Abstract
This paper examines how German media discourses reflect debates around integration, based on a newspaper corpus spanning the period 2008–2018. Considering these discourses, our research interest is focussed on how integration is constructed as a responsibility of those who are expected to integrate into society. To analyze how media might play a role in reproducing essentialist constructions of difference, we present a case study that combines methodologies of corpus linguistics and critical discourse analysis, and that examines discursive practices and strategies linked to dominant discourses about integration in Germany. Our findings show that in the German context, certain ethnicities (here: people of Turkish background) are framed as “other” and these groups are constantly being assessed in terms of integration in ways that others are not.
Keywords
Introduction
“Understanding how migrants are represented in the media is [. . .] key to understand the growing sense of crisis in both personal and collective identities that dominates the socio-discursive landscape of migration.” (Musolff and Viola, 2019)
The above quote goes to the heart of the paper here presented, as the focus of the discussion is on how German print media represent migrant identity, or rather ascribe identity to individuals identified as migrants. Our contribution examines how German media discourses reflect debates around integration, based on a newspaper corpus spanning the period 2008–2018. We combine methodologies of corpus linguistics and critical discourse analysis in order to recognize how media might play a role in reproducing essentialist constructions of both “us” and “them” (Morley and Taylor, 2012; Musolff and Viola, 2019; Perdue et al., 1990).
While changes to citizenship and immigration laws over the past two decades acknowledge the fact that Germany has long been a country of immigration, the formerly dominant ethnonational ideology about German identity still plays a role in public discourses. The discussion presented here is part of a larger research project that explores how the term “integration” in the context of migration is constructed in German public discourse. Understandings of, and positioning toward, migratory movements shape realizations of both borders and belonging (Fortier, 2012). Claiming the integration of migrants as a requirement of their belonging displays attitudes toward social structures in immigrant societies. These attitudes are reinforced and perpetuated through text and talk, yet often remain uncontested.
In 2018, Mesut Özil, born and raised in Germany and of Turkish background, became a person of public interest not because of his career in football, but because he retired from the German national team via a three-part statement on Twitter, following a controversial debate over a photo that he had posed for together with his team colleague İlkay Gündoğan and the Turkish President Recep Tayyip Erdoğan. 1 In his tweet, Özil (2018) said that what he “can’t accept, are German media outlets repeatedly blaming my dual-heritage and a simple picture for a bad World Cup on behalf on an entire squad.” The controversy as to whether taking such a picture was appropriate for a football player of the German national team quickly led to discussions about whether this incident was a symbol of unsuccessful integration, stating that Germany was now facing yet another debate on integration. The aim of the paper here presented is not to verify statements made during the discussions following the publication of the photo. Controversies over integration in Germany have been present long before this incident 2 and are likely to continue, for example as part of the dynamics around European integration, and as reflected in definitions of “Europeanness” (Wodak, 2018).
Rather, we present a case study that displays attributions attached to one person, Mesut Özil, and we examine discursive practices and strategies behind these attributions to link them to dominant discourses about integration in Germany. Considering these discourses, our research interest is focussed on how integration is often constructed as a responsibility of those who are expected to integrate into a society. Various aspects of this responsibility are discursively assigned to individuals who are expected to integrate; thus, integration becomes “a socio-political instrument [that] is connected to larger structuration processes that create and produce normativity and subaltern ‘others’, all the while proclaiming an integrative and inclusive policy” (Flubacher and Yeung, 2016: 606). Such structuration processes are often based on norms of the host society. Legally, those norms form part of requirements to receive German citizenship (Möllering, 2010). On a societal level, such norms are often not defined clearly, but rather perceived and discussed as practices (Fuller, 2019) both of the host society and of those yet to be integrated. 3 “Germanness,” for example, is not predominantly discussed as an ethnic trait anymore, but as a form of culture instead (Fuller, 2019: 334). Hence, attaining cultural belonging or success in integration, is seen as the responsibility of migrants, designating “the expected efforts of migrants to successfully partake in the project of their social and economic integration” (Flubacher and Yeung, 2016: 612). Such an approach defines integration as a one-way street, with the onus on migrants to adapt to the host society, referred to as “Bringschuld der Migranten” in the German context; a stance that has been criticized in favor of a two-way process (Fuller, 2018: 187; Schinkel, 2018).
At first sight this stance seems to give agency to the individual, but our conclusions suggest that there is a gap between what is expected and what is achievable according to these obligations. Discourses that posit integration as a dominant obligation of migrants constitute a divide between those who are, and those who are not yet integrated (Flubacher and Yeung, 2016; Fuller, 2018, 2019). This gap reduces or suppresses agency. According to Özil’s statement, he felt “German when we win, but I am an immigrant when we lose.” Hastings and Manning (2004: 293) have argued that “identity is always constructed in relation to alterity,” and that knowledge about “others” influences the shaping of one’s own identity. In our case study, we were able to find insights into how migrant identities are constructed by media, and how argumentations around such identity constructions reinforce alterity on the one hand, and the image of a culturally homogenous society on the other hand. Representing people as migrants or with a so-called migratory background thus positions “them” in comparison to “us” (Morley and Taylor, 2012; Van Dijk, 1998), and implicitly or explicitly constructs difference and boundaries.
Methodology
The case study presented in this paper forms part of a larger project which aims to examine discourses on migration and integration in Germany and Australia. The project as well as the case study follow an interpretivist research design which understands social interaction as a process of constituting meanings – reality – and, at the same time, co-constructed meanings as influencing social interaction (Berger and Luckmann, 1967). As stated in the introduction, we conducted discourse analytical research as it “primarily studies the way social power abuse, dominance, and inequality are enacted, reproduced, and resisted by text and talk in the social and political context” (Van Dijk, 2001: 352). This kind of research does not follow a single methodology, but is rather an “academic movement, a way of doing discourse analysis from a critical perspective, which often focuses on theoretical concepts such as power, ideology and domination” (Baker et al., 2008: 273), hence it can combine various and interdisciplinary methodological steps.
To analyze the data for our study, we chose a methodological triangulation of Corpus Linguistics, Qualitative Content Analysis, and Discourse-Historical Analysis to approach a corpus that consisted of newspaper articles reporting about or mentioning Mesut Özil. Thus, we integrated both the actual newspaper corpus as well as “the background of the social and political fields within which discursive events are embedded” (Wodak, 2018: 8) into the analysis. All three methodological approaches will be outlined in some detail in the following sections. In our article, we use several key terms that originate in the methodological approaches we combine, and that shall be defined here before we continue to describe our methodology in more detail.
We use the term “content” to refer to the subjects of a text or a part of the text, a text unit. The data in our corpus contained various subjects, or contents, that were identified and structured by means of Qualitative Content Analysis (for a more detailed description see below). The term “context” in our paper refers to linguistic and non-linguistic aspects. In corpus linguistics, context displays the instances around a search term in a concordance analysis, and considering the context around a search term allows to identify patterns of language use in general, and with regards to the search term specifically (Bubenhofer, 2009: 27). On a non-linguistic level, context in a wider sense can be seen as the societal situation around an oral or a written text. Context is considered “crucial” in Critical Discourse Analysis because it looks at language as social practice (Fairclough and Wodak, 1997), and in this practice the meaning of utterances is “context-determined” (Widdowson, 2004: 38). Furthermore, we understand “attribution” as the practice of ascribing features to someone or something. These attributions (un)intentionally serve to qualify or label and thus are related to context. In Discourse-Historical Analysis (for a more detailed description see below), “discursive qualification” helps to “achieve a particular social, political, psychological or linguistic goal” (Reisigl and Wodak, 2009: 94).
Corpus-linguistic approach
To identify the contents of the discourses reflected in our data, we applied a corpus-based approach as a first step. The aim was to establish a macro perspective on our corpus by deconstructing the entity of the texts into units that surrounded the key word or search term “Mesut Özil” 4 and thus offered insights into patterns and frequencies of utterances. In large data sets these contents run the risk of remaining unnoticeable when reading all texts in full length (Bubenhofer, 2018: 28). By using the subject of our case study as a search term, we gained a broad overview over “patterns of meaning, use, or attitude” (Gabrielatos and Baker, 2008: 6) in the concordance lines of “Mesut Özil.” Baker et al. (2008) have argued that while subjective positioning of researchers is immanent in any part of the research process, corpus-linguistic approaches still “offer the researcher a reasonably high degree of objectivity” (Baker et al., 2008: 277). The online database DeReKo (Deutsches Referenzkorpus, German Reference Corpus) by the Institute for German Language (IDS), a corpus which is available online via the “Corpus Search, Management and Analysis System” (COSMAS II) proved to be a rich database to select our corpus from, as it is one of the largest corpora of written texts in German (Bubenhofer, 2009: 175) 5 and at the time of data collection contained over 57 billion word forms (http://www.ids-mannheim.de/cosmas2/projekt/referenz/korpora.html). Within that database, we customized an individual corpus for our case study. It covered 10 years, from 2008 until 2018, 6 and contained all articles published by six newspapers: Der Focus, Spiegel Online, Die Süddeutsche, Der Tagesspiegel, Die Welt, and Die Zeit. These regional and nation-wide daily and weekly publications are both well distributed and sell well and are therefore considered to be leading German newspapers (Vollmer and Karakayali, 2018). We excluded the popular tabloid newspaper BILD from the corpus, as a large amount of work has already been done on this newspaper (see e.g. Heinkelmann-Wild et al., 2019).
Altogether, the corpus for our study consisted of 962,083,353 words derived from 2,013,563 texts. As smaller corpora have been questioned in terms of their representativeness (Koller and Mautner, 2004: 218; Stubbs, 1994: 204), a sufficient “size and coverage of the corpus” provided “that relevant patterns would emerge in the analysis” (Gabrielatos and Baker, 2008: 6). Within this corpus, the key word “Mesut Özil” occurred in 7065 instances (see Table 1).
Occurrences of the search term in the corpus.
Besides football-related contents such as appearances and performance in different teams, we found that concordances included the subjects of ethnicity, religious practices, and integration. To comprehend the extent and complexity of these topics, we then approached our data from the meso-perspective. To do so, “the researcher is normally required to analyse hundreds of lines of concordance data by hand, in order to identify wider themes or patterns in the corpus which are not so easily spotted via collocation, key word or frequency analysis.” (Gabrielatos and Baker, 2008: 6). For our case study however, instead of going through each concordance “by hand,” we decided to down sample (KhosraviNik et al., 2012) the number of by applying a Qualitative Content Analysis as the second step in our analysis. Thus, we were able to filter the large amount of concordance lines and focus only on those text units in which context and nuances seemed to be relevant to our research question, as they ascribed migrant identity and reproduced constructions of difference (Van Dijk, 1998). In fact, running a Qualitative Content Analysis filtered out 2382 of the altogether 7065 paragraphs of concordances as relevant to our research questions. As we were aiming for a “comprehensive and complete understanding” (Creswell et al., 2004: 8) of our data, down sampling the expanded concordances (Gabrielatos and Baker, 2008) allowed for a more detailed look at the data.
Qualitative content analysis
So far, a limited amount of research has been conducted applying a combination of Qualitative Content Analysis (QCA) with Critical Discourse Analysis (CDA). CDA “investigates language use beyond the sentence level” and “aims to denaturalize the role discourses play in the (re)production of noninclusive and nonegalitarian structures and challenges the social conditions in which they are embedded.” (Wodak, 2015: 1). Qualitative Content Analysis (QCA) is a mixed methods approach that combines quantitative and qualitative aspects of data analysis to describe qualitative data systematically (Schreier, 2013: 170). It reduces the amount of data to be analyzed in detail by assigning text units that are relevant to a specific research question to corresponding categories, thus allowing for the elimination of those data that do not speak to the research questions. The occurrence of these categories in the data is expressed in frequency, and the analysis then draws conclusions from such frequencies (Mayring, 2014: 19).
We found the combination of QCA with CDA an accessible and advantageous approach to our large data corpus as it proved to be a helpful symbiosis of patterns of frequency, meaning, and context (For studies that employed a mix of quantitative content analysis and discourse analytical methods, see e.g. Feltham-King and Macleod, 2016; Herrera and Braumoeller, 2004; Wilson, 1993). The method “reduces data, it is systematic, and it is flexible” (Schreier, 2013: 170) and proved to be reliable when handling large data sets, but most importantly served as the logical link between the corpus-linguistic approach (macro perspective) and a more in-depth critical discourse analysis (micro perspective). QCA provided a meso perspective on our data and subdivided the corpus into various text units that showed different levels of how migrant identities were ascribed in print media and how discourses on integration were reflected in such ascriptions. The method encourages researchers to follow a rather linear procedure of several steps that lead from the definition of the material to the interpretation of it (Kuckartz, 2014: 50; Mayring, 2014: 54; Schreier, 2012: 6). Thus, rather than only reflecting on frequency patterns, we were able to consider the situational context of such frequency patterns and thus combined both deductive assumptions and inductive findings emanating from discursive interaction.
To carry out a qualitative content analysis, we first designed a coding frame which is “at the heart of the method” (Schreier, 2013: 174). In a corpus consisting of a large number of texts, it is used to assign smaller text units to categories. These categories form the coding frame and reflect the theoretical framework guiding a research interest. In our study, categories derived from the theoretical framework were enriched by insights from the texts that we gained during the concordance analysis. An outcome of the concordance are lists of key words in context (KWIC). In COSMAS II, these KWIC lists can be accessed in a one-sentence format or a one-paragraph format. To pilot the consistency of our coding frame, we used one-sentence KWIC lists (a few words to the left and the right of a search term or key word respectively). Piloting a coding frame is essential, as in the main phase of coding, the coding frame is set concisely. When testing the coding frame, we considered it to be sufficiently extensive and manageable (Kuckartz, 2014: 121) once it could code all collocations of the search term “Mesut Özil.” At this stage, the piloting was completed and had contributed to the “methodological strength” (Mayring, 2014: 13) and reliability of our study (Feltham-King and Macleod, 2016: 6). We designed a coding frame consisting of 16 categories which were informed by the first findings we gained from our pilot study, indicating that although most of the text units containing the search term “Mesut Özil” would still contain football-related topics, newspapers texts were also reflecting societal and political discourses on the expectations related to successful integration within these texts. The 16 categories therefore reflected both topics related to football and those related to discourses on integration and migration and will be described in the following. Only text units related to the coding frame were coded.
Of all 16 categories, the first six ones were composed of descriptions of professional football playing skills and membership in professional football teams. Categories no. 7 and 8 covered the (non-) monetary value of football players. Categories 9 and 10 derived from the controversies around a specific discursive event, the photo of Mesut Özil and the Turkish President Erdoğan. Category no. 11 was assigned to attributions of Mesut Özil in context of his private and/or non-professional life. To identify text units in our data which reflected mentioning of ethnicity, we designed three categories that were coded for a socialization ascribed as being either Turkish (category no.12), or German (category no.13), or Turkish-German (category no.14). We then extended our coding frame to a category for text units describing religious, particularly Muslim practices (category no. 15) and to one which selected text units ascribing expectations regarding integration (category no.16).
The Qualitative Content Analysis assisted in filtering 7065 text units for occurrences of the search term “Mesut Özil” in relation to specific topics reflected in the categories which we described in the above section. Altogether, we coded 2083 of such text units. In our corpus of six newspapers covering a time period of 10 years, 33.71% of all newspaper articles mentioning “Mesut Özil” contained topics like expectations regarding integration, ethnicity, and religion besides football-related topics. The distribution and frequency of all 16 categories showed that the highest frequency was a coding of 601 times (25.23%) in a total of 2382 text units while the least frequent coding occurred 35 times (1.47%). Figure 1 gives an overview over the frequency of all categories as a result of our Qualitative Content Analysis.
Frequency of all 16 categories (figures refer to the total of 2382 text units containing the search term “Mesut Özil”).
As we had assumed, the category coding text units of Mesut Özil as a national football player occurred the most frequently (601 instances or 25.2% in a total of 2382 text units containing the search term “Mesut Özil”), followed by the category for football player skills (376 instances or 15.78% respectively). These categories capture attributions of professional aspects like competences, talent and career paths and indicate strong contribution to the success of a professional team and to the expectations of success and performance by those supporting the team. Furthermore, the frequent attribution of membership in a national team constructs the notion of national identity (Parr, 2015).
Coding the corpus with football-related categories for professional performance, career development, value, and performing for national/international/regional teams was helpful to understand the weighting of non-football related categories representing ethnicity, religion, and expectations regarding integration in comparison to that. The category for monetary value was represented quite frequently and coded over three times more often than the one for nonmonetary value (172 instances or 7.22% and 50 instances or 2.1% respectively), and the text units assigned under both categories contained a high number of metaphors. The category reflecting the controversies after the photo of Mesut Özil and Turkish President Erdoğan had been published was represented with a relatively high frequency of 238 instances (9.99%), but could not be included in further steps of the analysis here presented, as archiving the newspaper articles covering this event was not finished yet at the time we conducted our research. However, as a first reading shows that text units assigned to this category argue whether or not the fact that the photo has been taken impacts so-called successful integration processes, indicates that further research regarding this incident is called for.
Our analysis showed three remarkable results with regards to our research question about how discourses on integration are reflected in a selection of German newspapers. Firstly, the category that was coded for text units displaying expectations regarding integration was coded the third most frequently (291 times or 12.22% in a total of 2382 text units containing the search term “Mesut Özil”) and occurred more often than any ascriptions of ethnicity. Secondly, within the text units that described ethnicity, being of a Turkish-German ethnicity was attributed more often than being either Turkish or German. Thirdly, religious aspects in characterizations of a professional football player did occur.
To understand these results in a broader context, we conducted a similar analysis within the same corpus for two other football players who in print media are sometimes referred to as Polish-German. 7 The results showed that codings of expectations regarding integration only ranked third from last when it came to the text units describing Lukas Podolski and fourth from last with Miroslav Klose respectively. We also found that the attribution of being “Polish-German” was not significantly frequent for the other players. While Özil was born in Germany, Podolski and Klose were not. These findings indicate that in our study, the obligation to integrate is expressed more clearly in relation to people of Turkish ethnic heritage, and that the process of integration of “them” is under continual review by “us.”
Studies have shown that migration policies create hegemonies of difference (Flubacher and Yeung, 2016; Fuller, 2019; Morley and Taylor, 2012), and the analysis of our data indicates that these hegemonies of difference are reinforced by media. At the same time, grasping discourses of difference by only drawing on ethnicity did not seem to be sufficient in reflecting current discourses in our data, as “[n]ational identity emerges very much as a relational concept” (Hardt-Mautner, 1995: 179). By creating a category that we could assign to text units containing religious aspects, we could show that even if that was only the case for 1.97% (47 instances) text units, attributions of religion/religious practices, and traditions were present in newspaper articles mainly focussing on football.
Altogether, the Qualitative Content Analysis of our data helped to identify attributions that were as frequent as expected from media coverage of a football player (National football career, Football player skills), but it also showed that other attributions in the context of football (International football career, Regional football career, Position in the match, Development as a football player) were less frequent than attributions of expectations regarding integration and ethnicity.
Our first conclusions drawn from these results were that while the frequency of some codings occurred according to what we expected from newspaper coverage about a football player who competed in several world championships, the high frequency of the category coded for expectations regarding integration in comparison to other categories stood out. To gain a more in-depth understanding of how such expectations regarding integration are realized on a linguistic level, we focused on a Discourse-Historical Analysis of only these text units in the next step of our analysis.
The discourse-historical approach and discursive strategies
Discourse-Historical Analysis (DHA) belongs to the field of Critical Discourse Analysis and was first developed for an analysis of stereotypes in public discourses in Austria (De Cillia and Wodak, 2018; Wodak, 2015; Wodak et al., 1999, 2009). DHA shares methodological principles with other approaches subsumed under CDA (Wodak, 2015: 2), however, DHA is distinct in that “it should make the object under investigation and the analyst’s own position transparent” (Wodak, 2015: 3). In positioning our case study within a larger research project about discourses on integration and migration, we have displayed our stance toward, or within this discourse. We hope to contribute to identifying ideologies and concepts that “stay largely unchallenged” (Wodak and Meyer, 2009: 8) on the one hand, but on the other hand impact public discourses and become “agenda setters” (Vollmer and Karakayali, 2018: 119).
Obligations and expectations regarding integration position an individual toward ideologies in the context of integration. At the same time, they serve as a benchmark as to whether or not, or to which extent, an individual should be considered as being integrated. Individuals, groups, or institutions positioned toward discourses become social actors (Reisigl and Wodak, 2009; Van Leeuwen 1996, 2008), and are part of “asymmetric relationships.” Within these unbalanced relations, social actors “assume different social positions or belong to different social groups.” In other words, the discursive positioning of social actors is related to hierarchy and power (Ahlstrand, 2019: 10). Through the use of language, the positioning of social actors within hierarchical power structures is justified or legitimized in texts. A critical analysis of discourses thus needs to examine “[b]y means of what arguments and argumentation schemes do specific persons or social groups try to justify and legitimize the exclusion, discrimination, suppression and exploitation of others?” (Wodak, 2001: 73). Reisigl and Wodak (2001, 2009) conceptualized a range of argumentation strategies (for an overview including objectives and linguistic devices, see Reisigl and Wodak, 2009: 94). The aim of the following section is to display a selection of such strategies with examples from our corpus in order to gain a better understanding as to how language legitimizes, and thus standardizes “the social structures, cultural norms and physical legacies that discourse occurs within” (Kwon et al., 2014: 6).
Through the Qualitative Content Analysis, we could identify what kind of attributions occurred in the context of the search term “Mesut Özil.” Now, we move on to examine how expectations, responsibilities, and duties are introduced as dominant discourse topics. As mentioned in the previous section, we found 291 instances in which expectations regarding integration were expressed. Table 2 shows a selection from these 291 text units and the embedded discursive strategies, such as nomination (example 1), perspectivization (example 2), intensification and mitigation (example 3), and predication (examples 4-11). The social actors most frequently constructed were Mesut Özil, a “we-group” (Reisigl and Wodak, 2009: 114) realized for example through personal pronouns like “us,” and other relevant individuals in comparison to the player, who depending on the context of the utterance changed between teammates or politicians. Furthermore, the debate on integration itself, the acknowledgment of his successful integration, his integration process and religion were nominated as discursive events, processes, or phenomena the respective text unit focused on. As example no. 1 shows, social actors are both nominated and then attributed in comparison with each other and their migration experiences. Through perspectivization strategies, discourses about integration are represented in certain ways, which in our case study often problematizes integration and those in the process of it. The perspective implied in example no. 2 of the table below suggests that discussions around the integration of people with a so-called Turkish background would be more heated in comparison to others. Furthermore, we found that intensification and mitigation strategies modified utterances in the context of our search term “Mesut Özil,” either by weakening or by strengthening them. In example no. 3, a football match between Germany and Turkey that took place in Berlin in 2010 is framed with references to different debates on integration and migrants in Germany. What the match is supposed to be about is mitigated in terms of winning the match and intensified in terms of understanding the relevance of such a football encounter within the frame of current discourses. Besides the aforementioned discursive strategies of nomination, perspectivization, intensification, and mitigation, we found strategies of predication to be frequently distributed throughout our data. Predication strategies aim to qualify utterances or social actors, for example by attributing negative or positive qualities to them. In the present case study, identifying these strategies in particular helped to display how attributions reflect dominant discourses about integration in Germany.
Coding frame of 16 categories with application rules (summarized) and examples.
The examples show utterances that were in vicinity to the search term “Mesut Özil,” and that were then considered for further analysis. Examples 4, 5, 9, and 11 do not display the actual search term, because it did not occur in direct vicinity to the utterance considered for further analysis. Example 7 displays only one part of the search term (“Özil”), because it occurred as a repetition of the search term “Mesut Özil.”
The table integrates two methodological steps: text units containing expectations, duties or rights selected by means of the Qualitative Content Analysis, and discursive strategies that justify or legitimize the obligations and expectations related to integration, as identified through the Discourse-Historical Analysis approach.
Özil is categorized as part of a group, the football team, and then positioned toward other groups, relevant actors, and ongoing political debates. At the same time, even in articles about athletic performances in a football match, constructions of political and societal debates on integration occur as recurring or impacting discursive events juxtaposed to that match and the individual player(s) involved. Hence, social actors in the text are characterized against such debates and qualified in comparison to them. These predication strategies attach notions of quality.
Compared to the ongoing debates on integration, Özil is also attributed as being excited about being rewarded for his successful integration (Spiegel Online, 2010-11-12), while his role of being responsible for integration is qualified as one of a “secret integration commissioner” (Süddeutsche Zeitung, 2010-10-16), who is of symbolic value (Die Welt, 2010-10-21) for the Chancellor, even when “half-naked.” Again, even when attributed rather positively, his responsibilities for his own integration as well as the expectations of him being a role model for the yet to be successfully integrated “others” are modified by strategies that on the one hand intensify his role in society and on the other hand mitigate his athletic performance in reports about his participation in football matches. Then, “it would be naïve to think of it [the match] without a debate on integration attached” (Süddeutsche Zeitung, 2010-10-8).
The Qualitative Content Analysis displayed that the attribution of being a national football player was the most frequently coded one. Integration policies imply an expectation of professional participation (Henkelmann, 2012) and professional identities are formed by participation and performance in professional communities (Stryker and Burke, 2000). In our data, the professional anthroponyms (“football player”) carry notions of national identity, mostly bound to the national identity of a football team. Hence, the attribution of both professional and national membership assigns membership in two groups. Through qualifying professional anthroponyms by notions of nationality, attributions of being a member in the German national team function as a framework (“Ordnungsraster”: Parr, 2015: 3) against which the stage of the integration process of Özil can be measured.
Furthermore, debates on integration have always also highlighted language competency as an important factor in integration frameworks (Flubacher and Yeung, 2016: 601) as is born out in the establishment of testing regimes for migration and citizenship (e.g. Möllering, 2010). Language becomes a “symbol of shared social norms and values and indicates a willingness to integrate” (Fuller, 2019: 322). In our data, the announcement of Özil’s resignation from the national team is predicated as neither being in German nor in Turkish, although the matter of his resigning would be a German-Turkish one (Spiegel Online, 2018-6-8), implying that the language chosen to publish his resignation (English) was not the appropriate or expected one. Again, Özil and even his resignation seem to not meet expectations and the discussions about him instigate another debate on integration in Germany (Der Tagesspiegel, 2018-7-25).
Discussion
The study presented in this article used a triangulation of Corpus Linguistics, Qualitative Content Analysis, and Discourse-Historical Analysis to present a comprehensive understanding of how discourses on integration are reflected in newspapers in a particular case. Our interdisciplinary methodological approach assisted in identifying and critically examining conceptualizations of belonging, descent, and constructions of ideologies of power (Baker et al., 2008; Van Dijk, 2009; Wodak and Meyer, 2009). Although Van Dijk (2009) states that rather abstract ideologies are not necessarily linked to discourses per se, they can “assist to structure groups, memberships, and relations between groups and individuals” (p. 79), which we showed in the analysis of nomination and predication strategies in the case of Mesut Özil (Table 3).
Selection of discursive strategies.
As the results of both the Qualitative Content Analysis and the Critical Discourse Analysis show, there is a divergence in medial reflections of descent and belonging on the one hand, and of expectations toward integration on the other hand. While being attributed with features that according to the dominant discourse on integration (Fuller, 2019: 335) make it impossible to “achieve” Germanness, text units assigned to attributions of ethnicity were not the most frequent ones. Attributions of local belonging in the Ruhr area were coded less frequently in the case of Mesut Özil than for Lukas Podolski, although the former was born in that area, whereas Podolski migrated to Germany as a young child. Our findings clearly show that ethnicity does not solely serve as a marker for ascriptions of national belonging and integration, and that in the German context, expectations regarding a (successful) integration performance dominate.
In our findings, cultural practices were of a higher relevance for “assessments of success of integration in Germany” (Fuller, 2019: 324). These practices however, come with obligations that are to be fulfilled by those in the processes of integration (Fuller, 2019: 324). Integration then becomes a consequence of “expected efforts” (Flubacher and Yeung, 2016: 612). Frequently constructed responsibilities of becoming a so-called successfully integrated German were clearly describing what that would mean and when this goal would be achieved. Duties that arise in the process of integration, as well as the discursive construction of a migrant identity were assigned to Mesut Özil, reflecting that despite legal and societal changes regarding debates on integration in Germany, media discourses encourage a picture of integration as an ongoing process subject to ideologies about migrants and cultural practices of “the others.” Othering in our data was also constructed by attributing religious aspects to football-related context. Fortier (2012: 32) demonstrated a shift toward the fact that “cohesion, and national integrity become projected onto bodies and onto cultural markers (e.g. the skull cap or the hijab).” Fuller (2019: 335) in her discourse analysis of comments in German newspapers notices “practicing Islam” becomes a stereotypical reason to depict “Muslims/Turks as the unintegrated Other.” Muslim migrants have mostly been stigmatized as an intimidation to a host community (Ahmed and Matthes, 2017; Baker et al., 2013; Dunn, 2001).
In our analysis, we found that most predications of Özil characterized him having “migrant story” (Die Zeit, 2012-6-14), as living between cultures and nationalities (Focus, 2009-7-6), as living in two worlds (Focus, 2017-4-8) and denying to sing the national anthem (Der Focus, 2014-6-16; Die Welt, 2016-6-8). In these examples, discursive strategies served to construct Özil as struggling with expectations regarding integration, and as being in the process of a not yet successful integration. This important finding speaks to a tendency for negative representations of migrants in print media which reinforces ideologies of prejudices (e.g. Gabrielatos and Baker, 2008) and is highly problematic, because of the influence of (print)media on reinforcing discourses (Duffy and Rowden, 2005; Van Dijk, 2000). Such discourses are produced in a framework of beliefs that are shared by a majority of people and “circulated” through media (Vollmer and Karakayali, 2018: 122). Debates about migration and integration are shaped by “wider social understandings of difference” (Fortier, 2012: 31) and a migrant “other” is positioned against a “‘national standard’ of norms and values” (Bigo, 2002: 67). In the course of Germany’s development toward a country of immigration, discourses on integration have added to these widely accepted assumptions of differences, for example in the practice of addressing people with their so-called migration background. By constructing and reinforcing migrant identities, the newspapers we analyzed contributed to a growing sense of crisis (Musolff and Viola, 2019) in identities that are still posited as being related to alterity and difference (Hastings and Manning, 2004). Integration is constructed as impossible to be achieved on the one hand, while on the other hand there is a strong claim for the need to integrate. At the same time, those expected to integrate are being othered. Integration becomes an irresolvable dilemma. Media re-enforces ideologies that standardize integration as a one-way street and thus shifts the responsibility for integration to the subject rather than to society (Flubacher and Yeung, 2016).
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.