By Franca Bätz, student of Communications at the University of Münster.

Turkey's President Tayyip Erdogan talks to the media before a meeting with European Commission President Jean-Claude Juncker at the EU Commission headquarters in Brussel
Turkey’s President Tayyip Erdogan talks to the media before a meeting with European Commission President Jean-Claude Juncker (not pictured) at the EU Commission headquarters in Brussels, Belgium, October 5, 2015. REUTERS/Francois Lenoir

1. Introduction

“Now more than ever – The EU needs to negotiate again with Ankara“ (Zeit Online, May 12 2014) –  “Europe is not a place for Erdogan’s Turkey” (FAZ, August 12 2014) – “Now the EU explained: we should be seriously concerned about the independence of the Turkish judiciary and the protection of fundamental rights” (Die Welt, October 8, 2014). These are only excerpts of the German mass media debate about the Turkish accession to the EU, one of the most controversial issues in the discourse about European integration (de Vreese, Boomgarden, & Semetko, 2011). The debate about Turkish accession has been controversially loud and emotional, and has involved geographers, historians, theologians, political scientists, economists and sociologists. It is discussed to what extent Turkey meets the Copenhagen criteria, which include the rule of law, democracy and protection of minorities. For the first time even the question of the European character of a candidate country is a matter of debate: Is a country with a majority Muslim population European with regard to history, religion and culture? (Madeker, 2008)

After the 2004 European Council announced that Turkey meets the Copenhagen criteria, the EU entered accession negotiations with Turkey and consequenttly debates about Turkish accession were increasing (Madeker, 2008). Hence, all studies discussed in this paper were published in 2004. The aforementioned newspaper headlines illustrate, however, that the debate about the Turkish accession is still ongoing. Eurobarometer data show that 59% of the EU population oppose Turkish membership of the Union  (European Commission, 2007). Applying the framing approach, this paper analyses how citizens form an opinion about Turkish accession to the EU by using journalistic information. Opinions in the EU are not static but rather suggestible by new information (de Vreese et al., 2011). Therefore, it will first outline the framing approach and discuss the theoretical operation of frames. Then the framing approach is applied to three studies that shed light on actors and voices participating in the debate and detect existing media frames in coverage. Finally, the effects of these frames on public opinion are analysed. This serves the purpose of answering the question of this paper: What effects do frames about the Turkish accession have on public opinion within the EU?

2. History and Status of Accession Negotiations

The first contractual relationship between the EU and Turkey dates back to 1963 (‘Ankara Agreement’), but accession negotiations only began four decades later. Turkey was the second country (next to Greece) with which the European Economic Community (EEC) signed an Association Agreement. In addition to economic clauses, the agreement included the possibility of Turkey joining the EU. Turkey was not, however, awarded the official status of candidate country until 1999. Accession negotiations began in 2005., after the European Council announced that Turkey fulfils the Copenhagen criteria at the end of 2004. 14 out of 35 chapters were opened, but only one, namely ‘Science and Research’ could be provisionally closed. In December 2006 accession negotiations were partially suspended due to the ongoing conflict in Cyrpus and Turkey’s refusal to recognise the Republic of Cyprus. Since then, this decision has been renewed annually. The Turkish government has attempted to give new impetus to the negotiations by advocating a settlement of the Cyprus conflict on the level of United Nations. In addition, the Turkish Prime Minister reiterated the request for EU membership and the opening of further chapters in 2014, including Chapter 23 ‘Judiciary and Fundamental Rights’ as well as Chapter 24 ‘Justice, Freedom and Security’ (Federal Foreign Office, 2014).

The content of these chapters is viewed as the most important conditions for Turkish possible accession to the EU.  To illustrate, 85% of EU citizens think that human rights in Turkey have to be respected. 77% think that the state of economy needs to be improved (European Commission, 2007). Data of the Eurobarometer (2007) show that the majority of the EU population (59%) opposes Turkish EU membership. Only 28% of the EU population supports membership, whereas 13% are indecisive about the issue. Even the Turkish population does not evaluate EU membership as unambiguously positive. 38% of the Turkish population think that an EU membership would be a positive development and only 33% of the Turkish population are in favour of joining the EU (European Commission, 2014)

The following chapter will clarify how the issue of EU membership is framed in mass media coverage. Furthermore, it is analysed to what extent existing frames influence public opinion in the EU. More specifically, the impact of frames about Turkish membership on the opinion of the EU population will be investigated.

3. The Framing Approach

3.1 Basic Concepts

This chapter presents the basic assumptions and methods of the framing approach. Like other political issues, the topic of EU accession negotiations with Turkey is very complex, so recipients can only understand it selectively. Therefore, recipients form their opinion about this political issues by highlighting specific arguments while neglecting others. Thus, certain angles in the people’s cognitive apparatus arise. A Turkish EU membership can be viewed as a promising enlargement of the EU or as a cultural threat to a Christian-oriented community, depending on the angle of vision. This angle of vision is strongly influenced by journalistic information from the mass media. However, other communicators such as organizations, politicians or companies (strategic communicators), also have a certain point of view that they would like the public to adopt. The different perspectives that can be taken on the same issue, whether in media content or the people’s cognitive apparatus, are referred to as frames. Even in a public debate there is a certain point of view on an issue. What perspective will ultimately prevail in the public sphere depends on three different types of frames: (1) the journalist’s perspective appearing in the content of mass media, (2) the strategic communicator’s perspective and (3) the recipients’ perspective. To summarize, frames are understood as actor’s ‘horizons of meaning’ (Matthes, 2014).

Accordingly, framing is the process of selecting and attaching importance to specific information and perspectives:

“To frame is to select some aspects of a perceived reality and make them more salient in a communicating text, in seeking a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation for the item described.” (Entman, 1993, p. 52)

Entman (1993) names four functions of framing: problem definition, causal interpretation, moral evaluation and treatment recommendation. (1) The definition of a problem means that relevant actors and aspects of a topic are defined and the highlighted information is identified. (2) The causal attribution is a key issue of political formation of opinion. It is relevant when a condition is judged negatively or positively. (3) A moral evaluation is a moral or judgmental classification of an issue. (4) The treatment recommendation offers necessary actions to solve the problem (Matthes, 2014). A frame can also possess more, less or none of these four functions (Entman, 1993).

In its entirety the framing approach deals with the origin, change and effects of frames at the level of communicators, media content and recipients (Matthes, 2014). This paper only examines existing media frames about the Turkish accession to the EU as well as their impact on recipients’ attitudes.

3.2 The Framing Process

The perspective of reciepients on a political issue not only strongly depends on journalistic information, journalistic frames themselves are also further influenced by the framing of strategic communicators. Strategic communicators try to give importance to their frames by public relations (strategic frames). They want their frames to be adopted by journalists. However, instead of adopting strategic frames unreflectively, journalists bring their own viewpoint and interpretation (journalistic frames). Frames can determine the selection of news as journalists may select frames fitting into an existing scheme of previous coverage. Journalistic frames are closely linked to the so-called media frames that can be identified in journalistic information (Matthes, 2014). The frame adopted by the recipient represents the selective use of media content. Recipients accept and store some viewpoints, while they neglect others (Matthes, 2014).

How exactly, and to what extent do frames affect recipients? As mentioned before, frames emphasize some aspects about a topic which therefore become more salient. That means some information become more perceptible, significant or memorable. An increased salience increases the likelihood that the information is perceived, recognized, processed and remembered by the recipient. The salience of information may be increased, for instance, by their placement, by repetition or by combination with culturally familiar symbols. Kahneman and Tversky (1984) conducted an experiment through which they demonstrated that different frames produce different results by selecting and highlighting some features while others are omitted. The participants were asked to choose between two alternative programs to combat the outbreak of a disease. Two identical but differently framed options were the following: “If program A is selected, 200 of 600 people will be saved” (win-frame) and “If program B is selected, 400 of 600 people die” (loss-frame). Although being identical, program A was selected more often than program B, since program B was framed in terms of likely deaths rather than likely lives saved.

Repetitions or culturally familiar symbols increase the likelihood that recipients save information and viewpoints because the brain connects information by associations. If information is retrieved, it will be activated. At the same time, information connected with this association is activated, too. Whether or not information is activated depends on how many times information was accessed and when the last activation took place. The shorter the distance to the last activation and the more frequently activation occurs, the more available the information is. That is relevant since the effect of frames consists of three steps. First, the information needs to be cognitively available so that it can be activated in the first place. That is, the information of that frame must be understood immediately. After that, the information needs to be accessible, respectively the information of the corresponding topic must be immediately remembered when the recipient thinks of it. Afterwards, the activated information must be applicable to the prevailing situation. In case of multiple receptions, media frames also have an influence on whether certain information is deemed applicable or suitable. As a consequence, the recipient applies the frame in a situation and therefore has a framed view of an issue (Matthes, 2014).

The salience of frames demonstrates that frames increase the likelihood of an issue being understood in terms of that frame. Thus, frames have an impact on the recipients’ cognitions. By activating information repeatedly and sequentially, frames are able to influence recipients’ opinions and attitudes thereafter since certain frames are deemed more applicable (Matthes, 2014). Media frames can therefore influence how recipients perceive, understand and remember problems or situations and how they act accordingly. They have an impact on a large part of the audience (Entman, 1993). Price and colleagues (1997) conducted an experiment in which recipients we asked to read newspaper articles about government aid for universities. The findings show that the frames increased the accessibility of the corresponding issue. Recipients further consulted the activated information for evaluations albeit these effects were less strong. In their studies Chong and Druckman (2007, 2010) also found framing effects and their moderating factors: The stronger the arguments of a frame the stronger the framing effect. Recent studies also show that the availability of emotions increase framing effects (Aarøe 2011). Nevertheless, this cannot be described as a universal effect on all recipients. It is also important to add that frames are not only characterized to emphasize certain information and viewpoints, they also omit certain information. Recipients are also influenced by the fact that they have little or no facts about alternative perspectives. That is why exclusion and accentuation of particular information have to be denoted as equal constituents of the framing process (Entman, 1993).

3.3 Methods – Identification of Frames and their Effects

When media frames are extracted from journalistic content, intersubjective accountability is of great significance, otherwise there is a subjective interpretation of the components of a journalistic text. The challenge is to distinguish frame-related elements such as certain arguments, keywords and metaphors from other elements. There are four main methods based on content analyses to identify media frames. (Matthes, 2014).

First, there is the manually holistic frame analysis. Here, frames are coded when they occur in a journalistic text. There is a distinction between inductive and deductive approaches. Deductive studies derive frames from the literature and code them in standard content analyses. In inductive manually holistic analyses, frames are first generated via qualitative analyses and then coded as holistic variables in manual content analyses (Entman, Matthes & Pellicano, 2009). A disadvantage of this analysis is that it may identify researcher frames instead of media frames, which means that those frames that the researchers expected in the first place are the ones mostly identified. However, the inductive manually holistic frame analysis can examine the sample quickly and easily. Each text is coded if a certain frame occurs (Matthes, 2014). The study conducted by Koenig and colleagues used this approach and identified several assumptions in a sample before they were clustered into frames. Keywords were used as indicators of frames (Koenig, Mihelj, Downey, & Bek, 2006).

In addition to manually holistic frame analyses there are qualitative analyses of frames. These analyses proceed inductively as described above. In other words, frames are formed on the basis of the available texts. At the same time frames are described in detail and marked with examples from the text. A quantification like the manually holistic analysis does not take place. Furthermore, qualitative approaches only use small samples. The detailed description of the material can be seen as an advantage whereas the subjectivity of description accounts for a disadvantage. Besides, these analyses often lack a detailed explanation of how the frames were identified in the text. Accordingly, as with the previous method, qualitative approaches might identify researcher frames instead of media frames (Matthes, 2014). Madeker (2008) identifies her frames about the accession of Turkey by qualitative analysis.

For the sake of completeness, manual dimension reduction analyses and computer-based analyses also exist. The first process does not identify the entire frame but patterns of frame elements that occur repeatedly in media coverage, corresponding to the classification by Entman (section 3.1.). Subsequently these elements are gathered into clusters which form the eventual frames. Computer-based content analyses or frame-mapping interpret word connections as media frames. Both methods identify frames using computer algorithms to guarantee intersubjective accountability whereas the distance of frames from the text accounts for a disadvantage (Matthes, 2014).

Having presented these methods to detect media frames, this essay will now present a research design to examine effects of media frames. Investigating framing effects means to examine the influences of frames on recipients’ attitudes and cognitions. Framing effects can be analysed, among others, with an experimental research design. Test subjects are randomly divided into two or more experimental groups, depending on the number of frames that is tested. In a first step, the participants read articles, each with a different frame (Matthes, 2014). The study by de Vreese and colleagues (2011), whose results will be introduced in the next chapter, uses an experimental research design. After reading the newspaper articles, the participants are questioned about their attitude towards the Turkish accession to the EU. The test groups are compared with regard to the dependent variables. Differences between the groups are described as framing effects. There is also a control group reading no newspaper articles to ensure that the effects are causally explained by variations of the experimental groups (Matthes, 2014).

4. Framing of EU Accession Negotiations with Turkey

4.1 Existing Frames in Coverage

The mass media provide forums to discuss political topics, such as the Turkish accession to the EU. In contemporary society recipients perceive problems and constitute public opinion in the public arena. For these reasons it is necessary to analyse the public media discourse so that media frames about the discourse of Turkish EU membership and its influence on the attitudes of EU citizens can be identified (Madeker, 2008). There are three significant studies worth touchin upon in this respect, which will be discussed below.

In her study Madeker (2008) examines seven German quality newspapers with the intention to cover the spectrum of political opinion. These newspapers belong to the most widely circulated newspapers in Germany. Among them are Die Zeit, which tends to be liberal, the Süddeutsche Zeitung and the weekly magazine Der Spiegel, which are both left of the centre, the Frankfurter Allgemeine Zeitung (FAZ) and the Frankfurter Allgemeine Sonntagszeitung being right of the centre as well as Die Welt and the Welt am Sonntag, which tend to be conservative.

The first step is to examine these newspapers for relevant newspaper articles and the distribution of voices in favor and against EU membership of Turkey. Three quarters of all articles that contain a discourse about the EU membership of Turkey is publishes in conservative newspapers Die Welt and the FAZ. Unlike the newspapers left of the centre or liberal ones, they established themselves as main forums of discourse. The majority of contra positions on Turkish accession to the EU can be found in these newspapers as well. However, it does not seem that it is only the editors of conservative newspapers that able to influence the discourse of mass media public, as more than half of the examined articles of the liberal Süddeutsche Zeitung include contra positions too. The articles in Die Zeit and Der Spiegel can be neglected due to the small number of relevant articles. Therefore, it can be summarized that the EU membership of Turkey, at least in German media, is generally rejected across the board  (Madeker, 2008). This finding is also supported by surveys. In Germany the Allensbach Institute elicits similar results to the European Commission: The majority of the population (66%) rejects a Turkish membership (Allensbach Institute for Public Opinion Research, 2004).

The newspaper articles were not only examined in terms of their distribution in newspapers and with regard to their pro or contra positions, but also with regard to voices of different actors. The largest share of voices comes from the readers themselves as ‘letters to the editor’ (44%). However, they mostly consist of short posts. The readers’ influence in mass media discourses is also rather low. Journalists represent the group of the second most-heard voices. The journalist’s attitudes usually correspond with the orientation of the newspaper they work for. The voices from Intellectuals form the third largest group. Furthermore,  EU officials are represented twice as many times (11%) as representatives of Turkish origin. Of all German politicians, conservative politicians are the most dominant group represented in the media discourse. One reason may be the party program of the CDU/CSU which involved the rejection of the intended southeastern enlargement. Writers of Turkish origin (‘Letters to the editor’, intellectuals and politicians) they only account for 12% of all actors (Madeker, 2008).

The sample of Koenig’s study (2006) contains all articles published in the period from October 7 to 14 and from December 17 to 24 2004 that discuss a Turkish membership. The study also focusses on quality newspapers since it is assumed that these papers are the most important in forming a European public sphere. Nevertheless, regional newspapers in Germany and France, as well as tabloids in the UK and Germany are taken into account as well, so that the structure of print media in the different countries can be reflected to a certain degree.

The study also identified frames of the Turkish print media, and hence frames of EU countries can be compared with those in Turkey. A first frame is referred to as ‘Clash of Civilizations’ and assumes that the Christian-oriented Europe and Islam do not share the same values and have no common political basis. In addition to a conflict between Christianity and Islam there is also an assumed conflict between the EU’s secular, liberal form of government and Islam questioning a separation of church and state (Koenig et al., 2006). Strongly represented in all EU member states (over 40%), the clash of civlizations frame is dominant in Germany, France and Slovenia. In Turkey, on the other hand, this frame only accounts for 11% of all media frames. However, the frame usually appears when the Turkish media report on arguments concerning the ‘Clash of Civilizations’ frame in EU countries (Koenig et al., 2006).

Whereas the ‘Clash of Civilizations’ frame is widespread in France, the ‘Multiculturalist’ frame is particularly common in the UK and Turkey. This frame emphasizes the right to cultural differences which are tolerated or even celebrated. Although individuals belong to different cultures, universal values are able to outgrow various cultures to create a way of living together and to redefine cultural debates. In Germany, this frame can be found in one third of all articles. This frame embodies an antithesis to the ‘Clash of Civilizations’ frame (Koenig et al., 2006).

Another frame emphasizes the economic costs and benefits of accession (Koenig et al., 2006). This frame mainly appears in the UK but is also widespread in other countries. In Germany, the ‘Economy’ frame essentially has a negative valence. Turkey’s accession is presented as a financial burden for the EU. Most of the time this frame is accompanied by arguments of the ‘Clash of Civilizations’ frame. Unlike in the UK where economic benefits are used as arguments for Turkey’s accession, the discourse in France is similar to the German one (Koenig et al., 2006).

Instead of a liberal, multicultural coexistence of two cultures in terms of the ‘Multiculturalist’ frames, the ‘Liberal Individualist’ frame proposes Turkey to accept liberal and secular forms and methods of governance. Here a potential for social, political and cultural changes in Turkey becomes apparent. This frame can be identified in all countries analysed in the study (Koenig et al., 2006).

Furthermore, the study reveals a ‘Nationalist’ frame. However, it is only supported by the Turkish opposition. This frame emphasizes the interests of Turkey that have to be considered. In general, they do not voice their opposition against EU membership but rather propose a privileged partnership. The fact that this frame does not exist in EU countries does not mean that there are no nationalist positions in those countries. The ‘Nationalist’ frames, however, only includes arguments pitting the EU as a whole against Turkey. The definition does not contain posit individual nation-states against others (Koenig et al., 2006).

The third study by de Vreese and colleagues identified existing frames in Dutch media by means of a content analysis. In this vein the analysis of framing effects could be connected to existing frames. This includes cultural and economic frames, each with positive and negative valence, overlapping with the definitions of the frames in Koenig’s study[1]. They found out that negative frames account for 32% of coverage of the Turkish accession to the EU, while positive frames only account for 15%. The sample consists of news broadcasts of the public radio station NOS[2] and its biggest competitor RTL. Additionally articles of the five newspapers with the highest circulation in in the Netherlands were analysed. Analysed newscasts and newspaper articles were released in December 2004, in advance and during the EU summit about Turkisch accession to the EU (de Vreese et al., 2011).

4.2 Framing Effects

The question that arises from the identification of different frames is the extent to which these frames affect public opinion in the EU. Frames with negative valence are used to determine the impact of negative frames on the rejection of Turkey’s accession by the majority of EU population, as the Eurobarometer data suggest. As mentioned before, media frames increase the likelihood that a topic is understood in terms of the frame. As a result they impact recipients’ attitudes and opinions. De Vreese and colleagues (2011) assume that frames also have a direct impact on recipients’ attitudes without changing the importance of the information contained in the frame. It occurs if the frame shows new aspects to the recipient or if the frame has a valence. Framing effects can thus be achieved in both ways.

The experiment conducted by de Vreese and colleagues (2011) shows that valued frames have a direct impact on attitudes. While test subjects reading an article with positive frames assess the possible EU accession of Turkey more positively, negative frames exert a negative influence on their opinions and views. A further step is to identify indirect framing effects and thus, investigating to what extent frames increase the significance of a particular information. For this, frames are clustered in order to analyse the importance of culture, security and economy. Indirect effects of cultural frames are significant. The reception of cultural frames thus strengthens the role of culture which means that cultural aspects are even more significant when evaluating an EU membership of Turkey. As a consequence the support for Turkey’s accession decreases since the cultural frame according to de Vreese and colleagues only contains negative valence. In addition, anti-immigration attitudes, an exclusively Dutch identity, an ideology leaning to the right and weaker political sophistication have negative effects on attitudes towards Turkey’s possible accession (de Vreese et al., 2011). The significant difference between effects of negative and positive frames is important. Effects of positive frames are stronger than those of negative frames. The difference between the effects, however, is quite small (de Vreese et al., 2011).

The results of the content analysis show that negative frames account for 32% of coverage of the Turkish accession to the EU, while positive frames only account for 15%. Considering the significant results of the experiment, one can conclude: The reception of these frames is likely to influence public support for Turkey’s accession (de Vreese et al., 2011). A predominance of negative cultural frames suggests that negative effects may occur.

5. Conclusion

The study by Madeker (2008) initially identified conservative print media in Germany as the main forum of debate about Turkey’s EU membership. Contra positions can be found in both conservative and  liberal newspapers. The public debate in the German print media coincides with public opinion polls about the accession of Turkey. Turkish positions are hardly represented in public discourse. Koenig’s (2006) study deals with the question which frames dominate coverage in the EU and the effects they have on public opinion. He identified five different frames: a ‘Clash of Civilizations’ frame, a ‘Multiculturalist’ frame, a ‘Liberal Individualist’ frame, a ‘Nationalist’ frame and a frame dealing with economic consequences. Whereas in France and Germany ‘Clash of Civilizations’ frames are dominating, in the UK the economic and ‘Multiculturalist’ frames are most established. It should be noted that these media frames are not typically European but rather transcontinental.

The study put forth by de Vreese et al. (2011) indicates small but significant differences between effects of negative and positive frames. Therefore this paper concludes that due to the dominance of negative frames in coverage and their slightly stronger effects, these frames have negative effects on the population’s attitudes about Turkey’s possible accession to the EU. Even these indirect influences are significant regarding the Dutch population. Effects on the European public opinion cannot be demonstrated by means of this study so that further studies examining the effects of media frames in other countries are needed.

While looking at studies dealing with the framing approach, their criticisms should also be mentioned. If framing effects on recipients’ attitudes are found in experimental setting, it does not automatically mean that the audience is manipulated by media coverage. The everyday forming of opinion is not fully comparable with the stimulus provided in experiments (Matthes, 2014). Framing effects are measured immediately after presenting the stimulus. Thus, only short-term effects are measured. Future research need to consider investigating long term effects of media frames on public opinion with regard to Turkey’s accession.

Experiments are also strongly stimulus-oriented. However, frames are only effective if they are perceived repeatedly and if they are consonant with existing contents (Matthes, 2014). Unlike many studies analysing framing effects, de Vreese et al. (2011) include predispositions like the test of peoples’ political education. Predispositions are relevant as they moderate the effects of frames. Politically educated citizens are able to understand political information better and integrate them into existing attitudes. This moderation, however, only occurs with regard to positive frames. As a consequence, politically educated citizens are influenced by positive frames more strongly. Despite a limited comparability of conditions in the experiment and in real-life conditions, the experiment is a suitable method to measure and prove framing effects (Matthes, 2014).

Moreover, effects of current developments, such as the blocking of many Turkish websites by Turkish authorities, on the accession negotiations with the EU but also on European public opinion should be investigated (Süddeutsche Zeitung, April 6, 2015). For instance, it should be examined if new frames like a frame of censorship appeared in the mass media. Further it should be considered if these frames will contribute to a lower support of Turkey as a potential EU member state.

Thanks for writing this article, it was very interesting to read! It is very clearly written and contains good analysis. I’ve made some minor corrections and comments for you to look at. Other than that, the only thing I would like you to pay some attention to is the fact that sometimes you quote studies or newspaper articles from a certain European country and simply assume the results generalize for the EU as a whole. This is not necessarily wrong, but I feel like you would need to justify why this is the case, or alternatively, you need to broaden the scope to include newspaper headlines from other EU countries as well (for your very first paragraph for instance).


Aarøe, L. (2011). Investigating frame strenght. The case of episodic and thematic framing. Political communication, 28(2), 207-226. Allensbach Institute for Public Opinion Research (2004): Beklommenheit vor dem historischen Schritt. Die Bevölkerung sieht überwiegend die Risiken der Osterweiterung. [Anxiety before the historic step. The population mainly sees the risks of the eastward expansion] ´ Chong, D., & Druckman, J. N. (2007). Framing theory. Annual Review of Political Science, 10, 103-126. Chong, D., & Druckman, J. N. (2010). Dynamic public opinion. Communication effects over time. American Political Science Review, 104(4), 663-680.

de Vreese, C. H., Boomgaarden, H. G., & Semetko, H. A. (2011). (In)direct Framing Effects: The Effects of News Media Framing on Public Support for Turkish Membership in the European Union. Communication Research, 38, 179-205.

Die Welt (2014, 8. Oktober). Gravierende Defizite: EU rechnet mit Türkei ab. [Wide deficits: The EU settles a score with Turkey]

Entman, R. M. (1993). Framing: Toward Clarification of a Fractured Paradigm. Journal of Communication, 34, 51-58.

Entman, R. M., Matthes, J. & Pellicano, L. (2009). Nature, Sources, and Effects of News Framing. In: K. Wahl-Jorgensen & T. Hanitzsch (Eds.) The Handbook of Journalism Studies (pp. 175-190). New York: Routledge.

European Commission (2007). Eurobarometer 66: Public Opinion in the European Union.

European Commission (2014). Eurobarometer 81: Public Opinion in the European Union.

Federal Foreign Office (2014, 15. Juli). EU-Erweiterung: Türkei. [EU enlargement: Turkey] (18.03.2015).

Frankfurter Allgemeine Zeitung (2014, 12. August). Die Erdogan-Türkei hat in Europa nichts verloren. [Europe is not a place for the Erdogan Turkey]

Kahneman, D., & Tversky, A. (1984). Choice, values, and frames. American Psychologist, 39, 341-350.

Koenig, T., Mihelj, S., Downey, J., & Bek, M. G. (2006). Media Framing of the Issue of Turkish Accession to the EU: A European or National Process? Innovation, 19, 149-169.

Madeker, E. (2008). Türkei und europäische Identität: Eine wissenssoziologische Analyse der Debatte um den EU-Beitritt. [Turkey and european identity: An analysis of the debate about joining the EU from the perspective of sociology of knowledge] Wiesbaden: VS Verlag.

Matthes, J. (2014). Framing. Baden-Baden: Nomos.

Price, V., Tewksbury, D., & Powers, E. (1997). Switching trains of thought. The impact of news frames on reader’s cognitive responses. Communication Research, 24(5), 481-506.

Süddeutsche Zeitung (2015, 6. April). Türkische Behörden sperrten Twitter und Facebook stundenlang. [Twitter and Facebook were temporarily blocked by Turkish authorities]

Zeit Online (2014, 12. Mai). Jetzt erst recht: Die EU muss wieder mit Ankara verhandeln. [Now more than ever: The EU need to negotiate again with Ankara]

[1] The cultural frame with a negative valence (Cultural Threat) corresponds with the ‘Clash of civilizations’ frame. The economic frames with positive and negative valence (Economic Benefits/Threat) correspond with the economic frame including both positive and negative consequences. In addition, there is a frame focusing on geopolitical security issues (de Vreese et al., 2011).

[2] Nederlandse Omroep Stichting (Dutch Broadcast Foundation)