Faculty of Behavioural, Management and Social Sciences (BMS) European Public Administration (EPA)
University of Twente, Enschede Module 12 (3.4) Bachelor Thesis
Bachelor Circle: Managing the Migration Crisis First supervisor: Dr. Ringo Ossewaarde Second supervisor: Prof. Dr. Ariana Need
Pride and Prejudice:
The Proliferation of Nationalism and Islamophobia in the context of the European Refugee Crisis
A Discourse Analysis of National Online Media
in Austria, Germany, Switzerland and the United States between September 2015 and March 2016
Bachelor Thesis by Neda Laura Herrchen
Public Governance Across Borders (B.Sc.) Westfälische Wilhelms-Universität Münster
Date: June 30th, 2016
_______________________________________________________________________________
Abstract
The current so-called refugee crisis in Europe is associated with an increase of anti-migrant sentiments, which are often manifested in islamophobic rhetoric and an anti-Muslim climate.
Furthermore, a resurgence of nationalism is claimed in the context of the so-called ‘Brexit’. In light of these developments, the following research question is crucial: How can varieties of the islamophobic discourse in national online media be explained according to country-specific modes of nationhood and citizenship in the context of the refugee crisis since September 2015? The concept of islamophobia is frequently used in social sciences. Nonetheless, its definition is somewhat divisive depending on the prevailing national context and on the adopted perspective of the media evaluated. Thus this study focuses on identifying the prevalent national patterns of islamophobia for the selected countries Austria, Germany, Switzerland and the United States and links these results to country-specific modes of nationhood and citizenship. A discourse analysis of online articles by national newspapers is conducted to assess varieties of the islamophobic discourse as well as underlying meanings communicated by media since islamophobia is often concealed in distinct rhetoric means and formulations. The results of this study are of relevance as understanding the facets of and the factors behind islamophobic sentiments are crucial to avert growing political divergence and right-wing populism. The analysis shows that different socio- structural modes to nationhood and citizenship are stimulating different islamophobic patterns.
Thus, the expected linkage between prevalent islamophobic varieties and modes of nationhood and citizenship is observed.
Keywords: islamophobia, anti-Muslim, refugee crisis, migration crisis, nationhood, citizenship, national identity, nationalism, discourse analysis, media, newspaper, populism
_______________________________________________________________________________
Index
1. Introduction ... 1
2. Theoretical Framework ... 5
2.1. Varieties of the Islamophobic Discourse ... 5
2.2. Modes of Nationhood and Citizenship ... 8
2.3. The Role of Media ... 10
2.4. Concluding Remarks ... 11
3. Methodological Framework ... 12
3.1. Case Selection ... 12
3.2. Data Collection ... 12
3.3. Data Analysis ... 13
3.4. Concluding Remarks ... 17
4. Discourse Analysis ... 18
4.1. Cross-country Comparison of the Varieties of the Islamophobic Discourse ... 18
4.2. The Islamophobic Discourse and Modes of Nationhood and Citizenship ... 26
4.3. Concluding Remarks ... 29
5. Conclusion ... 31
References ... 34
Appendix A ... 38
Appendix B ... 53
Notes ... 70
Appendix C ... Extra File
List of Tables
Methodological Framework
Table 1 Coding Scheme – Thematic Varieties of the Islamophobic Discourse...15
Table 2 Discourse Analytical Concept...16
Discourse Analysis Table 3 Prevalence of Islamophobic Varieties...26
Appendix A Table A1 List of pre-selected and selected Newspaper Articles...38
Table A2 Exemplifying Phrases for the Coding Scheme...50
Table A3 Translation of the Key Words in the Coding Scheme...51
Table A4 Translation of the Coding Scheme...52
Appendix B Table B Exemplifying and Important Quotes from the Discourse Analysis...53
1. Introduction
“Until there is absolute, clear evidence, every headline doesn’t have to say ‘Muslim massacre’ or
‘Muslim shooters’, because it’s going to cause intolerance” (Goodstein, 2015, p. n.a.).
The political upheavals of the ‘Arab Spring’ and other destabilizing factors in the Middle East, such as the civil war in Syria and the emergence of the so-called Islamic state dramatically changed the European migration context (Park, 2015). Besides these factors, poverty and suppression in other countries intensified the numbers of refugees into an unequalled scale in Europe with complex and rapidly changing dynamics (Hampshire, 2015). In this context, populism against migrants, refugees and minorities recently established itself throughout Europe in nationalist and right-wing parties opposing further immigration, all of which focus on anti-Muslim and anti-Islam sentiments (Park, 2015). These sentiments are used by right-wing parties as instruments to mobilise the public and obtain support and votes during elections in Europe (Allievi, 2012). The success of populist parties is evident, e.g. the Front National in France, the SVP in Switzerland or the FPÖ in Austria. In Germany the islamophobic movement Patriotische Europäer gegen die Islamisierung des Abendlandes (PEGIDA) and the populist party AfD stress the recollection and preservation of shared ethnic and cultural identity (Pausch, 2015; Pfahl-Traughber, 2015). All European right-wing parties applauded the so-called ‘Brexit’ in June 2016, when the British citizens voted to leave the European Union (EU). They claim the ‘Brexit’ a triumph of anti-EU and anti-immigrant positions and call for similar referendums in France, Denmark and the Netherlands (Chrisafis &
Oltermannin, 2016). These political extremes and a general dissent about the role of Islam emerged throughout Europe, e.g. “whereas Merkel declared forcefully that Islam was also a religion of Germany, some in Eastern Europe have declared that they will welcome only a small number of refugees – and only if they are Christian” (Dervis, 2015, p. n.a.).
Such political and social exclusions are neither limited to Muslims (e.g. persecution of Christians in Syria, Iran etc.) nor new phenomena. Anti-Muslim and anti-Islam narratives circulated for decades and increased in particular since the 9/11 attacks in 2001 (Barbero, 2012;
Kunst, Sam, & Ulleberg, 2013). In addition, the multi-faceted, so-called refugee crisis is stoking islamophobia (SETA, 2016). The term itself was first described in 1997 as “a shorthand way of referring to dread or hatred of Islam – and, therefore, to fear or dislike of all or most Muslims”
(The Runnymede Trust, 1997, p. 1). In literature a general upsurge of islamophobia is described in Europe and the United States of America (US) (Ciftci, 2012; Dolezal, Helbling, & Hutter, 2010;
Ogan, Willnat, Pennington, & Bashir, 2014; Pratt, 2015). Although enmity against foreigners in general has decreased, negative attitudes against Muslims and Islam are growing since the 1990s (Meer & Modood, 2015). Whereas the first generation of the so-called guest-workers claimed mainly social and political rights, subsequent generations and Muslim immigrants now advocate
their religious and cultural claims. This perception of growing presence of Islam in ‘Western’1 states intensifies conflicts between immigrants and citizens of the host countries. As a consequence of increasing islamophobia any differentiation between moderate, radical or fundamentalist Islam, or even terrorism is ignored (Dolezal et al., 2010). Islamophobic sentiments are in general driven by and reactive to key events (Borell, 2015), for instance, both terrorist attacks in Paris in January and November 2015 impinged on the public debate regarding Islam and Muslims in Europe (SETA, 2016). In Europe and the US, this public discourse on Muslims and Islam is largely framed by the media (Ogan et al., 2014).
Apart from political developments on the right of the political spectrum, islamophobia also seems to have reached the societies’ mainstream (Altikriti & Al-Mahadin, 2015). A study by Creighton and Jamal (2015) investigates the relationship of Islam and anti-immigrant sentiments and identifies negative implications for the integration of Muslim immigrants because explicit hostility against Muslims is more freely expressed than against other immigrant out-groups. The image of Muslims as an un-integrable out-group is in place as “the closest a ‘Muslim’ may come to the speaking ‘us’ is to be ‘like us,’ where the inserted ‘like’ maintains the critical separation between ‘them’ and ‘us’” (Gardell, 2014, p. 133). This division emphasises the role of nationhood and citizenship in the context of islamophobia as these concepts also define out-groups in the national population. In literature it is argued that in European countries immigration is especially complicated due to their particular conceptions of nationhood, which stress in contrary to the US mainly ethnicity instead of civic aspects (Park, 2015).
False pride and persistent prejudices, as in Jane Austen’s novel, determine social barriers – may they be imaginative or real – between people of different origin. Whereas in the novel these barriers are overcome, in socio-political reality, pride and prejudice sometimes persist and lead to proliferation of nationalism and segregation. In this context, i.e. the increase of nationalist, populist and right-wing movements as a reaction to the arrival of refugees, the following research question is studied:
How can varieties of the islamophobic discourse in national online media be explained according to country-specific modes of nationhood and citizenship in the context of the refugee crisis since September 2015?
The research is based on an explanatory question and is examined by a qualitative discourse analysis of national online newspaper articles. The analysis is expected to detect national thematic characteristics of islamophobic patterns in online media for the selected countries, namely, Austria, Germany, Switzerland and the US. The media debate on islamophobia from September 2015 to March 2016 is reconstructed for each country. Besides the thematic varieties of the islamophobic discourse, the image of Muslims and Islam created by the media is structurally assessed considering the used rhetoric. This approach is vital to evaluate the role of the media in the upsurge
1 The term ‘Western` must be critically reflected as it is sometimes upholding the (cultural) distinction between the
‘West’ and the ‘East’ or the so-called ‘Orient’. In this thesis, the term is referring to the mere geographical meaning.
of islamophobic attitudes and to examine the underlying image created in the structure of the articles. More precisely, the research question has two central lines of argument:
a. The islamophobic discourses in Austria, Germany, Switzerland and the US fuel the marginalisation of Muslims and Islam by circulating stereotypes, prejudices, anxieties and racialization with ‘charged’ formulations and rhetoric in the media.
b. Country-specific modes of nationhood and citizenship are expected to generate a national prevalence of islamophobic patterns creating differences of the islamophobic discourses between the selected countries.
In addition to the main research question, two sub-questions are introduced to reflect these dimensions of the two lines of argument, which are examined in the fourth part of this thesis to derive an answer to the overall research question in the conclusion:
a. Which islamophobic patterns are prevalent in the national discourses?
b. Which islamophobic patterns detected relate to the country-specific modes of nationhood and citizenship and which patterns challenge the anticipated relationship?
The recent vote of the UK citizens to leave the EU underlines the scientific relevance of this study as the “Brexit vote coincides with a rising nationalism in the rest of Europe and [...] in the United States, exemplified by Donald Trump” (Lowry, 2016, p. n.a.). This emphasises the current importance of the modes of nationhood and citizenship. Whereas in 1996 Helmut Kohl claimed that the great problems of the 21st century cannot be solved by states, today, 20 years later, reverse sentiments and political relapses are striking throughout Europe favouring autonomy instead of reliance on the European community (Lowry, 2016). This marks a step back in the process of European integration. In light of this resurgence of nationalism and the current so-called European refugee crisis, the link between islamophobia and modes of nationhood and citizenship indicates a scientific gap requiring further investigation. Even political responses to the so-called refugee crisis seem to be in substance self-governed, i.e. actions taken by individual states. For instance, the German government surprised in September 2015 with the ‘open door’ policy allowing Syrian people a refuge despite the EU Dublin regulation and Angela Merkel’s confidence in the country’s ability to cope with the crisis (Benner, 2016; Delcker, 2015). The examination of national islamophobic discourses by means of a newspaper analysis is necessary due to these current developments: the study is expected to add scientific knowledge about anti-Muslim sentiments by giving an explanation for the national predominance of specific islamophobic patterns. This study adds a cross-country analysis and comparison of the media discourses on Islam and the marginalized out-group Muslims to scientific literature.
Concerning the social relevance of this study, the understanding of patterns and factors of islamophobia may help to facilitate the progress of integration by identifying communicative obstructions and to omit possible, latent or explicit prejudices against Muslims and migrants.
Although the phenomenon of islamophobia is steadily growing, it is lacking recognition, political responses and preventive measures by the EU leaving people affected by islamophobia marginalised (Cerulus, 2014). Immediate political actions are needed as attitudes against migrants and Muslims are to some extent intertwined in these marginalising trends (Creighton & Jamal,
2015) and a lowered inhibition threshold for using violence against refugees, asylum seekers or asylum shelters is registered (Kroet, 2016). The research question is also of relevance to a problem- focused coping with the so-called European refugee crisis as the quota of Muslims in non-Muslim majority states continues to rise and the growing migratory pressures are generating fear in Europe (Ogan et al., 2014). Islamophobia leads to growing polarisation of political attitudes and party fragmentation making political responses on a national and international level even more complicated. Latter developments obstruct the implementation of socially controversial policies, e.g. measures concerning the allocation of refugees (Henley, 2016). Understanding the motives behind islamophobic sentiments is crucial to avert growing political divergence and regain transnational and European solidarity.
In the following part, the theoretical framework for the research question is established in regard to the existing literature. After introducing the methodological approach of the discourse analysis and the appertaining coding scheme in part three, the analysis itself is conducted in part four. Within the newspaper analysis, the thematic varieties of the national islamophobic discourses as well as the image created by the media are examined in a cross-country comparison. The relationship between these varieties and the modes of nationhood and citizenship is analysed to arrive at a conclusion on the characteristics of the contemporary islamophobic discourse and its relation to the modes of nationhood and citizenship.
2. Theoretical Framework
As the research problem revolves around two issues, namely, the varieties of the islamophobic discourse and associated linguistic and rhetorical means as well as the link to modes of nationhood and citizenship, the theoretical framework is divided into two main parts. Since one focus lies on the varieties of the islamophobic discourse, the theoretical framework aims at covering and conceptualizing a broad scope of thematic patterns of islamophobia. A comprehensive concept of islamophobia is developed by examining relevant scholarly literature and by categorizing existing conceptualizations of the term islamophobia according to specific thematic patterns. Hence, the first part of this theoretical framework distinguishes three different patterns of the islamophobic discourse, namely, (1) stereotypes and prejudices, (2) anxieties and (3) racialization including their sub-concepts. In the second part the other focus of the study is conceived – modes of nationhood and citizenship – as a possible explanation for varieties of islamophobic sentiments. Additionally, the role of media is highlighted in the third part as the methodology relies on a discourse analysis of online newspaper articles and on the examination of linguistic and rhetorical means communicated in the media.
2.1. Varieties of the Islamophobic Discourse
The concept of islamophobia is frequently used in social sciences (Bleich, 2011; Cheng, 2015;
Ekman, 2015; Hafez, 2014; Lathion, 2015; Pratt, 2015). Nonetheless, a precise definition is not in place. According to Gardell (2014, p. 133), “in the concepts ‘‘xenophobia’’ and ‘‘homophobia,’’
the suffix ‘‘phobia’’ is used in a transferred sense to connote socially, culturally, and politically produced prejudice, aversions, and discrimination against specific categories of humans, in this case Muslims”. The term’s focus varies significantly between the topics of fears, prejudices and stereotypes, discrimination and racism in regard to Muslims and Islam (Kunst et al., 2013). This study aims at identifying the thematic range of islamophobic tendencies by examining the debate in each country. The inclusion of differing definitions of islamophobia is necessary to depict variances more precisely. The concepts are categorised into three major patterns, i.e. stereotypes and prejudices, anxieties, racialization, and related sub-concepts to acquire a broad range of the varieties of islamophobic sentiment.
(1) A first main pattern of the thematic varieties of the islamophobic discourse consists of stereotypes, which means fixed and oversimplified perceptions of a person or a group, and prejudices, such as irrational negative attitudes and enmities against a person or a group and the associated characteristics. One prejudice often described in literature is the perception of Muslims as a demographical hazard due to alleged high birth rates and immigration rates into ‘Western’
countries (Ekman, 2015). Islamophobia is to some extent based on and entangled with the older conception of orientalism (Bleich, 2011). The conception of orientalism summarizes the ‘Western’
perception of the ‘East’ as backwards, irrational and inferior (Barbero, 2012). Muslims and Islam are stereotyped as being culturally backwards, non-progressive and responsible for crime increases as well as accused of spreading hostility towards women (Cheng, 2015; Elchardus & Spruyt, 2014).
In this context, suspicion and rejection of Islam and Muslims is often based on generalised assumptions and biases, e.g. the indiscriminate conviction, that Islam and Muslims are inherently violent (Kumar, 2015). While orientalism situates the cultural otherness to the geographical
‘Orient’, islamophobia addresses the otherness within the ‘Western’ societies creating insecurity- related stereotypes and prejudices, i.e. “modern Islamophobia is preoccupied with the idea that Muslims pose an inner threat to the ‘West’” (Ekman, 2015, p. 1989). This assigned inner threat in combination with terrorist events such as 9/11 or the London bombings reshape the perception of Islam to be intertwined with terrorism and by character radical, which inevitably leads to a sense of increasing insecurity within the ‘Western’ societies (Meer & Modood, 2015).
(2) A second major pattern of the varieties of the islamophobia discourse revolves around islamophobic anxieties, i.e. not only prejudices and stereotypes are prevalent, but rather those existing generalised patterns of anti-Muslim sentiments generate indiscriminate cultural, social or political fears. When fears and anxieties concerning a fundamentalist and extremist form of Islam are generalised to all Muslims and Islam as such, islamophobia ensues (Pratt, 2015). Kinnvall (2015, p. 522f) argues that “visions of a Muslim takeover have led movements on the far right [...]
to increasingly stress Christian identity themes [...] [describing Islam] as a religion of fanaticism and intolerance, incompatible with democratic values and Western culture”. Hence, the pattern of anxieties is subdivided to include different theoretically relevant dimensions. In the following, cultural, social and political aspects are introduced:
Cultural anxieties – This pattern revolves around the fear of cultural deterioration in accordance with the perception of multiculturalism leading to the Islamisation of the ‘West’ (Ekman, 2015), that is a perceived shift of the society towards Islam. Hafez (2014) argues that the main distinction criterion determining identity and the affiliation to a group has changed from being race to being culture with a special focus on religion. Thus, not the ‘unknown’ is feared, but the perceived
‘otherness’ induces suspicion and anxiety (Lathion, 2015). This stresses the perception of the inability of Muslims to integrate into the ‘Western’ culture. Still, Lathion (2015) claims that the prevalent and known practices of Islam can in general be integrated into the European legal framework, which emphasises the arbitrariness of anti-Muslim attitudes as a result of cultural anxieties. Additionally, a ‘resurgence of religion’, i.e. an increased significance of religion in the public and private sphere (Jackson & Sørensen, 2013), as well as demands for a conformity of cultural norms are identified (Kofman, 2005). Both could eventually lead to (further) segregation, ghettoisation and integration problems of Muslims.
Social anxieties – Here the term islamophobia focuses on the aspect of social fear towards Islam and Muslim cultures (Kunst et al., 2013). Bauman (2015) stresses the dimension of social anxieties by depicting the fear of social decline as an emerging and exceptional factor of anti-migrant
sentiments reaching beyond a general distrust of strangers meaning that resentments and hostilities against refugees, migrants and asylum seekers result from anxieties about social standing and downward social mobility. This is also relevant in the context of anti-Muslim attitudes as anti- migrant and islamophobic patterns are to some extent entangled. The social anxiety pattern also involves the fear of the emergence of parallel societies and Muslim self-segregation by a growing presence of Islam within ‘Western’ societies (Phillips, 2006).
Political anxieties – Islam is accused of being a totalitarian political ideology creating political anxieties, especially based on concepts such as ‘indoctrination’ or ‘infiltration’ by Islamic law and the sharia (Arjomand, 2004; Ciftci, 2012). Cheng (2015, p. 582), for instance, states that in islamophobic discourses “Islam is considered a religion with only troubling aspects that has caused problems in Western society or will cause problems in the future if not tightly restricted and prevented from ‘spreading’”. However, Islam is not considered a problem in the sense that it demonstrates a threat to Christian hegemony in the ‘Western’ sphere, but rather that Islam impinges on the political and legal order of the ‘Western’ societies (Cheng, 2015).
(3) Racialization is the third pattern of islamophobic varieties. In general, islamophobia can broadly be defined as racism against Muslims and Islam. However, in this thesis another focus is chosen, namely, the ‘racialization’ of Muslims as described in literature (Meer, 2014; Sayyid &
Vakil, 2010). Muslims are perceived as a race instead of a religious community generalising all islamophobic stereotypes and prejudices as well as anxieties towards this religious group to stress Islam’s and Muslims’ alleged inferiority:
[Although] Muslims transcend categories of ‘race’, nationality, ethnicity and skin colour, because, technically, anyone can choose to become Muslim [...], there is strong evidence that Islam is conflated with a ‘non-white’ religion, and Muslims are racialised depending on the particular national context of the country where they reside [e.g. in Germany as Turks or in the United States as Arabs]. (Cheng, 2015, p. 563)
Thus, Muslims and Islam are homogenized and debased by the islamophobic discourse, i.e.
inherent cultural traits are depicted as preventing integration (Kofman, 2005). This generalisation is also described by Garner and Selod (2014, p. 4), i.e. “people read Muslim-ness onto individuals by using a combination of ideas about culture and appearance”. For instance, generalised characteristics such as the origin or particular complexions are taken as indicators of the homogenized Muslim community and Islam (Cheng, 2015; Garner & Selod, 2014).
To sum up the first line of argument, islamophobia is expected to manifest itself in different patterns within national debates on Muslims and Islam creating distinct varieties of islamophobic discourses. In addition, the debates in the selected countries are anticipated to vary in regard to the national prevalence of islamophobic patterns, which are in accordance with the presented theoretical framework (1) stereotypes and prejudices, (2) anxieties and (3) racialization as well as associated sub-concepts. These thematic varieties are examined by a discourse analysis to detect national islamophobic key patterns and to link these to the explanatory factors of country-specific modes of nationhood and citizenship.
2.2. Modes of Nationhood and Citizenship
Besides the concept of islamophobia, this study focuses on country-specific modes of nationhood and citizenship to assess their role in the national islamophobic discourses. This is essential as nationhood and citizenship influence the relationship between majority and minority groups in society (Dolezal et al., 2010). Existing literature identifies a close connection between national identity, religion and culture (Meer & Modood, 2015; Suvarierol, 2012). In consequence, it is expected that Muslims – generally a minority in ‘Western’ countries – and their relationship to the host countries’ population can be explained by distinctive modes of citizenship and understandings of nationhood. According to Brubaker (1992, p. 182), the concepts of nationhood, national identity and citizenship are closely related because “proposals to refine the legal criteria of citizenship raise large and ideologically charged questions of nationhood and national belonging”.
Apart from the association with orientalism, islamophobia can be embedded within the context of xenophobia, i.e. the general fear or hatred of foreigners, or ethnocentrism, which depicts the superiority of one’s own nation or (ethnic) group (Taras, 2013). Furthermore, nativism can be an islamophobic bias involving an ideology which claims that the state is exclusive for the native group, while non-native qualities are regarded as threats (Kinnvall, 2015). All three concepts relate to the formulation of ‘us’ and ‘them’, i.e. the distinctive construction of out-groups unable to integrate into national society. In general, the demarcation between in- and out-groups is anchored in specific characteristics, for instance, common values, history, language, customs or nationality (Martiniello, 1995). The European Report on Intolerance, Prejudice and Discrimination also claims that Muslims are often not considered an integral part of society, but remain migrants and foreigners despite their actual citizenship or place of birth (Zick, Küpper, & Hövermann, 2011).
Ignatieff (1994) argues that nationalism has three dimensions, i.e. political, cultural and moral claims. As a political claim, nationalism covers the self-determination of states, while it entitles to national self-defence as a moral claim. The cultural idea of nationalism defines the nation as the primary source of belonging. Nationalism has special relevance for this thesis and is primarily understood as a cultural claim, i.e. a way to define a sense of belonging and national identity. In the context of nationalism as a cultural claim, Ignatieff (1994) introduces two types of nationalism – civic and ethnic. In the following, both different modes of nationalism as well as different citizenship concepts will be elaborated. These modes are assigned to the selected countries according to previous literature and established scholars instead of collecting new data on the modes of citizenship and nationhood for each country:
Civic mode – The concept of civic nationalism claims the significance of law and citizenship instead of common roots in regard to the belonging to a nation:
[...] the nation should be composed of all those – regardless of race, color, creed, gender, language, or ethnicity – who subscribe to the nation’s political creed [...] [and] envisages the nation as a community of equals, rights-bearing citizens, united in patriotic attachment to a shared set of political practices and values. (Ignatieff, 1994, p. 6)
Scholars argue that citizenship, national identity and nationhood are addressed in a universalistic and civic sense in the US (Ignatieff, 1994; Joppke, 2007; Koopmans, Statham, Giugni, & Passy, 2005). Citizenship in the US is attained unconditionally at birth – according to the so-called jus soli or right of the soil – or through naturalisation (Gilbertson, 2006). In general, the USA is considered a classical immigration state, while in Western European states immigration developed rather involuntarily and against the preference of the citizens and political elite (Koopmans et al., 2005).
Ethnic mode – In contrary to civic nationalism, the ethnic type emphasises language, culture and religion claiming “an individual’s deepest attachments are inherited, not chosen. It is the national community that defines the individual, not the individuals who define the national community”
(Ignatieff, 1994, p. 7f). This approach is also reflected in citizenship law: citizenship in Germany, Austria and Switzerland is described in generally more ethnic-cultural terms (Brubaker, 1992;
Dolezal et al., 2010; Ignatieff, 1994). The citizenship in Germany was formally formulated by the so-called jus sanguinis – describing the right to nationality based on parentage – depicting a more ethno-cultural and identity-focused approach to nationhood and citizenship (Brubaker, 1992;
Martiniello, 1995). Within this approach citizenship exceeds a mere legal concept of citizens’
rights and derives at a sense of belonging to a community and emotional identification (Cesarani &
Fulbrook, 1996). Since the year 2000, a mixture of jus sanguinis and jus soli is introduced in Germany allowing a more liberal approach to the acquirement of citizenship under specific obligations (The Economist, 2013). Citizenship in Austria and Switzerland is bound to jus sanguinis or naturalization (Vink & de Groot, 2010) stipulating an ethnic-based mode of citizenship. For Austria, Switzerland and Germany the naturalization requirements are quite restrictive – while Germany recently has become more liberal, Austria is tending towards a more restrictive naturalization regulation (Hofhansel, 2008). Within the context of the ethnic mode, the visibility of Muslims and Islam in a society becomes a central issue:
Muslim spaces [...] are read by some as symbols of insularity and possible sites of insurrection, prompting questions about minority ethnic citizenship, national identity and belonging. Ideas about `race', ethnicity, and religious identity have thus become intertwined to construct images of outsiders that disrupt and defy a national sense of belonging rooted in whiteness and Christianity. (Phillips, 2006, p. 27)
Hence, the ethnic mode of citizenship and nationhood leads to further marginalisation and discrimination of minority cultures.
To conclude the second line of argument, it is anticipated that instead of focusing on ethnic- related issues regarding Muslims and Islam (e.g. cultural and social anxieties), in countries with civic-based modes of nationhood and citizenship the islamophobic discourse converges on other islamophobic patterns such as insecurity-related stereotypes and prejudices. The comprehension of the role of citizenship and nationhood within the islamophobic discourse is required because citizenship is increasingly ‘culturalised’, i.e. identity, belonging and specific cultural traits are becoming pivotal qualities within the integration debate (Suvarierol, 2012). National states are insisting on their roles as guardians of national identities and social cohesion while managing the
so-called European refugee crisis. This lack of European solidarity is refuting multiculturalism, especially, regarding marginalised, less skilled and culturally different immigrants. The differentiation between ‘us’ and ‘them’, insiders and outsiders, citizens and non-citizens mirrors the securitisation of the issue by the state creating anxieties regarding an external threat to deteriorate the attitude of the society towards outsiders and their rights (Cesari, 2012; Kofman, 2005). The term ‘securitisation’ describes the process of extreme politicisation and the construction of an issue as a security threat and existential concern. The securitisation of Islam and the associated portrayal of Muslims as a threat to national security is especially relevant in the current so-called refugee crisis as migration and asylum are often also securitised (Ibrahim, 2005). Cesari (2013) argues in the context of national community and symbolic boundaries with a differentiation between internal and external enemies. While Islam is perceived as incompatible with national values and Muslims as unable or unwilling to integrate in the first, the latter perceives Islam as a motor of terrorism and Muslims as internal enemies.
2.3. The Role of Media
Muslims are often portrayed in media and policy discussions sharing homogeneous cultural traits and collective integration inabilities despite the heterogeneity of Islam (Triandafyllidou, 2015). For instance, in light of such a homogenization, repetitive demands towards every Muslim to publicly distance him- or herself from terror attacks by fundamentalists imply an alleged complicity. It is expected that some media generate (in this case negative) emotional reactions and higher levels of prejudice against minority and out-groups (Das, Bushman, Bezemer, Kerkhof, & Vermeulen, 2009). Security concerns due to perceived linkage between terrorism and Islam result in a distortive perception of ordinary Muslims as a consequence of higher threat estimations (Borell, 2015).
People tend to be more judgemental and prejudiced after watching news covering terrorism as subconsciously suspicions and fears towards out-groups are triggered (Das et al., 2009).
In general, public media add to the shaping of public opinion and knowledge. The debate on out-groups is biased towards the social majority-group (T. A. van Dijk, 2000). Mainstream knowledge about and awareness of Islam and Muslims is shaped by their representation in popular (mass) media – where the connection to violence or terrorism is often established (Creighton &
Jamal, 2015; Moore, Mason, & Lewis, 2008). Islamophobic assumptions expressed in and circulated by the mass media reinforce the acceptance of stereotypes and prejudices within the context of islamophobic agitations (Ossewaarde, 2014). Islamic stereotypes are even more corroborated by newspaper rhetoric, since the media act as an catalyst and multiplier of islamophobic sentiments (Allievi, 2012). Despite the fact that Islam and Muslims are a part of a pluralistic society, their political, cultural and social claims are addressed in marginalising and discriminatory ways. Islam is covered by media, public and governmental institutions in ways that differ from the representation of other religions (Allievi, 2012). According to Ekman (2015, p.
1987), “islamophobic actors have pushed the boundaries of publicly accepted speech on Muslims and immigration [by endorsing and normalizing extreme standpoints on migration and migrants]“.
All in all, the perception and public portrayal of Muslims and Islam by the media is central as media coverage concerning these issues tends to be decontextualized and linked to discourses of fear and threats instead of being based on facts (Moore et al., 2008). Widespread acceptance and intensity of extreme and negative media representations of Muslims and Islam mirror islamophobia. This makes a structural evaluation of the image on Muslims and Islam created by the media essential.
2.4. Concluding Remarks
The conceptualisation of islamophobia varies in regard to contexts and tendencies – it can be associated with islamophobic prejudices, different forms of anxieties or racialization. As a result, this thesis conceptualises islamophobia as a general and broad notion summing up anti-Muslim attitudes and negative perceptions of Islam, which adopt different shapes in regard to national and contemporary contexts and the newspaper rhetoric. Hence, the very fact that islamophobia is not clearly defined allows the inclusion of various conceptualisations to inquire diverse islamophobic tendencies within the discourse analysis. By examining existing literature the link between the two lines of the argument in this research – varieties of islamophobia and modes of nationhood and citizenship – becomes more evident as the distinction into ‘them’ and ‘us’ is central to both. The modes of nationhood and citizenship divide into two types – a more liberal civic mode and a culturally oriented ethnic mode. As no data is collected for modes of nationhood and citizenship, the theories presented in literature are points of reference. The role of media is examined as it is expected that the selected newspapers contribute to and shape the islamophobic discourse with formulations and rhetoric corroborating the public acceptance of islamophobic attitudes. Hence, the theoretical framework regarding the role of media explains the choice of discourse analysis as the methodological instrument of the thesis. Not only the content of the newspapers is of relevance, but also underlying meanings carried by rhetoric. In the end of the paper, the theoretical framework itself is reviewed in light of the results and in the context of the current so-called European refugee crisis, the associated ‘shift to the right’ and decline of European solidarity to evaluate new tendencies in the islamophobic discourse and add to the existing scientific literature.
3. Methodological Framework
This section constructs the methodological framework for the thesis and introduces the method of the discourse analysis. As indicated in the theory section, the role of media is essential in the islamophobic discourse due to its influence on mainstream knowledge about Islam and Muslims. A content analysis would not capture all facets of islamophobia. In a discourse analysis, however, the underlying meanings are included and reflected such as rhetorical and linguistic formulations. This is crucial to reach beyond the mere content and to include also structural and subliminal islamophobic sentiments in the analysis. It is anticipated that these underlying meanings add to the negative image of Muslims and Islam. The study is conducting a discourse analysis of online articles by national newspapers to include the impact of media on islamophobia. This is necessary as media coverage is frequently biased towards Islam and Muslims (SETA, 2016). In the following, the case and data selection, the methods of data analysis, the coding scheme and the discourse analytical concept are introduced and the ways of proceeding are explained.
3.1. Case Selection
The selection of the included countries is based on the expected differences regarding the modes of nationhood and citizenship. Austria, Germany and Switzerland are expected to have a more ethnic model of citizenship and nationhood, whereas the United States is chosen due to its civic approach.
Germany is a key actor in managing the refugee crisis within the political framework of the European Union. Austria is also a vital player due to its transit status in regard to the so-called Balkan route. Although not a EU Member-state, Switzerland is included due to its public referendums – which entailed the ban of the construction of minarets (Kern, 2013) – and due to its close bilateral ties with the EU. The focus on the central European States is chosen, since the comparison of the three Neighbouring states with an ethnic approach to nationality and citizenship as well as cultural and ideological similarities could reveal national differences regarding islamophobic patterns more clearly. Besides, the USA is included to examine the discourse variances as a non-European state, i.e. where the refugee crisis is not as prevalent and immediate as in Europe. Additionally, the feasibility of this study is considered in the selection of the countries – especially regarding language barriers as the inclusion of newspapers in the native language of each country is essential to avoid biases in the analysis.
3.2. Data Collection
Regarding data collection, 80 online newspaper articles are included in total, i.e. 20 per country, to allow an outline of the national islamophobic discourse. Two different newspapers per country were included to cover the islamophobic discourse separately in all four countries. The intention is
not to compare the newspapers, but to thoroughly reconstruct the national debate in the chosen timeframe. Hence, eight newspapers are part of the analysis, specifically, Die Presse, Neue Kronen Zeitung, Frankfurter Allgemeine Zeitung, Bild, Neue Zürcher Zeitung, 20 Minuten, The New York Times and Daily News. The selection of the newspapers is based on four aspects: (1) the newspaper is written in the national native language to avoid biases in the analysis, (2) the nationwide circulation average, (3) the inclusion of both a liberal-conservative quality newspaper and a tabloid for each country, and (4) the newspapers have an official website with online articles. Newspaper and especially the more extreme tabloids anchor the Islam and Muslim image and often highlight extreme and non-representative examples of Islam instead of the moderate majority of Muslims (Altikriti & Al-Mahadin, 2015). Hence, the inclusion of these two genres is valuable for the study as the tabloids’ rhetoric is assumedly more extreme and negative than the conservative newspapers’
expressions (Moore et al., 2008).
The articles are pre-selected based on the inclusion of at least one of the following key words in the headlines of the online articles: *Muslim* and *Islam*. The search engine ‘Google search’ is used to find articles for each newspaper in the selected timeframe. The pre-selection is compromised of 233 articles. After the pre-selection, articles are included that most clearly contribute to or channel the direction of the islamophobic discourse. Hence, the final 80 articles, which are listed in the appendix A (Table A1, p. 38) next to all pre-selected articles, are chosen in regard to the content of the full article with simultaneous consideration of covering the defined timeframe between September 2015 and March 2016. The latter selection criterion is of relevance as during the period numerous national and international key events occurred: among others, the terror attacks in Paris, San Bernadino, Ankara and Brussels; the anti-Muslim statements of several politicians, e.g. by Donald Trump; the events in Cologne on New Year’s Eve; et cetera. Those events will – as it will most probably show – have impinged on the national islamophobic discourse leading to indiscriminate and explicit attitudes against Muslims and Islam in accordance with the theory by Borell (2015) on the reactivity of islamophobic sentiments.
3.3. Data Analysis
Discourse is defined in this thesis as an ensemble of ideas and conceptions which are shaped and reproduced through argumentative structures and sets of practices (Hajer & Versteeg, 2005).
Hence, the discourse analysis concentrates on the link of systematic structures of debates to prevalent social and political contexts (T. A. van Dijk, 2000). It regards the reality as socially constructed, which puts an emphasis on the underlying meaning of communication (Hajer &
Versteeg, 2005) and focuses the “[...] role of discourse in the (re)production and challenge of dominance. Dominance is defined here as the exercise of social power by elites, institutions or groups that results in social inequality, including political, cultural, class, ethnic, racial and gender inequality” (Teun A. van Dijk, 1993, p. 249f). The method is chosen as it can reveal prevailing power structures within the discourse on islamophobia. Thus, it is believed that this discourse is not
neutral, but linked to and intensified by political, societal and cultural power relations appearing in the discourse as well as in the newspapers’ rhetoric and language. This approach follows the argumentation that the islamophobic discourse follows a social framework on what is regarded normal and abnormal (Altikriti & Al-Mahadin, 2015). The analysis of media is expected to reveal national patterns of the islamophobic discourse, i.e. not the intensity of the prevalent islamophobic discourse is examined, but rather its thematic varieties as well as underlying structures of the media discourse. According to Teun A. van Dijk (1993), discursive marginalization of a group is not only based on the content, also on the structures and strategies of representing the ‘other’.
The debate in the media with rhetorical expressions by politicians or other persons of the public sphere are essential for the development and climate of the islamophobic discourse by propagating and “including particular slogans, stereotypes, specific terms and phrases” (Ossewaarde, 2014, p.
5). In the islamophobic discourse, such terms depicting islamophobia and stigmatizing Muslims are for instance ‘radical’, ‘fundamentalism’ or ‘terrorist’ (Moore et al., 2008). For the data analysis, a coding scheme is developed with a deductive approach based on the theoretical framework to avoid biases during the coding process. In the discourse analysis, all selected articles were coded in the qualitative data-analysis program ATLAS.ti with the scheme provided on the next page (Table 1).
This coding scheme derives from and mirrors the range of theories on islamophobic patterns presented in the theoretical framework and offers a provisional outline of the categories, coding rules as well as related keywords. For each category an exemplifying quote from the analysis is listed in the appendix A (Table A2, p. 50). All included keywords and the entire coding scheme are translated into German to analyse the newspaper articles in Austria, Germany and Switzerland (Table A3, p. 51 and Table A4, p. 52). Possible plural, stem forms (e.g. terror, terrorism, terrorist) and obvious synonyms (e.g. migration flow/refugee flow) of each keyword are considered in the coding process to avoid loss of valuable data. However, the coding is essentially guided by the coding rules. The key words are not determinative, but chosen to support and facilitate the search process. This approach is preferred due to developments during the first coding stage, i.e. the schemes were adjusted to include the coding rules as coding only based on key words excluded crucial quotes while including irrelevant quotes (e.g. when a connection to Muslims or Islam was missing). Some key words were added in the process, which are highlighted in green colour in the coding scheme, as these proved indicative to islamophobic patterns, while others (crossed out in the scheme) anticipated in the beginning were not mentioned at all in the selected newspaper articles.
Table 1: Coding Scheme – Thematic Varieties of the Islamophobic Discourse
Concept Sub-concepts Categories Coding Rule Key Words Thematic patterns from the
theory section Codes used in
the coding process
In the newspaper articles a connection is established to the images below:
Used to support and facilitate the coding process:
I. Stereotypes and
Prejudices
Stereotypes and Prejudices
Perception of demographic threat
Image of Muslims as a large part of the society is created (due to migration or birth rate).
invasion, flood, migration crisis/influx/flow, mass migration
Perception of backwardness
Image of Islam/Muslims as not progressive and in need of reform is created.
barbaric, backwardness, reform, progressive
Perception of oppression of women
Image of suppression of women and their rights by Islam and Muslims is created.
misogyny, patriarchy, veil, burka, gender equality/roles, suppression of women, women rights, hijab, role of a woman, obedience
Violence- related stereotypes and prejudices
Perception of physical oppression of women
Image of physical suppression of women by Islam and Muslims is created.
beating, rape, (sexual) assault, honour killing, abuse, forced marriage, molester, female genital mutilation Perception of
crime threat
Image of threat of crimes due to Muslims is created.
crime (increase/statistics), thief, dealer, suspect, imprisonment Perception of
violence threat
Image of Muslims or Islam as intrinsically violent is created.
violence, brutality, weapons
Insecurity- related stereotypes and prejudices
Perception of
radical Islam Image of radical Islam and
Muslims is created. radical, extremist,
fundamentalist, Islamist scene Perception of
inner threat to the ‘West’
Image of insecurity and terrorism due to Islam/Muslims is created.
jihad, terror, attacks, network, militant, suicide bomber,
‘Gefährder’
II.
Anxieties Cultural
Anxiety Fear of cultural deterioration
Image of cultural deterioration of the ‘Western’ cultural framework through Islam und Muslims is created.
Islamisation, mosque, religion,
‘Leitkultur’, (shared) values, headscarf, minarets, multi-culti
Social Anxieties
Fear of social decline
Image of social decline (e.g.
due to perceived demographic threat) is created.
welfare, benefit, unemployment, uneducated, job, burden, social decline/perspective
Fear of social
segregation Image of social segregation and inability or unwillingness to integrate is created (‘them’ and
‘us’).
segregation, parallel society, parallel structure, no-go-area, (unwillingness to) integrate
Political Anxieties
Fear of ideology and intolerance
Image of Islam as a totalitarian political ideology rather than a religion and of Muslims’
intolerance of other religions is created.
ideology, indoctrination, political agenda, salafism, (hate-
)preacher, hatemonger, infidels, caliphate, propaganda, Islamic doctrine
Fear of infiltration of Islam
Image of Islam infiltrating the
‘Western’ legal system with the sharia is created.
infiltration, sharia, Islamic law, Islamic rule, constitution, anti- democratic, parallel judiciary III.
Racialization Racialization Racialization Image of Muslims as not merely a religious but a racial or single unitary community is created.
the Muslim community, skin colour, complexion, (entry) ban, barring
Besides the varieties of the discourse, the images of Muslims and Islam reproduced by the media are of interest and whether these contribute to the inequality regarding the (public) representation of and knowledge about Muslims and Islam creating a negative image or generate a neutral or even positive image. The study concentrates on identifying focused backgrounds, emphases and contents of the national islamophobic discourses as well as specific and striking usage of language in the articles. Hence, the usage of rhetorical figures and language such as
metaphors and hyperboles or structural concentration on negative actions of the marginalized group is of interest (Teun A. van Dijk, 1993). In Table 2 (p. 16) the dimensions of the underlying structures are presented which are used to evaluate the image created by the media. The coded articles are hence reviewed based on this discourse analytical concept to assess the role of the media in islamophobic discourse. Muslim and Islam related topics are positively reflected in the articles when islamophobic attitudes are deplored, or neutral, when based on facts and described without any underlying meaning. Islamophobic sentiments are negatively reflected in the articles if explicit islamophobic attitudes are stated or a negative image is created through formulations (e.g.
generalisation, de-contextualization) and use of language (e.g. hyperboles, metaphors, exaggerations) (Teun A. van Dijk, 1993). As islamophobia revolves around the negative representation of Islam and Muslim the study focuses mainly on the negative image of the discourse analytical concept.
Table 2: Discourse Analytical Concept
2 As an indication for the sources of the quotes the bracketed numbers refer to the enumeration of pre-selected and selected articles in the appendix (Table A1, p. 38).
Concept Image Explanation Exemplifying phrases from the analysis Underlying
structures
Positive Categories in the coding scheme are positively reflected in the article, i.e.
islamophobic attitudes and images are deplored and only mentioned based on facts.
„Der Vorfall zeigt erneut, dass wir in Europa vor der immer wiederkehrenden Aufgabe stehen, über Probleme des Zusammenlebens in unseren pluralen, multikulturellen Gesellschaften zu diskutieren und Lösungen zu finden, ohne Rassisten oder Feinden des Islam das Feld zu überlassen.“ (5.21.)2 [The incident shows once again that we in Europe are facing the recurring challenge of discussing problems regarding social coexistence in our pluralistic, multicultural society and finding solutions without leaving the floor to racists or enemies of Islam.]
Neutral Categories in the coding scheme are neutrally reflected, i.e. islamophobic attitudes and images are stated based on facts without underlying meaning.
“Misstrauen gegenüber Religion allgemein und Angst vor dem Islam nach Paris, obgleich Muslime selber Opfer und
Anschlagsziel der Terroristen mehrheitlich in der Welt sind, beeinträchtigen die Flüchtlings- und Integrationsarbeit erheblich.”
(3.6.) [Suspicion of religion in general and fear of Islam since Paris, although predominantly Muslims themselves are victims and targets of the terrorists all over the world, interferes with the work with refugees and integration.]
Negative Categories in the coding scheme are negatively reflected, i.e. explicit islamophobic attitudes are stated or a negative image is created through formulations (assumptions, de- contextualization and generalisation) and use of language (hyperboles, metaphors, exaggerations, etc.).
• Hyperbole: “Derzeit kommen über die Flüchtlingsrouten Zigtausende Muslime nach Österreich und Deutschland.” (2.4.) [Currently, trillions of Muslims are coming to Austria and Germany on the refugee routes.]
• Metaphor: “Der Koran ist ein rauchender Colt.” (5.6.) [The Quran is a smoking colt.]
• Exaggeration: “Viele Muslime flüchten nach Europa, um hier ihre Menschenrechte wahrnehmen zu können. Und dann stellen sie fest: Das geht hier auch nicht mehr.” (2.4.) [Many Muslims flee to Europe to assume their human rights. And then they realize: This is here also not possible anymore.]
• Generalisation: “Zwei von drei Deutschen sehen einen Zusammenhang zwischen den Terror-Schlägen und dem Islam.
[...] Befragt wurden vom 24. bis 29. März 1873 Personen.”
(4.15.) [Two of three Germans see a connection between terror attacks and Islam. [...] Between May 24th and 29th 1873 persons were surveyed.]
3.4. Concluding Remarks
By means of the discourse analysis as the methodological approach, the coded articles are used to reconstruct the national debates on islamophobia for the chosen timeframes. The coding process aims at detecting nationally prevalent thematic patterns and the cross-country varieties of these patterns with the provided coding scheme. The research activities commenced with the construction of the coding scheme and the discourse analytical concept in accordance with the theoretical framework. After the selection of the newspaper articles, ATLAS.ti was used for the subsequent coding process. The data analysis followed mainly the coding rules provided in the scheme to evaluate varieties of the discourse and was guided by the provided key words. Each article was coded multiple times to avert mistakes and to ensure the codes are assigned correctly.
To reach a summarized overview of the detected and important varieties of islamophobic discourse, Table B (p. 53) in the appendix was created. The discourse analytical concept is taken into account during the coding process to detect especially negative images of Islam and Muslims reaching beyond the mere content of the islamophobic pattern. Thereby, power structures in the social and political context can be analysed as nationwide and popular newspapers are selected as data source and underlying structures of the media discourse and the created images of Muslims and Islam are examined in the analysis in the next section. By means of the ATLAS.ti output summarized in Table B and the results from the discourse analytical concept, the actual discourse analysis was approached and realised, which is provided in the next section.
4. Discourse Analysis
In accordance with the two sub-questions, the associated two lines of argument and the theoretical framework, the analysis section is divided into two main parts for each of the two central arguments of this study. To recapitulate: while the first argument revolves around the thematic variation of islamophobic sentiments and the usage of language to broadcast underlying meanings, both circulated by the media, the second argument claims that country-specific modes of nationhood and citizenship are a central explanatory factor for these varieties. Therefore, in the first part of this section the discourse analysis detects varieties within the islamophobic discourse and the associated national prevalence (answer to sub-question A). This analysis is used (1) to reveal transnational varieties of islamophobic attitudes and (2) to discuss the role of media – and their use of formulations and language. The inclusion of subliminal meanings conveyed in the newspaper articles is crucial as language is recognized as non-neutral. In the second part the linkage of the varieties detected to country-specific modes of nationhood and citizenship is examined (answer to sub-question B). It is anticipated that the ethnic approach to concepts of nationhood and citizenship furthers a focus on themes related to national identity and belonging (e.g. cultural and social anxieties), whereas the civic mode leads to other forms of islamophobic sentiment. While answering the second sub-question, striking islamophobic patterns are discussed, which were not anticipated or are contradictory to the theoretical framework. The concluding remarks summarise the answers to both of the sub-questions.
4.1. Cross-country Comparison of the Varieties of the Islamophobic Discourse
In the following, the thematic varieties of the islamophobic discourse are discussed in consideration of the underlying structures and conveyed negative images. A cross-country comparison is conducted to assess similarities and differences for the selected countries Austria, Germany, Switzerland and the US. Table B in the appendix B (p. 53) summarizes exemplifying and important quotes for the analysis in accordance with the coding scheme to give an overview.
Hence, instead of aligning with the coding scheme, this section emphasises contexts and interrelations of all different islamophobic (sub-)patterns to highlight striking features and reflect these in light of the theoretical framework.
Confirming the theory by (Borell, 2015) on the reactivity of islamophobia, the incidents in Cologne at New Year’s Eve 2015/2016 generated an intense debate on women’ role in Islam. All four countries address the incidents, however, only one marginal reference is made in an US- American newspaper (7.27.)3, In general, the image created is negative neglecting facts and differentiation. For instance, a direct connection between the suspects and Islam is immediately
3 As an indication for the sources of the quotes the bracketed numbers again refer to the enumeration of pre-selected and selected articles in the appendix (Table A1, p. 38).
