficulties of inter-coder reliability. By being very explicit about our coding procedures we leave the decision to the reader as to whether this attempt was successful and worth a follow up.

The inclusion of counter-frames, that is, frames which argue against the general idea which is expressed by the opposite frame, enabled us to make a distinction between the directions in which the three frames were applied, something which is often not included in a framing analysis. The Donor Frame was employed by most political actors in the two countries, either in the normal version or in its opposing variant, that is, the counter version. This cross-national employment of the Donor Frame is an indicator for the sharing of common understandings of what constitutes “Europe” among French and Dutch actors, namely being a successful joint-venture that is able to handle problems that manifest themselves increasingly on a large scale within Europe.

Also the David vs. Goliath Frame and the Invention Frame form common understandings, yet these were less present than the Donor Frame – especially among French actors, who employed these frames less often in comparison to their Dutch counterparts. Observed variations could be linked to the opinion an actor appeared to hold on the European constitution and to a lesser degree to the actor type. Actors with a relatively positive opinion (particularly newspapers) employed the Donor Frame more often; actors with a relatively negative opinion (particularly NGOs) preferred to employ the Donor Counter-Frame. However, none of the three frames can be labelled as negative (the David vs. Goliath Frame and the Invention Frame) or as positive (the Donor Frame) as such. This implies that, on some issues, framing goes beyond notions of “pro” or “con.”

The results of our cross-national comparison and focus on a diversity of actors lead to a more general conclusion about the notion of public sphere in an online environment. The Internet indeed provides a forum to (decedent) voices that are often overlooked by the mainstream news media, as was illustrated by the many NGO websites gathered in our collection. For citizens who want to be informed on European issues, the Internet contains a lot of communication about Europe produced by a variety of political actors. Most of these actors, especially NGOs as less-institutionalized actors, are not visible in the mass-mediated public sphere. Yet our study has also shown that these actors do not communicate in a very ‘rich’ manner when employing a frame (especially the French NGOs). Political parties as online actors score better on this aspect, providing more diverse frames in their communication about Europe. Nonetheless, we believe this situation to be an asset to European democracy, as is also pointed out by scholars such as Norris (2001, 2003). The question as to whether this online communication about Europe enhances citizen’s involvement in the European political process (thereby reducing the EU’s democratic deficit) has not been investigated in this study. We can only speculate on this matter.

As concerns the reconstruction of the three frame packages, we make no claim to be exhaustive; there are probably other alternative frames used to give meaning to Europe. As we wanted to be certain of acceptable levels of inter-coder reliability, a repertoire of only three frames seems to be both acceptable and unavoidable. Yet we have confidence that the three reconstructed frames cover important lines of reasoning about Europe as present in the collected material in the context of the referendum on the European constitution. Whether this is also true for political communication about Europe during any other event, or in other countries, can also be the object of future research.

Endnotes

[1] For the Netherlands, the search engines www.google.nl and www.ilse.nl were used. For France, www.google.fr and www.altavista.fr were used. Keywords: “Europese grondwet” for the Netherlands, and “constitution européenne” for France. The first 500 hits (50 pages) were checked for their relevance to increase the chance of finding not only the mainstream organizations, but also the minor/fringe organizations.

[2] NGOs include issue advocacy groups, social movement organizations, activist groups, special interest groups (including labour unions), and religious organizations.

[3] See: http://www.tenmax.com.

[4] Dutch newspapers included in the study: Volkskrant, NRC Handelsblad and Telegraaf. French newspapers included in the study: Le Monde, Le Figaro, Libération. If possible, on each day one relevant article was taken from the front page, and one from the first section of the newspaper. If the newspaper had devoted a specific section to the referendum, one or two relevant articles were taken from that section, too. In total, approximately 20 up to 25 articles were selected for each newspaper during the three week period.

[5] There are three reasons for having chosen Krippendorf’s α: (1) it is uncertain whether we can report on Scott’s π in our study, since it is questionable whether the absence/presence of a variable can be considered a ordinal-level measure (minimal demand for Scott’s π); (2) both Krippendorf’s α and Cohen’s κ compute higher reliability figures than Scott’s π when one value of a category is used much more often than others, which is the case for some variables in our study. So Scott’s π will not be reported on; (3) Krippendorf’s α corrects for small samples, in contrast to Cohen’s κ (Riffe, Lacy, & Fico, 2005: pp. 148-153).

[6] The following table shows the results:

References

Bennett, W.L., Pickard, V.W., Iozzi, D.P., Schroeder, C.L., Lagos, T., & Caswell, C.E. (2004). Managing the public sphere: Journalistic construction of the great globalization debate. Journal of Communication, 54(3), 437-455.

Dahinden, U. (2006). Framing: Eine integrative Theorie der Massenkommunikation. Konstanz: UvK.

De Vreese, C.H. (2005). News framing: Theory and typology. Information Design Journal + Document Design, 13(1), 51-62.

Entman, R. (1993). Framing: Toward clarification of a fractured paradigm. Journal of Communication, 43(4), 51-58.

Foot, K.A., & Schneider, S.M. (2002). Online action in campaign 2000: An exploratory analysis of the U.S. political Web sphere. Journal of Broadcasting & Electronic Media, 46(2), 222-244.

Gamson, W.A., & Lasch, K.E. (1983). The political culture of social welfare policy. In S. E. Spiro & E. Yuchtman-Yaar (Eds.), Evaluating the welfare state: Social and political perspectives (pp. 397-415). New York: Academic Press.

Gamson, W.A., & Modigliani, A. (1989). Media discourse and public opinion on nuclear power: A constructionist approach. American Journal of Sociology, 95(1), 1-37.

Gamson, W.A., Croteau, D., Hoynes, W., & Sasson, T. (1992). Media images and the social construction of reality. Annual Review of Sociology, 18, 373–393.

Gerhards, J. (2000). Europäisierung von Ökonomie und Politik und die Trägheit der Entstehung einer Europäischen Öffentlichkeit. In M. Bach (Ed.), Die Europäisierung nationaler Gesellschaften. Sonderheft der Kölner Zeitschrift für Soziologie und Sozialpsychologie (Vol. 40, pp. 277-305). Opladen: Westdeutscher Verlag.

Gitlin, T. (1980). The whole world is watching. Mass media in the making & unmaking of the new left. Berkeley, CA: University of California Press.

Grimm, D. (1995). Does Europe need a constitution? European Law Journal, 1(3), 282-302.

Habermas, J. (1974/1974). The public sphere: An encyclopedia article (S. Lennox & F. Lennox, Trans.). New German Critique, 1(3), 49-55.

Habermas, J. (1989). The structural transformation of the public sphere. Cambridge, England: Polity Press.

Koopmans, R., Statham, P., Kriesi, H., Della Porta, D., De Beus, J., Guiraudon, V., et al. (2004). The transformation of political mobilisation and communication in European public spheres - preliminary draft of end report. Europub.com project, available at: http://europub.wz-berlin.de.

Kunelius, R., & Sparks, C. (2001). Problems with a European public sphere. Javnost - The Public, 8(1), 5-20.

Margolis, M., & Resnick, D. (2000). Politics as usual: The cyberspace revolution. London: Sage.

Meyer, C. (1999). Political legitimacy and the invisibility of politics: Exploring the European Union's communication deficit. Journal of Common Market Studies, 37(4), 617-639.

Nelson, T.E, Clawson, R.A., & Oxley, Z.M. (1997). Media framing of a civil liberties conflict and its effect on tolerance. American Political Science Review, 91(3), 567–583.

Norris, P. (2001). Digital Divide: Civic engagement, information poverty, and the Internet worldwide. Cambridge, England: Cambridge University Press.

Norris, P. (2003). Preaching to the converted: Pluralism, participation and party web sites. Party Politics, 9(1), 21-45.

Pan, Z., & Kosicki, G.M. (1993). Framing analysis: An approach to news discourse. Political Communication, 10(1), 55-75.

Propp, V.J. (1958). Morphology of the folktale. Bloomington: Indiana University (Original work published in 1928).

Reese, S.D. (2001). Introduction. In S.D. Reese, O.H. Gandy & A.E. Grant (Eds.), Framing public life: Perspectives on media and our understanding of the social world (pp. 1-31). Mahwah, NJ: Erlbaum.

Rheingold, H. (1993). The virtual community: Homesteading on the electronic frontier. Reading, MA: Addison-Wesley.

Riffe, D., Lacy, S., & Fico, F.G. (2005). Analyzing media messages: Using quantitative content analysis in research. Mahwah, NJ: Erlbaum.

Risse, T. (2003). An emerging European public sphere? Theoretical clarifications and empirical indicators. Paper presented at the Annual meeting of the European Union Studies Association (EUSA), Nashville, TN, March 27-30, 2003.

Risse, T., & Van de Steeg, M. (2003). An emerging European public sphere? Empirical evidence and theoretical clarifications. Paper presented at the International conference: "Europeanisation of public spheres, political mobilisation, public communication and the European Union," Science Center Berlin, June 20-22, 2003.

Savigny, H. (2002). Public opinion, political communication and the Internet. Politics, 22(1), 1-8.

Schneider, S.M., & Foot, K.A. (2002). Online structure for political action: Exploring presidential campaign Web sites from the 2000 American election. Javnost - The Public, 9(2), 43-60.

Schuck, A.R.T., & De Vreese, C.H. (2006). Between risk and opportunity. News framing and its effects on public support for EU enlargement. European Journal of Communication, 21(1), 5-32.

Semetko, H.A., De Vreese, C.H., & Peter, J. (2000). Europeanised politics, Europeanised media? European integration and political communication. West European Politics, 23(4), 121-142.

Semetko, H.A., & Valkenburg, P.M. (2000). Framing European politics: A content analysis of press and television news. Journal of Communication, 50(2), 93–109.

Simon, A., & Xenos, M. (2000). Media framing and effective public deliberation. Political Communication, 17(4), 363-376.

Tankard, J.W. (2001). The empirical approach to the study of media framing. In S. D. Reese, O. H. Gandy & A. E. Grant (Eds.), Framing public life: Perspectives on media and our understanding of the social world (pp. 95-106). Mahwah, NJ: Erlbaum.

Trenz, H.-J. (2004). Media coverage on European governance: Exploring the European public sphere in national quality newspapers. European Journal of Communication, 19(3), 291-319.

Van de Steeg, M., Rauer, V., Rivet, S., & Risse, T. (2003). The EU as a political community: A media analysis of the 'Haider debate' in the European Union. Paper presented at the Annual meeting of the European Union Studies Association (EUSA), Nashville TN, March 27-30 2003.

Van Gorp, B. (2005). Where is the frame?: Victims and intruders in the Belgian press coverage of the asylum issue. European Journal of Communication, 20(4), 484-507.

Van Gorp, B. (2007). The constructionist approach to framing: Bringing culture back in. Journal of Communication, 57(1), 60-87.

Van Os, R. (2005). Framing Europe online: French political parties and the European election of 2004. Information Polity: An international journal of government and democracy in the Information Age, 10(3-4), 205-218.

Van Os, R., Jankowski, N.W., & Wester, F. (in press). Exploring the online European public sphere: The Web and Europeanization of political communication in the European Union. In: J. Harrison & B. Wessels (Eds.), Mediating Europe: New Media, Mass Communications and the European Public Sphere. Oxford, England: Berghahn.

Van Os, R., Wester, F., & Jankowski, N. W. (2007). Presentations of Europe on political party Web sites during the 2004 EP election campaign. Javnost - The Public, 14(2), 63-82.

Ward, D. (2001). The democratic deficit and European Union communication policy. Javnost - The Public, 8(1), 75-94.

Ward, S., Gibson, R.K., & Lusoli, W. (2003). Participation and mobilisation online: Hype, hope and reality. Parliamentary Affairs, 56(3), 652-668.

Copyright 2008 Communication Institute for Online Scholarship, Inc.

This file may not be publicly distributed or reproduced without written permission of the Communication Institute for Online Scholarship,

P.O. Box 57, Rotterdam Jct., NY 12150 USA (phone: 518-887-2443).