The "European refugee crisis" through a media lens. A cross-national study into press representations of displaced people combining corpus linguistics and argumentation analysis

1 University of Warsaw University of Seville Natalia Zawadzka-Paluektau The “European refugee crisis” Through a Media Lens. A Cross-National Study into Press Representations of Displaced People Combining Corpus Linguistics and Argumentation Analysis PhD Thesis Field of study: Linguistics Supervisors: dr hab. Łucja Biel, prof. ucz. Catalina Fuentes Rodríguez, catedrática de universidad Marina González Sanz, profesora titular de universidad Warsaw/ Seville 2023 2 Keywords: migrations, European refugee crisis, media discourses, corpus linguistics, discourse analysis, argumentation analysis Astract: The aim of the investigation is to contribute to the field of research on discourses surrounding the “European refugee crisis” by exploring how people migrating from Asia and Africa to Europe between 2015 and 2018 were represented in British, Polish, and Spanish newspapers. To this end, it applies a triangulation of two methodologies, argumentation analysis and corpus linguistics, to a corpus of press articles of over six million tokens. The investigation departs from the assumption that a study of news should account for the issues involved in its creation and reception in the given historical and sociopolitical context. Hence, it first explores the relationship between media exposure and the audiences' perception of the world as well as sheds some light on the factors influencing media message programming. Next, it looks at the historical and sociopolitical processes which might have contributed to shaping the present-day responses to immigration of the three discourse communities under the investigation. Investigating these aspects is considered to be a complementary source of hypotheses and insights which might help illuminate and support findings of the subsequent discourse analysis. The analysis follows several steps. First, tools associated with corpus linguistics are employed to conduct a collocation analysis of the highest-frequency terms denoting the displaced people, a wordlist analysis, and a more detailed examination of the most prominent trends revealed as a result of the preceding steps, using concordance and collocation analyses. Second, an analysis of argumentation is conducted on a downsampled portion of the corpus. Third, a special analytical 3 procedure is adopted to systematically retrieve and examine migration metaphors which are common in the corpus. The results show that the 2015-2018 migrations were predominantly conceptualised in terms of a crisis for Europe, whereas the displaced people tended to be discursively deprived of the right to seek asylum via a wide range of delegitimising linguistic strategies. This overarching trend is largely consistent across the three language subcorpora, with some variation in the frequencies of its particular linguistic realisations. Similarly, while both main analytical approaches point towards this general conclusion, they sometimes differ in the emphasis they put on specific findings, or contribute findings which are in line with the general picture, but which have not been observed using the other methodology. This provides evidence of the usefulness of the methodological triangulation applied to this analysis. The thesis concludes with a discussion of the implications of the way in which the newspapers of Poland, Spain, and the United Kingdom construe migrations and their principal protagonists, particularly with respect to its impact on the safety and well-being of the displaced people. 4 Uniwersytet Warszawski Uniwersytet Sewilski Natalia Zawadzka-Paluektau „Europejski kryzys uchodźczy” z perspektywy mediów. Porównawcze badanie obrazu przesiedleńców w prasie z wykorzystaniem lingwistyki korpusowej i analizy argumentacji Praca doktorska Na kierunku: lingwistyka Promotorki: dr hab. Łucja Biel, prof. ucz. Catalina Fuentes Rodríguez, catedrática de universidad Marina González Sanz, profesora titular de universidad Warszawa/ Sewilla 2023 5 Słowa kluczowe: migracje, europejski kryzys uchodźczy, dyskursy medialne, lingwistyka korpusowa, analiza dyskursu, analiza argumentacji Streszczenie: Celem pracy jest przyczynienie się do rozwoju badań nad dyskursami dotyczącymi „europejskiego kryzysu uchodźczego” poprzez przeanalizowanie, w jaki sposób ludzie migrujący z Azji i Afryki do Europy w latach 2015-2018 byli ukazywani w brytyjskich, polskich i hiszpańskich dziennikach. Do zbadania przekraczającego sześć milionów segmentów korpusu tekstów prasowych zastosowana jest triangulacja dwóch metodologii: analizy argumentacji i lingwistyki korpusowej. Punktem wyjścia analizy jest założenie, że badanie wiadomości medialnych powinno uwzględniać kwestie związane z ich tworzeniem i odbiorem w określonym kontekście historycznym i społeczno-politycznym. W pierwszej kolejności zatem, badanie zwraca uwagę na wpływ mediów na postrzeganie świata przez ich odbiorców oraz rzuca światło na czynniki wpływające na programowanie przekazu medialnego. Następnie przygląda się procesom historycznym i społeczno-politycznym, które mogły przyczynić się do ukształtowania reakcji na imigrację w trzech badanych społecznościach dyskursywnych. Uwzględnienie owych aspektów uznajemy za uzupełniające źródło hipotez i spostrzeżeń, które mogą pomóc naświetlić i potwierdzić wyniki późniejszej analizy dyskursu. Analiza składa się z kilku etapów. Najpierw, za pomocą narzędzi związanych z językoznawstwem korpusowym, przeprowadzono analizę słów o najwyższej frekwencji w korpusie, analizę kolokacji najczęstszych terminów stosowanych wobec przesiedleńców, a także bardziej szczegółowo zbadano najistotniejsze zaobserwowane w tej sposób trendy, z wykorzystaniem analiz konkordancji i kolokacji. Następnie, na próbie części korpusu, przeprowadzona została analiza argumentacji. Na koniec 6 zastosowano specjalną procedurę analityczną do systematycznego wyszukiwania i badania metafor migracyjnych, które są powszechne w korpusie. Wyniki pokazują, że migracje w latach 2015-2018 były głównie konceptualizowane w kategoriach kryzysu dla Europy, podczas gdy wysiedleńcy byli dyskursywnie pozbawiani prawa do ubiegania się o azyl za pomocą szerokiego wachlarza delegitymizujących strategii językowych. Ten nadrzędny trend jest w przeważającej mierze spójny we wszystkich trzech podkorpusach językowych, z pewnymi różnicami w częstotliwości jego poszczególnych realizacji językowych. Podobnie, podczas gdy oba główne podejścia analityczne prowadzą do tego ogólnego wniosku, czasem różnią się one naciskiem, jaki kładą na konkretne obserwacje lub dostarczają obserwacji, które są zgodne z ogólnym obrazem, ale których nie zaobserwowano przy użyciu drugiej metodologii. Świadczy to o użyteczności triangulacji metodologicznej zastosowanej w niniejszym badaniu. Pracę wieńczy omówienie implikacji, jakie niesie za sobą obraz migracji i uchodźców kształtowany przez prasa polską, hiszpańską i brytyjską w danym okresie, ze szczególnym uwzględnieniem jego wpływu na bezpieczeństwo i dobrobyt przesiedleńców. 7 Universidad de Varsovia Universidad de Sevilla Natalia Zawadzka-Paluektau La “crisis europea de los refugiados” a través de una lente mediática. Un estudio comparativo de representaciones periodísticas de las personas desplazadas combinando la lingüística de corpus y el análisis de la argumentación Tesis de doctorado Área principal de estudios: Lingüística Tesis dirigida por: dr hab. Łucja Biel, prof. ucz. Catalina Fuentes Rodríguez, catedrática de universidad Marina González Sanz, profesora titular de universidad Varsovia/ Sevilla 2023 8 Palabras claves: migraciones, crisis de refugiados, discursos mediáticos, lingüística de corpus, análisis del discurso, análisis de la argumentación Resumen: El objetivo de la tesis es contribuir al campo de investigación sobre los discursos en torno a la “crisis europea de los refugiados” explorando cómo las personas que emigraron de Asia y África a Europa entre 2015 y 2018 fueron representadas en periódicos británicos, polacos y españoles. Para ello, empleamos una triangulación de dos metodologías, el análisis argumentativo y la lingüística de corpus, a un corpus de artículos periodísticos de más de seis millones de segmentos. El análisis parte del supuesto de que un estudio de noticias debe dar cuenta de las cuestiones involucradas en su creación y recepción en el determinado contexto histórico y sociopolítico. Por lo tanto, primero exploramos la relación entre la exposición a los medios y la percepción del mundo por parte de las audiencias, además de arrojar luz sobre los factores que influyen en la programación de los mensajes de los medios. A continuación, analizamos los procesos históricos y sociopolíticos que podrían haber contribuido a dar forma a las respuestas actuales a la inmigración de las tres comunidades discursivas que forman el objeto de esta investigación. Se considera que investigar estos aspectos es una fuente complementaria de hipótesis y puntos de vista que podrían ayudar a iluminar y respaldar los hallazgos del análisis del discurso. El análisis sigue varios pasos. En primer lugar, se emplean herramientas asociadas con la lingüística de corpus para realizar un análisis de colocación de los términos de mayor frecuencia que denotan a las personas desplazadas, un análisis de listas de palabras y un examen más detallado de las tendencias más destacadas reveladas como resultado de los pasos anteriores, utilizando los análisis de 9 concordancia y colocación. En segundo lugar, se realiza un análisis de la argumentación sobre una porción del corpus. En tercer lugar, se adopta un procedimiento analítico especial para recuperar y examinar sistemáticamente las metáforas de la migración que son comunes en el corpus. Los resultados muestran que las migraciones de 2015-2018 se conceptualizaron predominantemente en términos de una crisis para Europa, mientras que las personas desplazadas tendieron a ser privadas discursivamente del derecho a buscar asilo a través de una amplia gama de estrategias lingüísticas deslegitimadoras. Esta tendencia general es en gran medida consistente en los tres subcorpus, con algunas variaciones en las frecuencias de sus realizaciones lingüísticas particulares. De manera similar, si bien los dos enfoques analíticos principales apuntan hacia esta misma conclusión general, a veces difieren en el énfasis que ponen en hallazgos específicos, o aportan hallazgos que están en línea con el panorama general, pero que no se han observado utilizando la otra metodología. Esto evidencia la utilidad de la triangulación metodológica aplicada a este análisis. Finalmente, se discuten las implicaciones del imagen de las migraciones y los refugiados formado por la prensa de Polonia, España y el Reino Unido, particularmente con respecto a su impacto en la seguridad y el bienestar de las personas desplazadas. 16 Table 4.4. Verbs Collocating Frequently with Inmigrante [‘Immigrant’] in the Spanish Subcorpus 125 Table 4.5. Nouns and Adjectives Collocating with Migrant in the UK Subcorpus 126 Table 4.6. Verbs Collocating with Migrant in the UK Subcorpus 129 Table 5.1. Most Frequent Context Indicators in the Polish Subcorpus 131 Table 5.2. Most Frequent Context Indicators in the Spanish Subcorpus 133 Table 5.3. Most Frequent Context Indicators in the UK Subcorpus 136 Table 6. Most Frequent Topic Indicators in the Corpus 139 Table 7.1. Topic Indicators Indexing the Topic of Threat in the UK Subcorpus 143 Table 7.2. Nouns and Adjectives Collocating with Attack (as Noun) in the UK Subcorpus 144 Table 7.3. Nouns and Adjectives Collocating with Terrorist in the UK Subcorpus 146 Table 7.4. Topic Indicators Indexing the Topic of Threat in the Polish Subcorpus 148 Table 7.5. Nouns and Adjectives Collocating with Atak [‘Attack’] in the Polish Subcorpus 150 Table 7.6. Topic Indicators Indexing the Topic of Threat in the Spanish Subcorpus 153 Table 7.7. Nouns and Adjectives Collocating with Ataque [‘Attack’] in the Spanish Subcorpus 154 Table 9.1. Refugees’ and Immigrants’ Geographical Identities in the Polish Subcorpus 159 17 Table 9.2. Refugees’ and Immigrants’ Geographical Identities in the UK Subcorpus 160 Table 9.3. Refugees’ and Immigrants’ Geographical Identities in the Spanish Subcorpus 162 Table 9.4. First Time Asylum Applicants in the EU (Eurostat, 2016, 2017, 2018, 2019) 164 Table 10.1. Collocations Refugee + Child and Migrant + Child in the UK Subcorpus 168 Table 10.2. Collocates of Child + Refugee/ Migrant in the UK Subcorpus 169 Table 11. Religion-Related Collocates of the Main Query Terms 174 Table 12.1. Gender-Related Terms: Frequency Information 181 Table 12.3. Collocates of Kobieta [‘Woman’] and Mężczyzna [‘Man’] in the Polish Subcorpus 184 Table 12.4. Collocates of Mujer [‘Woman’] and Hombre [‘Man’] in the Spanish Subcorpus 186 Table 13.1. Explicit Linguistic Realisations of the Securitisation Conclusion 196 Table 13.2. Arguments For the Securitisation Stance 203 Table 13.3. Subarguments for Argument S1. 210 Table 14. Arguments For the Humanitarian Aid Stance 260 Table 15.1. Metaphors of Immigration 288 Table 15.2. COMMODITY Metaphors in the Spanish Subcorpus 290 18 Table 15.3. COMMODITY Metaphors in the UK Subcorpus 292 Table 15.4. COMMODITY Metaphors in the Polish Subcorpus 294 Table 15.5. WATER Metaphors in the Polish Subcorpus 297 Table 15.6. WATER Metaphors in the UK Subcorpus 298 Table 15.7. WATER Metaphors in the Spanish Subcorpus 301 Table 15.8. ANIMAL Metaphors in the Corpus 306 Table 15.9. GUEST Metaphors in the Corpus 310 Table 15.10. WAR Metaphors in the Corpus 313 19 List of Figures Figure 1. Organisation of Discourse 95 Figure 2. Argumentative Structure 97 Figure 3. Distribution of High-Frequency Words in the Polish (Left), Spanish 130 (Centre), and the UK (Right) Subcorpora 130 Figure 4. Distribution of Topic Indicators in the Polish (Top), Spanish (Centre), and the UK (Bottom) Subcorpora 138 Figure 5. Collocation Network: Terrorism — Islam — Refugees/ Migrants in the UK Subcorpus 146 Figure 7. Collocation Network: Terrorism — Islam — Refugees/ Immigrants in the Polish Subcorpus 152 Figure 8. Detections of Illegal Border Crossing by Main Nationalities Between 2015 and 2018 (Frontex, 2016, 2017, 2018, 2019) 166 20 1. Introduction While immigration in general is widely problematised as one of the most important issues in the contemporary Western world (Jacobs et al., 2018, p. 474), the “European refugee crisis” has been thought of as the second top challenge for the 1 European Union after the 2008 financial crisis (e.g. Corbu et al., 2017; Greussing & Boomgaarden, 2017). The term refers to the movement of people from war and other crises-stricken parts of the world (notably, Asia and Africa) to Europe, which began in 2015 when approximately one million refugees crossed the European borders (Frontex, 2015). This was consistent with an increase in the global refugee population: while there had been a steady number of refugees around the world since the beginning of the 20th century, it rose from 3% to 3.4% by the end of 2016 (United Nations High Commissioner for Refugees [UNHCR], n.d.a). The vast majority of the displaced people, however, remained in neighbouring lowand middle-income countries facing economic and development challenges (such as Turkey, Pakistan, and Uganda), while only a small proportion disposed of the resources to undertake a journey north (UNHCR, 2019). Nevertheless, for a few years beginning in 2015, Europe-bound migration remained at a steady level (insignificant fluctuations notwithstanding) until 2018 when it slowly began to decrease (Frontex, 2019) . 2 Following other scholars who have studied what came to be known as the “European refugee crisis” 1 (e.g. Greussing & Boomgaarden, 2017; Krzyżanowski et al., 2017; Rheindorf & Wodak, 2018; Sager, 2018), this term is approached critically in the present study as ideologically-motivated rather than an adequate denomination of the events in question. See Section 12.1 for a more detailed discussion. The situation in Europe changed dramatically when Russia invaded Ukraine in 2022, which forced 2 millions of Ukrainians to seek refuge in neighbouring European countries, most importantly Poland. This, however, does not form part of the present study, which only covers the period between 2015 and 2018 (but see, for instance, Zawadzka-Paluektau, 2022). 21 The accelerated movement of forcibly displaced people from Asia and Africa was due to many, often intertwined, factors such as socio-political oppression and the civil war in Syria; the war in Afghanistan (those were the top two nationalities of refugees); as well as general violence, military conflicts, persecution, drought, and poverty (often caused or exacerbated by the global warming) (e.g. Cosgrave et al., 2016, pp. 15-17). Furthermore, regimes which had formerly collaborated on border control with Europe (such as Libya, Tunisia, and Syria) collapsed as a result of the economic crisis whereas other states, like Turkey and Egypt, began to use migrations for political gain (Maddaloni & Moffa, 2018, p. 610). Most refugees who attempted to migrate to Europe chose either the Eastern or the Central Mediterranean Route. The former — from Turkey to Greece — was used predominantly by Syrians, Iraqis, and Afghans. The latter led from North Africa to Italy or Malta through the “chaotic, lawless, and mired in conflict" Libya (Goodman et al., 2017, p. 106). Furthermore, a significant number of migrants travelling from the Middle East took the Western Balkan Route, extending through Macedonia and Serbia to Croatia, Slovenia, and Hungary. The Western Mediterranean Route — from Morocco to Spain — became particularly important towards a later stage of the “crisis” (MacGregor, 2019). The European Union opened its borders to groups of refugees but only conditionally and briefly. The most notable event which marked the switch to a more strict securitisation approach was the controversial deal with Turkey (March 2016) where it agreed to take back those who arrived in Europe from its territory (Human Rights Watch, 2016). While this led to a decrease in arrivals via this route (Sigona, 2017, p. 458), at the same time, the — already substantial — number of sea deaths began to rise as people were forced to choose other, more precarious routes 22 (International Organisation for Migration, n.d.). Those who managed to reach Europe suffered further difficulties such as restrictions on movement, poor access to housing and public services, health issues, and low safety levels (e.g. Artero & Fontanari, 2021; Busetta et al., 2021; Griffith & Yeo, 2021; Scalettaris et al., 2021), as well as detention in overpopulated refugee centres (e.g. Vaughan-Williams, 2015; Tyler, 2017). Many struggled to obtain the refugee status (e.g. Mayblin, 2019). The most influential definition of refugee, which constitutes the hallmark of refugees’ rights to protection and one of the fundamental rules of international law, is the one established at the 1951 Geneva Refugee Convention. Article 1.A.2 defines a refugee as any person who owing to well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality and is unable or, owing to such fear, is unwilling to avail himself of the protection of that country; or who, not having a nationality and being outside the country of his former habitual residence as a result of such events, is unable or, owing to such fear, is unwilling to return to it. 3 It thus postulates that refugees are entitled to protection and to being provided means to live while they await one of three options: to be repatriated if and when conditions in their country of origin are safe; to be integrated in the country where they first arrive; or to be resettled in “third countries” (Chan, 2019, p. 371). Most importantly, they have the right of non-refoulment, which means that they should be allowed to apply for asylum in the first country they arrive in and cannot be returned to where they came from. However, the process called the Refugee Status Determination is highly dependent on the host society's adherence to the 1951 Although this definition is sometimes problematised, for example for privileging the public sphere 3 over experiences in the private sphere, which tend to concern women more than men (such as, for instance, domestic violence, rape, or female genital mutilation; Christou & Kofman, 2022). 23 Convention (Tirosh & Klein-Avraham, 2019, p. 381), and does not take place in a social vacuum: public perceptions about refugees affect their official status and, therefore, their exclusion or inclusion in host societies. These perceptions are, in turn, shaped by public discourses, especially in contexts when, as was the case with the “refugee crisis”, citizens have no first-hand experience with the given event and its participants, as well as when the event is (relatively) sudden and fast-developing. In fact, while it was by no means the first historical instance of people migrating to Europe — or, more narrowly, the European Union — in search for asylum (in fact, the EU received proportionally many more asylum claims during the last mass displacement in the 1990s. [Menjívar et al., 2019, p. 12]), the recent “refugee crisis” attracted unprecedented political, media, and (consequently) public attention (e.g. Kotišová, 2017; Krzyżanowski, 2018), with the former (the media and politicians) becoming crucial in defining the issue, evaluating the situation, and suggesting solutions. The goal of this investigation is to contribute to the field of research on discourses surrounding the “refugee crisis” by exploring how refugees migrating from Asia and Africa to Europe were represented in European press. It is important to note here that the study adopts Groom’s (2010, p. 59) definition of discourses as “sets of meanings and values which are associated with specific communities or institutions, and which are produced and reproduced through characteristic and often highly conventionalised linguistic choices”. These “meanings and values” may, in turn, be conceptualised in terms of ideologies, i.e. “the taken-for-granted assumptions, beliefs and value systems which are shared collectively by social groups” (Simpson, 1993, p. 5). In other words, discourses created by and for different communities are meant to convey their dominant ideologies. This also means that, at the same time, they strive 24 to exclude or marginalise any alternative ways of seeing the world, especially when they are pronounced by powerful speakers or institutions (Hart, 2014, p. 4). Therefore, they can be employed as tools in the service of maintaining power relations between different groups (van Dijk, 2000). Their aim is thus highly argumentative (Fuentes Rodríguez, 2013, 2017; Hart, 2014), but often in an implicit manner. As a consequence, discourses “need to be inferred through the ways that particular social groups are represented or through the generalisations and assumptions that are embedded in statements” (Baker, 2016a, p. 138). In the context of this study, the ways that refugees are represented in newspaper articles are expected to reveal the underlying dominant discourses of immigration in Poland, Spain, and the United Kingdom. Thus, the overarching research question of this thesis is: (I) How were refugees represented in market-leading British, Polish, and Spanish daily newspapers during the “European refugee crisis” (2015-2018)? Furthermore, in order to give more focus to the analytical process, a series of subordinate research questions was also devised: (1) Which aspects of the migrations are the focus of British, Polish, and Spanish news reports covering the “refugee crisis”? (2) What naming strategies are employed with respect to the displaced people? (3) How do different naming strategies determine the way refugees are represented? (4) How is refugee identity construed? (5) What processes and activities are the displaced people involved in? (6) How do the analysed texts evaluate refugees and their migration to Europe, and how is the evaluation supported? (7) What are the differences and/or similarities in the ways newspapers in Poland, 25 Spain, and the UK represented refugees during the “refugee crisis”? The study also aims at filling some of the gaps left by previous research into discourses surrounding the “refugee crisis”. Specifically, it can be noted that previous analyses had one (but usually both) of the following limitations. On the one hand, they investigated news reports published over a short period of time, usually several months (e.g. Chouliaraki et al., 2015) or days (e.g. Lams, 2018) of the year 2015, whereas the movement of the displaced people did not decrease until 2018 (Frontex, 2019) and, consequently, continued to receive media attention. On the other hand, these studies tended to be based on small samples of texts, which limited their representativity and, therefore, validity. This is related to issues of methodology as mostly qualitative, manual investigations were undertaken. This analysis is, therefore, conducted on a large corpus (over six million tokens) spanning four years since January 2015 until December 2018. It focuses on print media discourses published in three European countries: Poland, Spain, and the United Kingdom. A focus on three countries is a middle-range research strategy, which “permits broader generalisations than single-country studies, while allowing for more contextual and historical evidence than would be possible in statistically oriented studies of a larger sample” (Bleich, 2009, p. 363). Furthermore, the countries selected for this investigation differ regarding the extent to which they were impacted by the “refugee crisis”, as well as a number of other factors likely to shape their responses towards immigration. Specifically, Poland is an Eastern-European country of emigration rather than immigration, with very little contact with people from outside Europe and the sphere of influence of Christianity (e.g. Krzyżanowski, 2018, p. 79; Górak-Sosnowska, 2019, p. 297). During the “refugee crisis”, too, any migration movements across or towards Polish territory were almost non-existent; 32 as well as for people who regularly consume news (Schranz et al., 2018). While it has been shown that low-educated individuals are particularly susceptible to give in to the media logic, research suggests that even education does not protect entirely from its influence (e.g. Billiet et al., 2014; Matthes & Schmuck, 2017). Another study has demonstrated that both the least and the most politically engaged are especially susceptible of buying into the media logic (Fitzgerald et al., 2012). The latter finding seems particularly significant as the most politically oriented citizens are also most likely to vote (Denny & Doyle, 2008; Glenn & Grimes, 1968; Voogt & Saris, 2003), volunteer for campaigns (Verba, Burns, & Schlozman, 1997), engage in blogging and other online political activities (Best & Krueger, 2005; Woodly, 2008), get involved with talk radio (Hofstetter, Barker, Smith, Zari, & Ingrassia, 1999), and persuade others about politics (Hansen, 1997; Katz & Lazarsfeld, 1955). (Fitzgerald et al., 2012, p. 493). In other words, Fitzgerald and colleagues' study suggests that the potential opinion leaders are among the most prone to be influenced by the media. Similarly, studies conducted by Van Aelst and Walgrave (2016) and Mayblin (2019) point to the fact that the elites are not exempt from relying heavily on the media. Specifically, they have demonstrated that the news media are among the primary sources of information about current events for politicians and policymakers. This is particularly important in light of, on the one hand, their active use of information received from the media in their daily work, for example in shaping asylum policies (Mayblin, 2019); on the other hand, the fact that political and other elites then construct and spread their own discourses which further consolidate the effects exerted on the citizens (Van Aelst & Walgrave, 2016). 33 The remainder of this Section discusses the specific areas in which the media have been shown to exercise the greatest influence on their audiences: the construction of knowledge (Section 2.1); the elicitation of emotions (Section 2.2); the formation of attitudes and stereotypes (Section 2.3); and the evolution of behaviours (Section 2.4). 2.1. The role of the Media in Generating Knowledge Numerous scholars posit that media discourse is the main source of knowledge about the world for the majority of society, and that without the news, the major resource of the media (Jackob, 2010, p. 589), people would know very little beyond their own daily personal experiences (e.g. Jacobs et al., 2018, p. 474; Van Dijk, 2014, p. 53; Castells, 2009, p. 136; McLaren et al., 2018, p. 175). For example, a study conducted by Aalberg and colleagues (2013) in eleven different countries has shown a strong positive correlation between the public’s knowledge of international events and the amount of international hard news media coverage. What is more, aside from generating knowledge on issues, the media also decide about their salience (e.g. McCombs & Shaw, 1991; Esess et al., 2014). For instance, scholars argue that the perception of immigration as an important issue in the recent years grew as a result of the significant amount of media coverage it has been receiving (Kotišová, 2017, p. 244; Esser et al., 2016). In a similar vein, Herda (2010) has revealed a correlation between TV news exposure and overestimation of the number of immigrants (which is, in turn, a predictor for feelings of threat). 34 Finally, it is important to note that, apart from generating the public's knowledge on the world, the media, especially newspapers, also possess a “memorysetting function”: by creating “the first draft of history” they establish what will be considered as having mattered today by the future generations (Tishon & KleinAvraham, 2019, p. 385). 2.2. The role of the Media in Eliciting Emotions The way a news story is framed elicits emotions. Specifically, how the problem is defined, how the causes are diagnosed and what solutions are suggested, which aspects of the story are included or omitted, which ones are highlighted or downgraded, and how the issue is evaluated (Entman, 1993, p. 52) all have a strong influence on the readers' emotional reactions (Kühne & Schemer, 2015). For example, whether a given story attributes the blame for an accident to one of the actors or rather focuses on the suffering of the victims will determine which feeling is elicited in the readers: anger in the former case or sadness in the latter (Kühne & Schemer, 2015). Cues suggestive of an actor's emotional experiences are another strong predictor of the readers' lasting emotional reactions to news as they “learn to fear the things that frightened models, to dislike what repulsed them, and to like what gratified them” (Bandura, 2001, p. 281). With respect to immigration, research has shown that news content may elicit negative emotions towards out-groups. For instance, Atwell Seate and colleagues (2018) have demonstrated that the media cultivate anxiety and anger toward minority groups in the United States. 35 What is more, emotions are not only the end result of news framing, but they may also mediate other effects. Research shows that emotions play a part in forming judgment, decision making, attitude and opinion change, as well as governing information processing and recall (Nabi, 1999; 2003; Kepplinger et al., 2012; Matthes & Schmuck, 2017; Kühne & Schemer, 2015; Sapolsky, 2017). For example, Kühne and Schemer (2015), by investigating the emotional effects of anger and sadness news frames on cognitive mechanisms have demonstrated that the anger frame “increased the accessibility of information about punishment and the preference for punitive measures, (…) [whereas] the sadness frame increased the accessibility of information about help for victims and the preference for remedial measures” (p. 387). Similar results were reported by Kepplinger and colleagues (2012). Another study (Atwell Seate & Mastro, 2015) has revealed that news depictions of immigration affected behaviour indirectly through feelings of contempt. These effects are particularly strong for highly identified in-group members (Atwell Seate & Mastro, 2015) as well as those who are unable to process complex information and are therefore more likely to form their judgment about an issue solely on the basis of emotions (Matthes & Schmuck, 2017). 2.3. The role of the Media in Engendering Attitudes and Stereotypes An attitude can be defined as “the association of a social object or social group concept with a valence attribute concept” (Greenwald et al., 2002, p. 5). Attitudes, unlike behaviours or emotions, are considered to be long-term and are formed as a result of long-term exposure to specific messages (Hargreaves, 2001, p. 25, 27). Multiple studies have demonstrated that the media alter the standards people use in 36 evaluating other people, things, and concepts (e.g. Iyengar et al., 1991, p. 95; Brewer et al., 2003; Reeves & de Vries, 2016). With respect to immigration, specifically, research indicates that negative news portrayal of immigrants increases the activation of negative attitudes in the public and inversely — exposure to positive portrayals reduces them (e.g. Schemer, 2012; van Klingeren et al., 2014; Saleem et al., 2015; Wenzel & Żerkowska-Balas, 2018). In fact, it is a much stronger predictor than personal/ local experience (Evans & Mellon, 2019, p. 80). For example, Switat (2015) points out, in the context of Poland, that the negative attitudes towards Arabs might be explained by media exposure: when a survey asked about Egyptians and Palestinians (both Arab nations) instead of Arabs, attitudes became noticeably more positive. Furthermore, such frames may not only impact on the readers' explicitly expressed attitudes, but also on what Matthes and Schmuck call “automatic gut reactions” (2017, p. 558): the more subtle, implicit attitudes, which they found to be resistant even to the moderating effects of higher education. The formation of stereotypes is another outcome of media discourse reception. Stereotypes can be defined as “associations of a social group concept with one or more (…) attribute concepts” (Greenwald et al., 2002, p. 5). They are usually misleading or even offensive, and their detrimental effects increase for infrequently encountered groups, such as refugees, as the possibility to verify the veracity of stereotypes through personal experience is limited (Saleem et al., 2017, p. 862). The effects of biased depictions of immigrants on the formation of stereotypes has been amply demonstrated in studies (e.g. Matthes & Schmuck, 2017; Igartua, 2013; Saleem et al., 2017; Dixon, 2017; Atwell Seate & Mastro, 2015; Arendt, 2017; Saleem et al., 2017; Devlin & Grant, 2017). 37 2.4. The Influence of the Media on Human Behaviours Studies have also shown that people rely on the media in order to make sense of events and act accordingly. Moreover, the underlying assumption that the media convey decisions made by knowledgeable groups (which is related to the gatekeeping trust) may even release them from self-prohibitive reactions to behaviours their moral codes of conduct normally forbids (Bandura, 2001, pp. 279-280). In particular, the media's justification of harmful conduct, diffusion of responsibility, minimisation of destructive consequences, blaming and dehumanising members of an out-group, as well as spreading negative attitudes and stereotypes about them may all contribute to enhancing the public's harmful behaviour by sparing them from self-deterring reactions (Bandura, 2001, pp. 279-280). For example, Saleem and colleagues (2017) have found a direct correlation between the portrayal of Muslims as aggressive and the public's support for military action in Muslim countries as well as for other policies intended to harm Muslims domestically and internationally. Research has also shown that violence is more likely to be perpetrated against ‘others’ who are represented in the media as either the perpetrators (Scheufele & Brosius, 2001) or victims of violence (Esser & Brosius, 1996). Other studies focus on voting behaviours. For instance, Burscher and colleagues (2015) found that the numbers of people voting for anti-immigration political parties rise alongside increases in the amount of media coverage on immigration. 38 2.5. Conclusions This Chapter has explored the relationship between media exposure and the audiences' perception of the world. Specifically, previous contributions indicate that the media heavily influence the cognition, perception, emotions, and behaviour of individuals as well as collectives. It is also important to note that all the media effects discussed above are interdependent: cognitions affect emotional reactions and vice versa, and both cognitive and emotional reactions may influence behavioural intentions (Kepplinger et al., 2012). What allows for the media to exercise such an — arguably unparalleled — impact on the audiences are high levels of media consumption, their cumulative effects, reduced personal experience with issues covered by the media, and the gatekeeping trust (Elsass et al., 2014; Lee, 2007; Stoycheff et al., 2018). 39 3. Factors Dictating Message Programming News production is not a straightforward operation: it consists of several different steps which are, in turn, continually impacted on by a variety of factors (e.g. Reese, 2001; Bläsi, 2004; Firmstone, 2008). The awareness of how issues are selected and information is gathered, what the writing and publishing stages involve, as well as who has to approve the material before it reaches the public should, arguably, be an important consideration in a discourse analysis. Therefore, this Section provides a brief overview of the processes which lead to the production and distribution of news stories. In general terms, the job of journalists consists of, first, selecting newsworthy stories out of the extensive amounts of raw material of global events; second, gathering information and corroborating it with different sources; third, making a story into an article; and finally, submitting the article for approval and publishing (Firmstone, 2008). In order to make this process as efficient as possible specific news routines (patterned, repeated practices of media workers) have had to be developed (Shoemaker & Reese, 1996; Becker & Vlad, 2009). These routines are further guided by the various constraints which can be grouped from the most micro to the most macro into: individual factors (journalists’ individual norms and values), organisational factors, and external influences (e.g. Reese, 2001; Becker & Vlad, 2009; Moon & Hadley, 2014). Because the various factors which influence the news production process may sometimes seem conflicting, Reese (2001) advocates considering them in terms of a hierarchy, in which each successive level subsumes the ones prior. The organisation of this Chapter follows this hierarchy. 40 3.1. Factors Dictating Message Programming on the Individual Level At the level of an individual reporter, their beliefs, values, and norms play an influential role both in the process of selecting and covering newsworthy stories (e.g. Bläsi, 2004, p. 3). With respect to the former, for example, Firmstone (2008) in her ethnographic study of ten British newsrooms has revealed a strong relationship between individual journalists' judgments about salience of issues and the amount of coverage these issues receive. With respect to the latter, Kovach & Rosentiel, in their highly referenced work (2003), emphasise how the norm of journalistic objectivity has been neglected, which has led journalists to express their stance towards issues in their texts. This claim has been corroborated, among others, by a study conducted by Soontjens (2019), which has revealed a significant increase in the proportion of interpretative newspaper articles published in the last three decades in Belgium. It was manifested in the presence of a journalist and their opinions, values, and beliefs in news items; the journalistic interpretation and focus on the “why” as the center of an article, while the factual “who – what – when – where” serve as support rather than basis; and the strong presence of evaluation of actors and events. Another study, conducted by Goyanes and Rodríguez-Castro (2019) in Spanish newsrooms, provided evidence that journalists introduce bias and censorship into news articles at their own incentive, seemingly without any external or internal pressure. The above results can arguably be extended to other democratic countries, as comparative research suggests that journalistic norms and values share similar characteristics across different elective democracies (Deuze, 2005). 41 3.2. Factors Dictating Message Programming on the Organisational Level At the level of media organisations, factors such as type of media and existing formats within a given medium as well as editorial strategies and procedures exert significant impact (Bläsi, 2004, pp. 2-3). For instance, journalists working for TV stations must put a stronger emphasis on obtaining video or photo material than those working for the radio (Bläsi, 2004, p. 3). However, it is time and financial constraints that, according to many scholars, exercise the strongest influence on the news production process (e.g. Matthews & Al Habsi, 2018; Moon & Hadley, 2014). With respect to the former, increasing importance is given to getting the information to the public quickly due to the 24-hour news cycle (Reese, 2001, p. 184; Soontjens, 2019, p. 957-958). At the same time, while the volume of newspapers has been increasing, the number of journalists working for them has been decreasing, which means that fewer journalists are tasked with producing more copy (Lewis et al., 2008). Furthermore, a growing amount of news is written by freelancers who, as a result of feeling insecure about their unstable incomes, may attempt to produce as much content as possible as fast as possible (Lecheler & Kruikemeier, 2015, p. 166). This, in turn, may lead to paying less consideration to source-checking and the veracity of facts, reusing old content instead of producing original items (Soontjens, 2019, p. 958), “borrowing” news from other media outlets (Riffe et al., 2018), and relying on “pre-packaged” material from official sources (Duguid, 2010). With respect to the latter, financial considerations determine news routines to an increasing extent due to the newspaper crisis (i.e. the fall in sales of print newspapers and the difficulty in adjusting their operations to an online model) (Goyanes & Rodríguez-Castro, 2019). The media are financially dependent mainly on 48 Gozálvez, 2004, p. 228; Bielsa, 2005, p. 142; Orengo, 2005; Lewis et al., 2008, p. 50); Cottle, 2009; Lee et al., 2015, p. 904; Morgan et al., 2015; Curran et al., 2017; Riffe et al., 2018; Matthews & Al Habsi, 2018, p. 749), even though they admit that it often requires adjusting to local contexts (e.g. Bielsa, 2005, p. 142; Orengo, 2005, p. 170; Lee et al., 2015, p. 904). Others, however, downplay the role of globalisation by emphasising the importance of national news outlets (including online media) and the significant variations in the coverage of events around the world (e.g. Lee et al., 2002; Allberg et al., 2013; Flew, 2018; Flew & Iosifidis, 2019; see also Curran et al., 2017, pp. 118-119). In fact, with respect to the coverage of the recent “refugee crisis”, the majority of previous contributions point to significant variations between different countries (it must be noted, however, that these results are, naturally, limited to the European context; e.g. Berry et al., 2015; Chouliaraki et al., 2017). Finally, the media are influenced by — and become some of the main carriers of — the dominant ideologies (e.g. Downey et al., 2014; van Dijk, 2015). The term “dominant ideology” refers to a set of ideas which “sociocognitively ‘ground’ the system of domination, and provide its rationale and legitimation” (van Dijk, 2015, p. 494). It presents dominant interests and ways of viewing the world as natural and universal, while those of the marginalised are excluded or presented as aberrant (Lynn & Lea, 2003, p. 427; Downey et al., 2014, p. 880). Specifically, with respect to media reporting on non-dominant groups such as migrants and refugees, ideologies provide the means by which subservience to the dominant group is justified and legitimised (van Dijk, 2000, p. 35). They are based on the racist belief that certain groups are innately superior to other groups based on their skin colour, culture, norms, and values, and that the two belong in separate geographical areas (Bowling & Westenra, 2018). Importantly, only the former are granted the “moving power”, i.e. 49 “the capacity to choose between moving and staying” (Allen et al., 2017, p. 222) . The 5 influence of ideology on the processes of news production subsumes all other factors, i.e. it may be realised at the individual, organisational, and extra-media levels. Furthermore, its impact is particularly significant due to its ubiquity, multifacetedness, and flexibility, as well as its evasiveness in discourse; furthermore, its presence is often explicitly denied or not discussed at all, especially in the case of racism (van Dijk, 1992; Lentin, 2008; Lentin, 2015). 3.4. Conclusions This Chapter has shed some light on the — largely obscure — factors which influence media message programming. Specifically, scholars point out that the public are to a significant extent unaware of the issues which shape the news production processes (Stoycheff et al., 2018, p. 186) such as journalists’ personal beliefs and values, and the media organisation’s procedures which are, in turn, constrained by financial and time concerns, as well as a variety of extra-media factors (e.g. Bläsi, 2004; Matthews & Al Habsi, 2018; Moon & Hadley, 2014). The latter include the clientelistic linkages between political and economic elites and the media, news values and imagined audiences, other media, as well as the wider processes of digitisation, globalisation, and the all-encompassing ideologies. Other scholars use the concept of mobility justice to explain “how power and inequality inform the 5 governance and control of movement, shaping the patterns of unequal mobility and immobility in the circulation of people, resources and information” (Sheller, 2018, p. 14). 50 4. The “Refugee Crisis” — Sociopolitical and Historical Context This chapter discusses “the broader sociopolitical and historical context in which the discursive practices are embedded and to which they are related” (Rheindorf & Wodak, 2018, p. 16) in order to secure a better understanding of the discursive outcomes in question, i.e. the press coverage of the “refugee crisis” in Poland, Spain, and the United Kingdom. Naturally, it does not claim to be a full and complete account of the historical and sociopolitical realities of the three countries of the analysis as this is outside of the scope of the present thesis; it does, however, attempt at providing a selection of contextual information “necessary for unambiguous interpretation” (Heylighen & Dewaele, 2002, p. 293), especially of the potential differences in the coverage of the “refugee crisis”. Specifically, it focuses on British, Polish, and Spanish differing historical experiences (especially experiences of contact with other nations), sociopolitical climates, geographic positions, contemporary attitudes towards immigrants, media systems, immigration policies, and statuses as host (Spain), destination (UK), and non-host countries (Poland) during the 2015-2018 “refugee crisis”. 4.1. Historical Context Historically, all three countries have considerable experiences of contact with other nations. With respect to Poland, while nowadays it is one of the least ethnically diverse European countries, it used to be inhabited by a gamut of minority groups of different origins and religions (Kundera, 2016, p. 109). Before the Second World War, minorities constituted almost one third of the total population of Poland, which was unparalleled even in the rest of the then very multicultural Central-Eastern Europe 51 (Kundera, 2016, p. 112). On the one hand, for a long time Poland acted as “an asylum for heretics” by providing refuge to people fleeing persecution they suffered in other parts of Europe, such as Bohemian Hussites, Russian Old Believers, or the Ashkenazi Jews expelled from the Holy Roman Empire (Juchimiuk, 2017, p. 306). On the other hand, Poland became ethnically diverse as a result of its expansion which began in the seventeenth century and slowly incorporated various smaller nations — as a result, for many centuries Polish domination reached from the Baltic to the Black Sea (Fiut, 2003, p. 152; Sanford, 2012, p. 4). While some scholars stress the principle of religious freedom which attracted minorities persecuted elsewhere (e.g. Juchimiuk, 2017), it is important to note that the multicultural Poland was by no means free of discriminatory practices, for example against Jews or Arians, which were often based on feelings of ethnic superiority (e.g. Sowa, 2014). The end of the eighteenth century marks the beginning of occupation of Poland first by Russia, Prussia, and Austria, and then by the Soviet Union from 1944 (after two decades of renewed independence between the two world wars). It was accompanied by mass emigration movements . Nazi war crimes (in particular against 6 Jews and Romani) further diminished the number of minority citizens in Poland (Stanford, 2012, p. 11). The movements of people resulting from the post-war shifting of borders, as well as the resolve of the authorities of the USSR-controlled Polish People’s Republic to build a more homogenous nation (for example by delegalising religions other than Christianity) also contributed to the decrease in the number of persons of foreign origin on the Polish territory (Sanford, 2012, p. 14; Kundera, 2016, Much of Polish culture and philosophy is deeply rooted in the experience and memory of political 6 emigration. It contributed to the image of an emigrant as “a patriot in exile fighting for his [or her] country”, which persists in Polish historical discourse to this day (Fomina & Kucharczyk, 2018, p. 197). Thus, in Polish patriotic thought politically motivated migration has a higher status than migration for economic reasons. 52 pp. 118-125). The historical processes which have shaped the Polish nation as ethnically homogenous and Roman-Catholic continue to influence its sociopolitical consciousness and remain deeply-rooted in public discourses (Sanford, 2012; Kopytkowska et al., 2017). Spain, too, has had a long and robust history of contact with other nations: often entwined with the history of Arabs, marked by the cohabitation with Jewish and Romani minorities and by the colonial conquest of America. With respect to its proximity to the Arab world, it has long been envisaged as the Southern rampart of “Fortress Europe”. Nowadays, this is particularly evident in the heavily-securitised Spanish cities of Ceuta and Melilla, located on the African continent. Spain has also retained a clear memory of Arab conquest and domination which in some parts of the Peninsula lasted more than seven centuries and has left an immense heritage in language and culture. Consequently, the Reconquista remains strongly associated with the development of the Spanish national identity (García Fitz, 2009). The year 1492 when Reconquista became complete was also marked by two other significant events related to the Spanish experience of contact with other nations: the expulsion of Jews from the Peninsula and the so-called discovery of America. With respect to the former, the persecution of Jews in Spain was based on the same tropes which could be observed across all Europe at the time (Yovel, 2018). With respect to the latter, although Spaniards emphasised the putative civilising mission of colonisation, the economic factor was clearly a strong motivation to organise overseas expeditions and to subsequently struggle determinedly to maintain them (Schmidt-Nowara, 2008; Sweet, 1997). Similarly, although Spain’s colonial subjects received far better treatment compared to the situation of slaves and the indigenous populations in other European colonies, especially the British ones 53 (Schmidt-Nowara, 2008; Lange et al., 2006), discrimination based on racism was deeply rooted in the colonial mentality (e.g. Lange et al., 2006; Sweet, 1997; Fuchs, 2005). With respect to the UK, it can, on the one hand, be defined in terms of the processes of annexations, unions, and separations of its constituent countries over hundreds of years; and on the other hand, in terms of its overseas conquests. With respect to the former, its evolution as a uniform state was a turbulent process, fraught with conflict and violence, and often accompanied by large migration flows (Richards, 2018). With respect to the latter, while Britain joined the European colonial endeavour relatively late, it soon surpassed its competitors in terms of territory and wealth, and by 1815 constituted the largest Empire in the world (e.g. Levine, 2013). Similarly to other colonial powers, its structure was also based on racial discrimination (albeit, as many scholars suggest, to an extent superior to those observed in other colonial territories [e.g. Schmidt-Nowara, 2008]). Both Spain and Britain resisted losing control over their overseas territories, which often led to violence against the emerging nationalist movements. Similarly, both continued to benefit economically from their unequal relations of power with the (former) overseas territories (Schmidt-Nowara, 2008), and neither made any effort to alleviate the consequences of their often sudden withdrawals (Levine, 2013, p. 229). For example, a study conducted by Mahoney (2003) points to the fact that the former centres of European imperial activity in South America nowadays face much more significant difficulties in economic and social development than the peripheries, and attributes it to the harmful legacy of imperial institutions and patterns of economic dependence, among others. Similar observations may be applicable to Africa (Mahoney, 2003, p. 51; Ocheni & Nwankwo, 2012). 54 This political, social, and economic instability contributed to the migration flows of the former colonial subjects to the former metropoles. As Levine (2013) mentionsą in the context of the British Empire: if the beginnings of the Empire may be said to lie with the migration of white settlers, the end of the Empire might likewise be characterised by the phenomenon of migration, but a reverse migration in which, after the Second World War, colonial subjects arrived in increasing numbers in Britain. (p. 224) Following an initial period of unregulated migration, restrictions were adopted (unless the country experienced shortages in labour) and the rights of immigrants were gradually limited (Levine, 2013; Richards, 2018). These tendencies have governed the immigration policies to this day, especially in the UK, as is discussed in Section 4.2. 4.2. Political Context This Section discusses the political climate during the “refugee crisis” in the three countries of the analysis as well as their migration policies. With respect to the former, as many scholars have noted, the “refugee crisis” coincided with — or, according to some, caused (e.g. Corbu et al., 2017, p. 6; Dines et al., 2018, p. 443) — a significant increase in the popularity of populist political parties in many European countries. Their mostly xenophobic, anti-immigration discourses are believed, in turn, to have impacted heavily on the public debate on migrations across the entire political spectrum (e.g. Berry et al., 2015, p. 13; Chan, 2019, p. 372). While Spain constitutes a notable exception to this wider trend, both the United Kingdom and Poland have seen a growing importance of right-wing, antiimmigration, and EU-sceptic political movements in the recent years. With respect to 55 the UK, this trend had a particularly noteworthy consequence, namely the successful 2016 campaign to withdraw from the European Union, where issues relating to race and migration figured prominently (Goodfellow, 2018; Virdee & McGeever, 2018; Koram & Nisancioglu, 2017) . Specifically, the proponents of leaving the EU managed 7 to convince the 52% of voters that immigrants were a threat to national identity, security, and economic wellbeing and that escaping the EU’s dictate would allow the country to regain control of its borders in spite of the fact that, as Favell and Barbulescu emphasise (2018, p. 5), the UK was free to shape its immigration policy independently from the EU. It is worth noting that this discourse remained unchallenged even by the Remain campaign (Komaromi, 2016). While some scholars claim that the rising levels of concern over immigration were mostly due to the opening of the borders to the 2004 and 2007 EU accession countries (Evans & Mellon, 2019, p. 78), others discern a clearly racial component, especially in the Leave campaign’s discourse on the “white working class” (Antonucci et al., 2017; Bhambra & Holmwood, 2017). It postulated that this particular social group was being “left behind” and therefore suffered the most severe consequences of the economic austerity because people of colour were gaining a disproportionate advantage (Antonucci et al., 2017; Bhambra & Holmwood, 2017). In other words, as a result of too much attention being devoted to the improvement of the living conditions of non-white citizens, the needs of white workers were neglected, and thus their wellbeing decreased. Working and living conditions of the lowest segment of the British society were, in fact, in decline, both for white and ethnic minority workers (Bhambra, 2017, p. S221). However, it was the latter rather than the former who remained at a disadvantage with rising unemployment (which, at the same time, was This is also evidenced by the fact that the perceived salience of immigration rose significantly in 2015 7 only to fall back to the previous levels after the referendum (Blinder & Richards, 2020, p. 4). 56 falling for white people), deteriorating conditions at work, and lowering wages (Goodfellow, 2018, p. 150). The discourse of “white working class left behind” was also used post-Brexit to explain the voting result by commentators from the entire political spectrum (Bhambra, 2017), even though no correlation between voting leave and selfidentifying as working class had been found (Antonucci et al., 2017). Instead, studies suggest that the middle class played a much more significant role in delivering this result, as it constituted 59% of all Leave votes, whereas only 24% were cast by the lowest social classes (Dorling, 2016). Furthermore, as an investigation conducted by Flemmen and Savage (2017) has revealed, it was the elites (economic and cultural) who were most likely to express imperial racism, which was the main undercurrent of the Brexit campaign (e.g. Abbas, 2019). In a similar way, Evans and Mellon (2019, p. 77) notice with respect to the rise of the United Kingdom Independence Party, the main Leave proponent, that its success "could never have been achieved without substantial support from within the professional and managerial middle classes”. The “refugee crisis” also figured prominently in Brexit discourses, even though the UK accepted hardly any refugees (Bhambra & Holmwood, 2017, p. 574; Favell & Barbulescu, 2018, p. 5). In fact, many scholars emphasise that the UK has a particularly draconian immigration and asylum system characterised by “restrictive policies, absurdly complex and ever-changing Immigration Rules, harsh and arbitrary decision-making, criminalisation of mobility and indefinite immigration detention” (Griffith & Yeo, 2021, p. 4; Goodfellow, 2018; Berry et al., 2015, p. 11; Bloch, 2001). Its aim is to discourage people from migrating to and settling in the country, as attested by the range of border securitisation measures with regard to the 57 former, and the “hostile environment” policies, with respect to the latter (Mayblin, 8 2019; Griffith & Yeo, 2021). For example, the UK has the largest detention estate in all Europe (Goodfellow, 2018, p. 155) with no upper time limit of detention, which is extremely considered harmful to detainees (Coffey et al., 2010; Silverman & Griffiths, 2019). It prohibits asylum seekers from entering the labour market, thus making them dependent on welfare payments, which amount to a third of the income of the poorest 10% of UK households (Mayblin, 2019, p. 7). Over the last two decades, a variety of anti-immigration measures were introduced such as, for example, deporting immigrants based on police contact rather than conviction or revoking driving licences and closing bank accounts of people without sufficient leave to remain in the country (Griffith & Yeo, 2021). What is more, evidence regarding the displaced people which does not come from the Home Office (e.g. scholarly or public service investigations) is discarded as unreliable (Mayblin, 2019, pp. 13-17) even though, as Griffith and Yeo (2021) have shown, the official data is often extremely flawed (pp. 14-15). Ultimately, the “hostile environment” has been shown to generate substantial human, social, and financial costs, with risks to safety, security, and health, without, however, impacting negatively on migration numbers (Griffith & Yeo, 2021, p. 11). In a similar way, Poland has also seen a rise of anti-immigration movements which, however, prior to the “refugee crisis”, were practically absent from its political and public agenda (Fomina & Kucharczyk, 2018, p. 187; Krzyżanowski, 2018, p. 76). This was mostly due to the fact that Poland used to be a country of emigration rather than immigration (e.g. Krzyżanowski, 2018, p. 79; Górak-Sosnowska, 2019, p. 297). Following the post-1989 opening of Polish borders immigrants from lower-income The phrase, currently used with respect to immigration, originally referred to dangerous overseas 8 locations such as warzones (Griffith & Yeo, 2021, p. 4). 64 expressed support for closing the borders. At the same time, however, Spaniards were the least likely among European nations to link refugees with terrorism, and to doubt that refugees were really refugees (Wike et al., 2016; Ipsos Mori, 2018). Similar results from the Pew Spring 2018 Global Attitudes Survey were observed for the United Kingdom, where only 37% claimed that immigration should be restricted, and only 17% considered immigration to be one of the most important issues facing their country in 2018; nevertheless, it is important to note that this number decreased from 44% at the height of the “refugee crisis” (Silver, 2018, using Eurobarometer surveys conducted between November 2014 and March 2018). A vast majority claimed that they would be comfortable having an immigrant as a neighbour, friend, work colleague, doctor, family member, or supervisor (Special Eurobarometer, 2018). On the other hand, however, few believed that the country should allow more immigrants to move to their country (16%) (Connor & Krogstad, 2018), almost half considered the number of immigrants in their country to be too high, and 43% felt uncomfortable with the changes brought on by immigration (Ipsos Mori, 2018). When asked about refugees, more than one third agreed that the country could not accept any; 74% believed that there were terrorists among refugees; and more than half doubted that refugees were really refugees (Ipsos Mori, 2018). With respect to Poland, in the same manner that anti-immigration political movements and discourses were absent from the public sphere prior to the year 2015 (see Section 4.2 above), the general population also did not harbour strong antiimmigrant feelings; in fact, support for immigration had been stable for some time (Heath & Richards, 2016, pp. 4-5). In this respect, too, the change was sudden and significant. According to the data gathered in 2018, almost half of Poles said that 65 fewer immigrants should be allowed to settle in their country, whereas only 9% were in favour of more immigration (Connor & Krogstad, 2018). A quarter of the respondents said they would not be comfortable having an immigrant as a friend, neighbour, or work colleague; this number rose to a third when they were asked about a doctor, and almost 50% expressed aversion to having an immigrant manager (Special Eurobarometer, 2018). Whereas only a third of Poles believed that there were too many immigrants in their country, 41% said that they were uncomfortable with the changes induced by immigration (Ipsos Mori, 2018). With respect to closing the borders to refugees, Poland was among the countries with the highest support for this proposition (51%); in Europe it was surpassed only by Italy and Hungary (Ipsos Mori, 2018). A significant majority of Poles believed that terrorists were pretending to be refugees, and that refugees were not really refugees (Wike et al., 2016; Ipsos Mori, 2018). Similarly, the majority in Poland (but also in the UK) saw in refugees a major threat to their country (Wike et al., 2016). The above-quoted studies as well as the reviews conducted by Dempster and Hargrave (2017) and Dempster and colleagues (2020) indicate that generally in Europe (as well as across the world ) attitudes are more positive towards refugees 12 than immigrants or asylum-seekers, and that there is a clear preference for migrants who use regular means of entry as well as for the high-skilled and well-educated (although respondents are unsure whether immigration really has a positive impact on the economy [Dempster et al., 2020, p. 14]). Furthermore, Europeans are clearly more opposed to non-white and more culturally distinct immigrants (Heath & Richards, 2016, p. 6; Dempster & Hargrave, 2017, p. 11; Blinder & Richards, 2020, It is important to note that Europe has for a long time been exceptionally negative towards 12 immigration compared to the rest of the world (Dempster et al., 2020, p. 12). 66 pp. 5, 7; De Coninck, 2020; Dempster et al., 2020, p. 14) . Two groups are 13 particularly vulnerable to negative attitudes: the Roma and Muslims (Heath & Richards, 2016; Wike et al., 2016). Negative sentiments towards the latter have become more popular recently (Wike et al., 2016, p. 9), and are especially strong in Poland, Austria, Hungary, Belgium, and France (Goodwin et al., 2017). Europeans tend to believe that Muslims want to be distinct from the larger society — this includes almost three quarters of respondents in the generally more tolerant Spain (Wike et al., 2016, p. 25). There is also significant support for the proposition that at least some Muslims approve of terrorist groups: almost two thirds of respondents in Poland and Spain, and almost half in the UK believe this (Wike et al., 2016, p. 28). Notoriously, older respondents are more negative; it should be emphasised, however, that this is a generational effect (i.e. people are not becoming more negative with age but rather different generations’ attitudes are shaped by different factors) (Winkler, 2015). Stronger anti-immigrant attitudes are found on the right end of the ideological spectrum (Wike et al., 2016) as well as among the less-educated and those with more modest incomes (Heath & Richards, 2016). There are mixed reports with respect to the impact of gender and religion: while Heath and Richards (2016) show that these factors do not play any role in determining attitudes towards immigrants, according to the study conducted by De Coninck and Matthijs (2020) in four different European states, men and Christians appear to be more biased against immigration. As becomes clear when results of different studies are juxtaposed, examining public attitudes towards immigrants and refugees is not a straightforward task due to the inconsistencies and sometimes even contradictions in the polling data (especially so in the case of Spain and the UK). An explanation offered by Dempster and According to a different study, however, respondents in both the UK and Spain were equally 13 concerned about EU and non-EU immigration (Blinder & Allen, 2016, p. 6). 67 Hargrave (2017, p. 12) postulates that, in these countries, the largest segment of the population falls with the category of “conflicted” or “anxious” middle. The “anxious middle” are “less ideologically motivated than groups most confidently ‘for’ or ‘against’ immigration, and much more ambivalent towards refugees and migrants and their impacts on society” (Dempster & Hargrave, 2017, p. 12). Their answers to survey questions may thus be inconsistent and conflicting, and often fall in the middle of the Likert scale (i.e. these individuals are less likely to “strongly agree” or “strongly disagree” with a given proposition). For example, this might explain why the vast majority of a population would not mind being treated by an immigrant doctor or having an immigrant neighbour, whereas support for accepting more immigrants is significantly lower. Similarly, based on evidence from the UK, Ashcroft (2013) demonstrated that respondents are unlikely to believe that immigrants contribute to making their area a better place to live, but at the same time, many recognise their contributions to the National Health Service. This, as Dempster and Hargrave (2017, p. 12) emphasise, “underscore[s] the complex web of concerns around which public attitudes towards refugees and migrants are formed” as well as creates important opportunities for counter-discourses to influence these attitudes. On the other hand, it also demonstrates how impactful anti-immigration discourses may be in swaying the public opinion towards more extreme views. Furthermore, when interpreting survey results, several important issues should be taken into consideration. First of all, polling data is notoriously “sensitive to ambiguities and bias in question wording and ordering, and vulnerable to changes in methodologies and timing” (Dempster & Hargrave, 2017, p. 9). For example, Blinder and Allen (2016, p. 4) discuss how dramatically the responses to a question changed depending on how it was worded. Similarly, the lack of consistent, 68 commonly agreed-on definitions of the key terms presents a challenge. For instance, in the UK context, statistics usually count as immigrants people coming to the country for at least one year (Blinder & Allen, 2016, p. 9). However, respondents may intuitively define seasonal workers, “second generation migrants”, or representatives of different minorities in general as immigrants. The conflation of terms such as “migrant”, “refugee”, “asylum-seeker” in public discourses (Baker et al., 2008) further adds to the confusion. As a result, answers given by survey respondents may be based on images of specific groups (such as immigrants or refugees) which highlight only some of their defining features while downplaying others, and which might be different for different respondents (Blinder & Allen, 2016, p. 9). Second of all, responses are likely to be affected by the mode in which the poll is conducted (on-line, phone, or face-to-face), the person asking the questions (e.g. woman versus man), and by the desirability bias: respondents are often unwilling to disclose their true views on topics deemed controversial, but instead answer in ways that they consider to be preferred by the interviewer (Drouhot et al., 2022, p. 8). Third of all, national polling data hides geographical differences (Dempster & Hargrave, 2017, p. 9). It should also be understood in context, whereas few studies provide more detailed examinations of attitudes towards immigration beyond selected countries of the global North (Dempster & Hargrave, 2017, p. 9). Altogether, this indicates that survey results should be approached with caution as they may prove to be unable to truthfully represent the complex realities of public attitudes. 69 4.4. Media Landscape Finally, for the purpose of the present study it is important to briefly describe the main characteristics of the media systems of Poland, Spain, and the UK as they might also influence discourse production and reception (see Chapter 3). While this Section is mainly concerned with a more general overview of the media landscapes in question, more detailed information regarding the particular newspapers which compose the analysed corpus is provided in Chapter 6. Above all, it is important to note that the three countries have very different traditions and practices related to the mass media. For example, according to Hallin and Mancini’s (2004) oft-quoted categorisation of Western media systems into three models, Poland, Spain, and the UK each belong to a separate category . 14 With respect to Poland, its current media landscape was largely formed following the fall of communism in 1989 (e.g. Klimkiewicz, 2021, p. 56). The lack (or insufficiency) of rational regulations during that period of system transformation (Bielawski, 2010, p. 25) resulted in several notable consequences for the media. First, they became rapidly commercialised as no barriers against foreign investment were erected — as a result, the media market was to a significant extent seized by foreign corporations at the beginning of the 1990s. (Szynol, 2010). Currently, private foreign corporations have the main share of the Polish media market (e.g. Klimkiewicz, 2021, p. 54). Second, the sociopolitical climate of the transformation period decidedly encouraged right-wing views, whereas leftist inclinations were ill-favoured, which led to the lack of balanced representation of different political views in the press (Kubiak, Other scholars, however, argue that Poland is not such a clear-cut case but rather combines elements 14 from Hallin and Mancini’s (2004) liberal and polarised pluralist media system categories (DobekOstrowska, 2015, p. 27). 70 2018, p. 104; Szynol, 2010, p. 84; Gulyás, 2003, p. 86). At the same time, however, right-wing actors have often claimed that their point of view is underrepresented in the Polish media (Klimkiewicz, 2021, p. 56). For example, the current right-wing government tends to justify the increasingly partial reporting of the public media as a necessity to balance the “liberal bias” (Klimkiewicz, 2021, p. 56). This, arguably, exacerbates the high polarisation of the Polish media (Klimkiewicz, 2021; Peruško, 2021, p. 46). Furthermore, a majority of Poles are convinced that the media are not objective, and as many as 80% report being unsure about the truth about events due to the vast discrepancies in the different media coverage (Centrum Badań Opinii Społecznej [‘Centre for Public Opinion Research’], 2019). Third, in many ways the pathologies of the former regime persisted under the new democratic system, which has resulted in a relatively low quality of journalism, characterised by significant political influence, as well as a growing popularisation and sensationalisation of news coverage (Bielawski, 2010, p. 25; Esser et al., 2016, pp. 3-4; Szynol, 2010; Klimkiewicz, 2021). Tabloids currently enjoy the biggest circulation on the market. For example, Fakt, introduced in 2003, in just two months grew to be the most-read newspaper and continues to maintain unsurpassed readership. It is closely followed by another tabloid — Super Express. At the same time, the broadsheet Gazeta Wyborcza, the first post-transformation Polish newspaper, has been noting significant decreases in readership (Związek Kontroli Dystrybucji Prasy [‘National Circulation Audit Office’], 2019). The Spanish transition into democracy was a much more gradual process than that of Poland, and it accorded a vital role to the press. According to Gunther and colleagues (1999, p. 14), the press aided the process by acting as “conduits for information about the strategy for political change (…), as well as platforms for the 71 articulation of political demands by newly emerging political and trade union organisations”. For example, the principal stockholders of El País, which grew to be the most-read daily in Spain, as well as the world’s leading newspaper published in Spanish, “included representatives of all of the ‘political families’ which would govern during the transition to democracy” (Gunther et al., 1999, p. 13). This demonstrates the importance of the tight cooperation between politics and the press during that period. The press in Spain continues to have strong political ties to this day, made even more significant by its financial dependence upon the state, which might suggest a high level of instrumentalisation (Hallin & Mancini, 2004, p. 67; Papatheodorou & Machin, 2003, p. 36). It is also considered to be exceptionally polarised with the three market-leading newspapers: El País (PSOE) on the one hand, and ABC as well as El Mundo (Partido Popular) on the other, constituting the two opposing camps (Papatheodorou & Machin, 2003, p. 41; Hallin & Mancini, 2004, p. 130) . 15 Importantly, no tabloids or sensationalist popular dailies are published in Spain. As a result, similarly as in the political realm, no outright mouthpieces for extremist or racist discourses exist (Hallin & Mancini, 2004, p. 97). Attempts by foreign investors — including the German publisher Axel Springer, as well as the British Daily Mirror — to establish sensationalist newspapers in Spain have failed repeatedly (Papatheodorou & Machin, 2003, p. 41; Hallin & Mancini, 2004, p. 97). This has, arguably, led to much higher standards in Spanish news coverage. In fact, one cross-national study found the lowest percentage of infotainment in Spanish press (Jebril et al., 2013). This also translates into much higher levels of trust towards journalism than, for example, in Poland or the United Kingdom (Splichal & Dahlgren, The fourth most-read newspaper is the Barcelona-based La Vanguardia, ideologically close to the 15 Catalan Nationalist CiU (Hallin & Mancini, 2004; Córdoba Hernández, 2009). 72 2016, p. 11), even though there is no official institution responsible for press control and regulation (Franquet Dos Santos Silva et al., 2018, p. 513). On the other hand, however, the lack of more accessible, less sophisticated press may have led to a much lower daily newspaper readership than in other European countries. Spain has one of the lowest newspaper reading indexes in Europe (Salaverría et al., 2005). The elitism of Spanish press may be accounted for by Spain’s relatively late economic development and the inadequate level of education under Franco which has resulted in high rates of illiteracy, especially among women (Papatheodorou & Machin, 2003, p. 35). Furthermore, reading newspapers may also be considered as a sign of political involvement, which may discourage some from seeking information in this source (Papatheodorou & Machin, 2003, p. 41; Hallin & Mancini, 2004, p. 67). Contrary to the Spanish case (as well as other European countries), the UK national media system is unusual in the central position it awards to popular press which to a large extent dominates over quality press (Hallin & Mancini, 2004, p. 159; Taylor, 2014, p. 376). Furthermore, even the so-called “quality newspapers” may to some extent resemble tabloid press. For instance, as Baker and colleagues (2013, p. 7) note, “The Guardian has a daily supplement, referred to as a ‘tabloid section’, that focuses on more populist reporting of celebrity gossip and quirky and ‘human interest’ stories”. As a result, the degree of sensationalisation and commercialisation of the UK press is considered to be unusually high (Hallin & Mancini, 2004, p. 247; Jebril et al., 2013). For example, Jebril and colleagues (2013) in their cross-national study found the highest percentage of infotainment in British media. In a similar vein, Van Dalen et al.’s 2012 study has revealed that British political journalists are more entertainment-oriented that their counterparts from other Western 73 democracies. What is more, the UK press is frequently denounced for its strong political ties. Many scholars posit that it serves the interests of the elite rather than the public, and openly campaigns for political parties (Hallin & Mancini, 2004, pp. 211-212; Baker et al., 2013, p. 8). Especially tabloids manifest strong political orientations, usually in “the form of a right-wing populist stance, emphasising nationalism, anticommunism, traditional views on gender and on many social issues” (Hallin & Mancini, 2004, p. 211). Their stance often involves explicitly discrediting, and thus fuelling, of racial inequalities within the society (Hallin & Mancini, 2004, p. 212). The UK’s regulatory body, the Press Complaints Commission has often been denounced as largely ineffective in countering misleading and discriminatory news coverage (Baker et al., 2013, pp. 11-12), or even as “part of the problem rather than part of the solution” as it is financed by newspapers themselves, and its Code Committee consists of those newspaper editors it is expected to oversee (Petley, 2006, p. 61). These observations seem particularly significant considering the fact that the potential of the UK press is by no means limited to domestic readers as the high status of the English language and the relative appeal of its culture, combined with online access to major national newspapers, make it a particularly attractive news source for speakers and learners of English around the world (Baker et al., 2013, p. 5). 80 6. Corpus The corpus compiled for the purpose of the following analysis contains 6,379,921 tokens, which should allow for examining the usage of high frequency words, including the most frequent lexis (Baker, 2009, p. 314). It consists of news articles published in four market-leading national paid newspapers per each country of the analysis, i.e. twelve newspapers in total. The monthly circulation was the sole 17 criterion of choice of news outlets, with the oft-applied criteria of political inclinations or reporting style (e.g. Georgiou & Zaborowski, 2017; Cihan KocaHelvaci, 2019; Pstrąg, 2022) left out of the account as they could not have been applied consistently to newspapers published in different national contexts (see Abdelhady, 2018, p. 3 and Partington et al., 2013, p. 191 for a similar argument). Studying market-leading newspapers allows for drawing as generalisable results as possible with regards to how the press may influence public perceptions of a specific event or a (group of) actor(s), in this case the “European refugee crisis” and the displaced people. The process of corpus compilation consisted of two stages. First, texts were downloaded from relevant databases (EBSCO , My News ) or the newspapers’ 18 19 digital archives . The search key terms were: refugee*, immigrant* and migrant* 20 for the UK subset; refugiad*, inmigrante*, and migrante* for the Spanish subset; and uchodźc*, imigran*, and migran* for the Polish subset (where * is a wildcard, The monthly circulation was established based on the data from Statista Research Department for 17 the year 2018 for the UK and the Spanish samples and Związek Kontroli Dystrybucji Prasy [‘National Circulation Audit Office’] (2019) for the Polish sample. https://www.ebsco.com/ 18 https://mynews.es/ 19 https://classic.wyborcza.pl/archiwumGW/0,0.html?disableRedirects=true; https://www.se.pl/; 20 https://www.fakt.pl/; https://archiwum.rp.pl/; https://www.theguardian.com/gnm-archive; https:// www.mirror.co.uk/; https://www.thesun.co.uk/ 81 which stands for zero or more characters). Next, the results were subjected to a second stage of selection: to ensure that the “refugee crisis” was topical, only news entirely or mainly related to it were included in the final version of the corpus . This 21 was done manually. Columns and letters to editors expressing personal views of the authors, which are “less accountable” and do not necessarily concur with the newspaper’s general stance (Baker et al., 2013, p. 12), were excluded from the corpus as well. The composition of each language subcorpus is detailed in the next three subsections. 6.1. The Spanish Subcorpus The Spanish subcorpus includes the following titles: El País, El Mundo, ABC, and La Vanguardia. All of them can be classified as broadsheet dailies (tabloid publications are absent from the Spanish press market), and are highly polarised ideologically (as mentioned in Section 4.4). The most read newspaper in Spain both in print and online, El País, has since its beginnings been ideologically close to the Spanish Socialist Party — PSOE (Córdoba Hernández, 2009, pp. 226-234). On the other end of the spectrum, there are El Mundo and ABC, second and third newspapers with the biggest circulation, both associated with the Popular Party (Partido Popular). However, ABC much more markedly displays its españolismo, liberal conservatism, catholicism, and loyalty to the monarchy, whereas El Mundo tends to position itself as an independent, or even nonconformist voice within the Spanish media landscape, despite its visibly right-wing inclinations (Córdoba Hernández, 2009, p. 234-246). Examples of texts which were discarded at the second stage of corpus compilation included but were 21 not limited to: news on migrations towards the USA or Australia, profiles of famous people with migrant background, news where the “refugee crisis” was only mentioned briefly, e.g. when it was named as one out of many topics of discussions during a political campaign, etc. 82 The fourth most-read newspaper in Spain is the Barcelona-based La Vanguardia, ideologically close to the Catalan Nationalist CiU (Córdoba Hernández, 2009; Mármol Queraltó, 2021, p. 51). With respect to national politics, its stance is not dissimilar to the views embraced by El País; however, it tends to employ a rather neutral tone in its reporting, aiming to appeal to a wider range of readers (Córdoba Hernández, 2009, pp. 246-253). With respect to the percentage of texts published by the different newspapers in the subcorpus, the biggest number of articles was collected for El País, followed by El Mundo, as can be observed in Table 1.1. However, with regard to the token count, they are both superseded by ABC, which indicates that the articles published in the latter are longer. In total, the ABC articles account for close to 30% of all the tokens in the subcorpus, whereas those published in El País and El Mundo — 27% and 24%, respectively. Finally, La Vanguardia accounts for 20% of the Spanish subcorpus, with both the lowest number of texts and tokens. The total number of texts in the Spanish corpus is 3,086 (2,478,336 tokens). Table 1.1. Composition of the Spanish Subcorpus Title ABC El País El Mundo La Vanguardia Total number of texts 898 1,010 953 746 Total token count (percentage within the subcorpus) 719,449 (30%) 661,311 (27%) 590,051 (24%) 507,525 (20%) 83 6.2. The Polish Subcorpus The Polish subcorpus includes two broadsheet dailies — Gazeta Wyborcza and Rzeczpospolita (along with its Sunday edition Plus Minus) — as well as two tabloids — Fakt and Super Express. With respect to their monthly circulation figures, the distribution of readership among Polish daily newspapers has for a long time remained largely unchanged (insignificant fluctuations notwithstanding). The two tabloids consistently attract the biggest numbers of readers of both their print and online editions (while Fakt sells more print newspapers, its immediate competitor attracts more online visitors), whereas the two most popular quality newspapers invariably trail far behind. The market of quality press in Poland is divided between them across ideological lines, with the slightly more popular Gazeta Wyborcza appealing to the left-inclined readers. The Polish subcorpus contains 2,089 texts in total (1,276,991 tokens), of which the biggest number were retrieved from Gazeta Wyborcza (898 texts), followed by Fakt (741). With respect to the token counts, however, the disproportion between the two sub-subcorpora is much more significant due to the average length of texts published in the two outlets, as can be observed in Table 1.2. With respect to the remaining two newspapers — Rzeczpospolita and Super Express — only 259 and 191 texts, respectively, were available in their digital archives. 84 Table 1.2. Composition of the Polish Subcorpus 6.3. The UK Subcorpus Similarly as in Poland, the most popular national dailies in the UK are tabloids with The Sun, Daily Mail , and Daily Mirror occupying the first three positions of the 22 most-read, paid national newspapers in the United Kingdom. Their Sunday editions — The Sun on Sunday, Mail on Sunday, and Sunday Mirror, respectively — have also been included in the corpus. The fourth bestselling daily in the UK, according to Statista’s (2018) data, is the broadsheet Guardian (with The Observer as its Sunday counterpart). It is also, along with the BBC, the most successful online news provider in the UK (Lewis et al., 2008). With respect to their political affiliations, as Baker and colleagues (2013, p. 8) note, British newspapers are partial, especially to right-wing views, to an extent unseen elsewhere in Europe. Tabloids in particular manifest strongly their political orientations (Hallin & Mancini, 2004, p. 211). However, the corpus collected for this analysis can be, in this respect, considered to be more or less balanced, with only two of the four newspapers embracing explicitly right-leaning views: The Sun (nevertheless, it caters to a very diverse readership, which includes minorities, and is therefore careful around issues relating to race and immigration [Philo, 2007, p. 190]) and Daily Mail (which is, in fact, perceived as the most rightTitle Gazeta Wyborcza Fakt Rzeczpospolita + Plus Minus Super Express Total number of texts 898 741 259 191 Total token count (percentage within the subcorpus) 705,327 (55%) 297,486 (23%) 194,668 (15%) 79,510 (6%) Daily Mail, however, is sometimes referred to as a middle-market newspaper rather than a tabloid 22 (Baker et al., 2013, p. 69). 85 wing daily newspaper in the UK [YouGov, 2017]). The other two — Daily Mirror and The Guardian - can be classified as left-leaning (YouGov, 2017; Baker et al., 2013, p. 9). The UK corpus of news on the “refugee crisis” contains 2,624,594 tokens spread over 2,138 news articles. Table 1.3, which displays the distribution of the articles and tokens among the four newspapers, shows that the only broadsheet newspaper in this subcorpus accounts for the biggest number of texts and tokens. Specifically, 38% (measured in tokens) of this subcorpus consists of texts published in The Guardian. The Sun and Daily Mail, on the other hand, follow a similar distribution of both texts and tokens and account for 24 and 25% of the subcorpus, respectively. Finally, Daily Mirror is represented in 13% of the subcorpus. Table 1.3. Composition of the British Subcorpus Once collected, the entire corpus was uploaded, lemmatised, and POS-tagged in Sketch Engine (Kilgarriff et al., 2004; Kilgarriff et al., 2014). The subsequent analytical procedures, along with their theoretical foundations, are described in detail in the following Chapter. Title The Guardian & The Observer Daily Mail & Mail on Sunday The Sun & The Sun on Sunday Daily Mirror & Sunday Mirror Total number of texts 724 576 590 248 Total token count (percentage within the subcorpus) 990,646 (38%) 649,240 (25%) 640,995 (24%) 343,713 (13%) 86 7. Theoretical Framework and Methods The present study is conducted using a triangulation of two methodologies, the validity of which has been assessed satisfactorily by numerous previous contributions (e.g. Baker et al., 2008; Baker & Levon, 2015; 2016; Marchi & Taylor, 2009; Taylor & Del Fante, 2020; among others). Specifically, it applies quantitative methods associated with corpus linguistics to the entire dataset, and argumentation analysis, as it is conceived of within the larger theoretical framework of Pragmatic Linguistics (Fuentes Rodríguez, 2013; 2017), to a sample of the corpus. This Chapter discusses their theoretical underpinnings, describes the adopted sampling and analytical procedures, and, finally, mentions briefly the potential benefits of methodological triangulation. 7.1. Corpus Linguistics Corpus linguistics can be defined as “the use of a large body of electronically coded text which is subjected to analysis of linguistic patterns via computer software” (Baker & Levon, 2015, p. 222). This analysis was conducted, using Sketch Engine (Kilgarriff et al., 2004; Kilgarriff et al., 2014), on a specially constructed multilingual corpus of press articles of more than six million tokens (as discussed in Chapter 6). Three main techniques of exploring linguistic patterns in a corpus, which are central to corpus linguistics (Stefanowitsch, 2020, p. 53), were employed, namely concordances, collocations, and frequency lists. It has been suggested (e.g. Ajšić, 2021, p. 89) that corpus linguistics techniques work best when they are used in conjunction with one another. 87 The first phase was the analysis of collocation, which can be defined as “a way of demonstrating (relatively) exclusive or frequent relationships between words (or other linguistic phenomena)” (Baker, 2010, p. 24). In other words, collocation analysis indicates items that have a statistically salient connection with a given node. Therefore, it points at the most prominent patterns in a given discourse as “no matter how suggestive, an association which is not repeated enough will be less influential than an association that is more firmly established in the discourse” (Brezina et al., 2015, p. 160). Consequently, collocations are often used to give the most general and comprehensive idea of the analysed discursive phenomenon. Furthermore, by showing which words frequently occur near other words, collocations can help access additional information about their contextual meanings, including connotational and evaluative potential. There are two key concepts in this area: semantic preference and discourse prosody. The former is “the relation, not between individual words, but between a lemma or word-form and a set of semantically related words” (Stubbs, 2001, p. 65). For example, both Baker (2006) and Taylor (2018) have demonstrated a strong semantic preference of words such as migrant, refugee, asylum-seeker, etc. for quantification (realised using a wide range of different linguistic means). Discourse prosody, on the other hand, tends to be more suggestive of attitudes, and is often described as “a consistent aura of meaning with which a form is imbued by its collocates” (Louw, 1993, p. 157). It is usually evaluated as positive or negative (e.g. Partington et al., 2013) although, as Bednarek (2008, p. 133) notes, “it should be kept in mind that there are many semantic prosodies that do not relate to ‘(un)pleasantness’, and that evaluation is much more multi-faceted” (see also Grabowski [2022, p. 121] for a similar argument). For example, Baker and Levon’s 88 (2015) analysis of news articles about masculinity found a tendency to portray black men as criminals and Asian men as engaging in sexual and violent behaviour with underage white women (both are examples of negative discourse prosodies), whereas white men were often construed as victims. Taken together, as Mautner (2001, p. 161) summarises, semantic preferences and discourse prosodies show us what kinds of social issues a particular lexical item is bound up in, and what attitudes are commonly associated with it. Importantly, collocational patterns are not merely instantiated in text, but also cling to the lexical items themselves. This suggests that they are not limited to subjective associations but point to more widespread attitudes and beliefs within a given speech community. Furthermore, when an item becomes imbued with a particular set of connotations, these will be triggered even when the item in question is not accompanied by the particular collocate(s) and when its apparent meaning in a given context is neutral. This is referred to as lexical priming by Hoey (2005), and has been supported by numerous psycholinguistic experiments (e.g. Durrant & Doherty, 2010; Millar, 2011). For the collocation analysis, Sketch Engine’s word sketch function was employed, which displays “a one-page summary of a word’s grammatical and collocational behaviour” (Kilgarriff et al., 2014, p. 9). To identify collocates of a word, it uses the span of five words either side of the search term . It then arranges them in 23 columns according to the strength of their collocation, where each column represents a different grammatical relationship. The measure of determining the statistical Baker and colleagues (2013, p. 36) estimate that the five word span ‘offer[s] a good balance between 23 identifying words that actually do have a relationship with each other (longer spans can throw up unrelated cases) and giving enough words to analyse (shorter spans result in fewer collocates)’. 89 significance of a collocation chosen for this study was logDice score which, according to Jaworska and Kinloch (2018, pp. 115-116) seems particularly suitable for comparing collocations across corpora. In contrast to other commonly used statistics such as Mutual Information or T-test, Log Dice is a ratio with a maximum value (theoretically 14, but practically 10 or below) and it does not depend on the total size of the corpus (Rychlý 2008). This allows the researcher to have a consistent comparison measure across multiple data sets. This is particularly important in the context of the present study, which is based on a corpus composed of three language subsets of different sizes. Furthermore, logDice highlights exclusivity, i.e. “the extent to which the two words appear solely or predominantly in each other’s company, usually expressed in terms of the relationship between the number of times when they are seen together as opposed to the number of times when they are seen separately in the corpus”. (Gablasova et al., 2017, p. 160) This means that logDice highlights those collocates that are particularly likely to be associated with the given node even if it appears alone in text in the same way that, using Gablasova et al.’s (2017, p. 160) example, zig brings to mind zag. Consequently, words that tend to appear in each other’s company "may be more easily recognized, acquired, and stored as a unit” (Gablasova et al., 2017, p. 160). Finally, contrary to other association measures, logDice does not penalise frequency, i.e. it does not prioritise rare combinations such as proper names, and specialised or technical expressions (Gablasova et al., 2017, p. 164). Therefore, it is suitable for indicating collocations formed by more common words. 96 reinforce the receiver's view of the world, their opinions and beliefs . In other words, 27 all discourse has the potential of convincing the receiver to think, believe, or act in a particular way (Fuentes Rodríguez & Alcaide Lara, 2002, p. 9). News reporting, in spite of its claim to objectivity and neutrality, is not an exception: in fact it possesses a particularly pronounced argumentative function which is linked to the media’s economic, political, and ideological agenda (see Chapter 3). As argumentation arises from disagreement (real, imagined, or anticipated), it is associated with interaction aimed at exerting impact on the receiver (Fuentes Rodríguez & Alcaide Lara, 2002, p. 17). As a consequence, the speaker needs to take into consideration the receiver’s knowledge, beliefs, and experiences in order for the argumentative strategy to succeed. In other words, an argumentative text “foresees reactions from the audience even if the latter is virtual or apparently absent”, which in turn “explains how an utterance that does not voice the argument of the Other is still informed by it and provides an answer [to it]” (Amossy, 2005, p. 89). With regard to news reports, their main dialogue partners/ potential opponents are not only the readers, but also business and political elites as well as other media (for a discussion on the relationship of the media with these three groups of actors, see Chapter 3). The speaker, on the other hand, is the person responsible for the content of the text, although she does not necessarily coincide with the author (Fuentes Rodríguez & Alcaide Lara, 2002, p. 18). Specifically, with respect to news production, the journalist who writes a piece of news is not the only person who decides on the stance the text embraces and the ways of its expression, as discussed in Chapter 3. There exist, however, texts with no argumentation (Fuentes Rodríguez & Alcaide Lara, 2007, p. 15). 27 97 With respect to the text, there are several principal elements forming its argumentative structure (see Figure 2). The claim (also referred to as the conclusion, stance, standpoint, or thesis) is a negative or positive position regarding the given proposition. It may be stated explicitly or left implicit: as a result, its identification is one of the central challenges for an argumentation analysis (van Eemeren et al., 2014, p. 13-14). Nevertheless, for the text to be able to reach its persuasive aim, the conclusion should always be accessible to the receiver through an analysis of the argumentative structure in which it appears — the arguments and topoi (as discussed below), as well as the context of the sentence, paragraph or text, or the broader context of the text production, distribution, and reception (Fuentes Rodríguez & Alcaide Lara, 2002, p. 40). Figure 2. Argumentative Structure CLAIM TOPOS ARGUMENT SUBARGUMENT EXAMPLE 98 Note: this simplified schema of the argumentative structure is meant to show the relationships between its various elements. Obviously, in naturally-occurring discourse, any claim might be supported by a number of arguments, which might, in turn, be based on more than one topoi, be illustrated with numerous examples, etc. The claim is supported by arguments, defined as reasons which justify a conclusion, which are characterised by their orientation and the degree of argumentative force (Fuentes Rodríguez & Alcaide Lara, 2002, pp. 26, 55-63). The argumentative force of an argument can be modified by the use of connectors, repetition, evaluative lexis, and modals, which can either strengthen or lower the force of the argument and present it as sufficient or insufficient (Fuentes Rodríguez & Alcaide Lara, 2002, pp. 26, 55-63). In other words, different arguments may be placed at different positions on the scale of the strength of argumentation. For example, the use of even endows an argument with the highest level of argumentative force since it also construes it as unexpected. On the other end of the scale, there are arguments which are presented as simply sufficient by the use of items such as at least, after all, in the end; or as insufficient, e.g. only (Fuentes Rodríguez & Alcaide Lara, 2002, pp. 26, 55-63). Arguments may, in turn, be developed by means of subarguments, thus forming a complex argumentative structure (van Eemeren & Grootendorst, 1984, pp. 91-92; Lo Cascio, 1991, pp. 103-104; Fuentes Rodríguez & Alcaide Lara, 2002, pp. 80-95). On the other hand, in the absence of subarguments, the argumentative structure takes one of the two forms: simple (where only one argument supports a conclusion) or multiple (which employs more than one argument in support of the thesis) (van Eemeren & Grootendorst, 1984, p. 90; Lo Cascio, 1991, p. 102). 99 Newspaper articles tend to be based on complex argumentative structures, as this analysis will also show. Not all arguments are valid. Arguments following invalid or otherwise faulty reasoning are called fallacies. As “contaminators of argumentative discourse”, fallacies fail to lead to correct conclusions and are often used for manipulative and deceptive purposes (van Eemeren et al., 2014, p. 24; Rubio-Carbonero & ZapataBarrero, 2017, p. 208). According to Van Der Valk (2003, p. 320) the most frequent fallacies in discourses on ethnic issues include: slippery slope (asserting than a relatively small step is bound to start a chain of events with major, usually negative, consequences), the straw man (addressing a misrepresentation of the opposing position rather than the real subject of the argument), argumentum ad hominem (an attack on the character, motive, or any other attribute of the opponent), argumentum ad misericordiam (appeal to mercy), and overgeneralisation. The argument and the conclusion are often connected by means of topoi. Topoi are general principles on which reasoning relies, almost always presented as being an object of consensus in a more or less large community (Anscombre, 1995b, p. 35). In other words, topoi serve as bridges allowing a premise to reach the conclusion by means of a generally accepted assumption in the particular sociological and historical context. Fuentes Rodríguez and Alcaide Lara (2002, p. 41) discuss an advertising phrase “young but more than prepared” as an example of the use of the topos implying that youth equals carelessness. As the example shows, topoi are often generalisations and can, therefore, be used for persuasive and manipulative effects. Since topoi form part of common knowledge in a given community, they are usually left implicit, as in the example discussed above. Nevertheless, they may also be stated explicitly in order to avoid ambiguity, in highly argumentative texts, or when the 100 speaker suspects that the receiver may not be aware of a particular topos (in which case they are referred to as argumentative basis by Fuentes Rodríguez and Alcaide Lara [2002, p. 49]). The entire argumentative structure, comprised of a claim, arguments, and topoi, is based on logical reasoning, where the relations between ideas are implicit or stated explicitly using grammatical tools called connectors (Guyot-Clément, 2012, p. 27-31). In Figure 2 above the relations between the different elements of an argumentative structure are represented by arrows. This study examines argumentation employed in news reports on the “refugee crisis” by focusing on a sample of texts selected from the larger corpus (whose composition is discussed in Chapter 6). Two hundred and eighty eight texts were chosen, i.e. 96 texts per language corpus (six texts x four newspapers x four years), published following “critical discourse moments” (Lams, 2018) of the “refugee crisis”, i.e. some of the most significant events with the potential to influence on how the movement of the displaced people was perceived. A wide range of different events were selected in order not to skew the results towards, for instance, the political arena. At the same time, however, it is important to note that these events served only as sampling points, and the selected texts do not necessarily concern the events following which they were published (as this would switch the focus of the analysis). Table 2.3 displays the list of the “critical discourse moments” selected, along with their brief descriptions. 101 Table 2.3. Critical Discourse Moments of the “Refugee Crisis” Date Event Description 2015 September 2 death of Alan Kurdi The photos taken by Nilüfer Demir of the body of the three-year-old Alan Kurdi, victim of human smuggling, are often credited with the change of public attitudes towards refugees to what Chouliaraki and colleagues (2017, p. 15) referred to as “ecstatic humanitarianism”. Two days later Angela Merkel declared an exception to the EU asylum rules and announced the “Refugees Welcome” programme which allowed a million displaced people to settle in Germany. April 18 sinking of a boat carrying 700 migrants The sinking of a smuggling boat on its way from Libya to Italy, which resulted in the death of approximately 700 people, was reported to be “the largest single loss of life in the Mediterranean in decades” (Miglierini, 2016, par. 1). November 13 terrorist attacks in France It is widely agreed that, following the series of coordinated terrorist attacks in Paris and Saint-Denis, for which the Islamic State of Iraq and the Levant claimed responsibility, a moral panic frame began to dominate the media reporting on the “refugee crisis” (e.g. Traub, 2016; Chouliaraki et al., 2017; Lams, 2018). 2016 January 1 sexual attacks in Germany Over 1,000 attacks, including 492 cases of sexual assault, were reported to have been perpetrated by men with an “Arab or North-African appearance” on New Year’s Eve 2015 in Cologne, Germany (Yurdakul & Korteweg, 2021, p. 47). These events became mediatised to a significant extent and were often discussed with respect to the “Refugees Welcome” programme. March 18 EU-Turkey deal The statement of cooperation between the European Union and the Turkish government had as its main objective to stop people travelling from Turkey to the Greek Islands by irregular means and was widely reported by the press (e.g. Fotopoulos & Kaimaklioti, 2016, p. 273), announced the preceding year. June 23 EU membership referendum in the UK It has frequently been noted that issues relating to migrations were a particularly salient theme in the discourses surrounding the Brexit referendum (Goodfellow, 2018; Virdee & McGeever, 2018; Koram & Nisancioglu, 2017; see also Section 4.2 of this thesis). 2017 September 24 elections in Germany As a result of the election, the populist radical-right party, Alternative for Germany, became the third biggest party in the Bundestag. One of the main tenets of its political campaign had been immigration, in particular opposition to Angela Merkel’s Willkommenspolitik towards refugees (Mudde, 2017). 102 Table 2.3. (Cont.) Critical Discourse Moments of the “Refugee Crisis” Each selected article was then analysed with respect to its argumentative dimension: claims (expressed explicitly or implicitly), arguments, and topoi were identified and examined. The results of this part of the investigation are presented in Chapter 9. 7.3. Methodological Triangulation The methodological triangulation employed in this study — combining the quantitative approach with the more qualitative perspective usually associated with Critical Discourse Analysis — is a way of providing a more robust analysis which accounts for both the most frequently articulated representations of refugees and the more obscure, nuanced, and context-sensitive patterns (e.g. Dörnyei, 2007; Marchi & Taylor, 2009; Baker & Levon, 2015; 2016; Marchi & Taylor, 2018; Taylor, 2018). It also allows the researcher to validate data through cross-checking the results of the Date Event Description December 18 elections in Austria The Austrian People’s party formed government in coalition with the far-right, anti-immigration Freedom party. 2018 January 9 Libya migrant shipwreck Up to 100 migrants went missing when their boat sank off the Libyan coast. April 8 elections in Hungary The anti-immigration Fidesz party won a landslide victory in Hungary’s general elections for the third time. The party had been responsible for the construction of a wall at the border with Serbia and Croatia in order to impede the movement of displaced people across the Hungarian territory. June 12 Spain gives safe harbour to the Aquarius Aquarius, a rescue ship carrying 629 people, was denied entry by Italy and Malta and spent more than a week at sea before being given permission to dock in Valencia, Spain. 103 particular approaches (Egbert & Baker, 2016, p. 201; Baker & Levon, 2015; 2016; see also Chapter 12 of this thesis). Finally, it has traditionally been considered to be an effective means of counterbalancing their potential weaknesses (e.g. Baker et al., 2008; Virtanen, 2009). On the one hand, with respect to discourse analysis, a range of criticisms has been raised. Most notably, Stubbs (1997) and Breeze (2011) discussed the principal charges against the currently prevalent approach to discourse analysis, namely Critical Discourse Analysis (which shares many features with Pragmatic Linguistics): the fact that it “can be seen as heavily conditioned by political choice, rather than scientific criteria, which might be thought to take on a secondary role” (Breeze, 2011, p. 501); that its theoretical bases are not always clearly defined and draw on eclectic — and not always compatible — sets of concepts from different traditions; that its methods, therefore, lack coherence, which does not always lead to rigorous analysis and does not provide sufficient support for the interpretations that are subsequently drawn from it; that it does not take into account audience perceptions; that it pays insufficient amount of attention to the immediate context; and that it is overwhelmingly negative. Furthermore, discourse analysis methodologies have been accused of cherry-picking the texts for analysis which “prove a point”, are in some aspect unique and/ or familiar to the researcher (Baker, 2012; Baker et al., 2013; Brezina 2018a, p. 266) and of using small data samples which may lack some of the features of discourse under consideration or “contain them in too small frequencies for results to be reliable, particularly when issues of statistical significance are not addressed" (Baker et al., 2008, p. 275). It also tends to be selective regarding the aspects of texts which are subject to the analysis (Stubbs, 1997, p. 3). Consequently, as Lischinsky (2018, p. 60) concludes: “when texts and features for analysis are 104 selected on the basis of the researcher’s intuitive judgement (Marchi & Taylor 2009: 3), there is no guarantee that they truly represent the distinctive patterns that characterise a discourse”. Finally, Baker and colleagues (2013) advise that discourse analysis is particularly prone to the analyst’s bias. Nevertheless, it should be mentioned that to counter some of the criticisms regarding traditional (critical) discourse analysis, a specific research perspective was adopted in this analysis — namely, Pragmatic Linguistics — which provides both the theoretical underpinnings as well as a consistent and comprehensive methodological framework, as discussed above. Similarly, an objective sampling method has been applied to the process of text selection. On the other hand, relying on computer software and statistical analysis allows the researcher to obtain more credible results based on larger, more representative and balanced samples of authentic language which reduces idiosyncratic variation (e.g. Biber & Flagstaff, 2009; Biel, 2009; Biel, 2017). As Stubbs (1994, p. 204) noted some patterns of language use are not directly observable, because they are realized across thousands or millions of words of running text, and because they are not categorical but probabilistic. Of course, some patterns may be discernible, in a rough way, via intuition. But in order to describe such distributions systematically, significant amounts of text must be stored on computer and searched, and quantitative methods must be used to describe the patterns. Stubbs thus also points to the fact that corpora are analysed in a systematic and exhaustive way, “meaning that the corpus does not simply serve as a database of examples from which some can be chosen ad libitum and others neglected but that the whole (sample of the) corpus is taken into consideration” (Gries, 2006, p. 4). As a result, corpus linguistics can also shed light on marginal discourses which could be ignored if only qualitative methodologies are applied, for example because they run 105 counter to researcher intuition, but which are nevertheless frequent or salient in a given discourse (Baker, 2010; Baker, 2016a). Another important advantage of corpus approaches is the high-level of research transparency. This allows other scholars to check the validity of findings, as well as to offer different interpretations of the same data (Biel, 2009). However, this approach also has some limitations. For one, even though it is by no means limited to a quantitative approach, corpus linguistics does rely heavily on frequencies and statistical information. Frequency counts in themselves can sometimes be misleading, and certain types of data — which a discourse analyst might find to be useful in contextualising findings, for example — may be discarded (Baker, 2010; Virtanen, 2009). What is more, corpus linguistic methods tend to investigate discourse patterns at the level of the entire corpus rather than in individual texts (Egbert & Schnur, 2018, pp. 159-160). This might lead to, for instance, prioritising repeated patterns of language use over outstanding texts with more socio-cultural relevance. Furthermore, they have often been reported to be more suited to indicate differences rather than similarities between corpora as well as presences rather than absences. Finally, CL is often criticised for neglecting the wider context of discourse production and reception due to the large size of data, which is said to lead to a lower explanatory and interpretative potential (Nartey & Mwinlaaru, 2019). To sum up, the methodological triangulation was applied to this study in the hope that, in the words of Marchi & Taylor (2018, p. 4): mixing the two should bring together social relevance and statistical relevance, thus overcoming the peril of quantitative studies ‘counting only what is easy to count’ (Stubbs & Gerbig 1993: 78) and of qualitative studies ‘find[ing] what they expect to find’ (Stubbs 1997: 2). 112 least salient group of verbs collocating with the query term comprises references to helping refugees. With respect to the Spanish subcorpus, Table 3.3 displays the nouns and adjectives which collocate most frequently with refugiado. It is important to note here that despite its lower frequency, its collocational range is bigger than the collocational range of uchodźca. This may indicate that the themes present in the Spanish news reports on refugees are more numerous and varied, compared to the seemingly more formulaic way of covering the topic in the Polish press. As can be observed in Table 3.3, newspapers in Spain construed refugees predominantly in regard to their movement. The items which comprise this category of collocates can be divided further into, first, words which refer to the action of arriving somewhere (in Europe, above all, as the analysis of the concordances has indicated) and being relocated (within Europe or to third countries). These constitute 40% of the words in this category. Secondly, words referring to refugees’ being expelled from Europe account for 20% of this category. Thirdly, the remaining 40% are references to movement in general, the majority of which belong to the semantic field of water and construe migrations in terms of water flows: flujo [‘flow’], oleada, ola [‘wave’], avalancha [‘avalanche’], afluencia [‘influx’], expulsión [‘expulsion’], and marea [‘tide’] (see Section 10.2 for a discussion of the use of water metaphors in the corpus under the investigation). 113 Table 3.3. Nouns and Adjectives Collocating with Refugiado [‘Refugee’] in the Spanish Subcorpus The second top category of collocates of refugiado comprises items denoting geographical spaces. These include references to refugees’ (temporary) accommodation in Europe, e.g. Calais or campo [‘camp’], and, above all, to their geographical identities, e.g. iraquí [‘Iraqi’] and sirio [‘Syrian’], the latter being the Category Collocates Movement llegada [‘arrival’], flujo [‘flow’], oleada, ola [‘wave’], entrada [‘entry’], avalancha [‘avalanche’], reubicación [‘relocation’], afluencia [‘influx’], éxodo [‘exodus’], expulsión [‘expulsion’], devolución, retorno [‘return’], distribución [‘distribution’], ruta [‘route’], paso [‘passage’], reasentamiento [‘resettlement’], contingente [‘contingent’], movimiento [‘movement’], marea [‘tide’], tránsito [‘transit’], redistribución [‘redistribution’], destino [‘destination’], deportación [‘deportation’] Geographical space sirio, (de) Siria [‘Syrian’], procedente [‘from’], palestino [‘Palestinian’], afgano [Afghan’], (de) Calais [‘(from) Calais’], iraquí [Iraqi’], (de) Grecia [‘(from) Greece’], (de) Eslovenia [‘(from) Slovenia’], (de) Europa [‘(from) Europe’], (de) Somalia) [‘(from) Somalia’], (de) Turquía [‘(from) Turkey’], saharaui [‘Saharan’], español [‘Spanish’], (de) Italia [‘(from) Italy’], de Idomeni [‘(from) Idomeni’], (de) Oriente (Medio) [‘(from Middle) East’], eritreo [‘Eritrean’], somalí [‘Somalian’], (de) Moria [‘(from) Moria’], (de) país [‘(from) country’], campo [‘camp’], centro [‘centre’], albergue [‘hostel’], campamento [‘encampment’] Numbers mil [‘thousand’], solo [‘single’], número, cifra [‘number’], mayoría [‘most’], ciento, centenar [‘hundred’], millón [‘million’], parte [‘part’], cantidad [‘quantity’], decena [‘dozen’], medio [‘half’] Conflict crisis, problema [‘problem’], situación [‘situation’], cuestión [‘question’], cargo [‘charge’], materia [‘matter’], tema, asunto [‘subject’] Plight vulnerable, drama, condición [‘condition’], tragedia [‘tragedy’], vida [‘life’], tráfico [‘trafficking’] Institutional & administrative cuota [‘quota’], política [‘politics’], derecho [‘right’], gestión [‘management’], registro [‘registration’], estatuto, estatus [‘status’] Aid acogida [‘accommodation’], integración [‘integration’], recepción [reception’], protección [‘protection’] Incentive político [‘political’], de guerra [‘war’], potencial [‘potential’] Religion musulmán [‘Muslim’], cristiano [‘Christian’] Family familia [‘family’], reagrupación [‘reunification’] Age menor [‘underage’], joven [‘young’] Collectivity grupo [‘group’] 114 strongest collocate of refugiado when all grammatical categories are combined (AF=563; RF=227; logDice score = 12.66). Describing refugees with respect to their numbers is the third most significant tendency in the Spanish corpus. The most frequent modifier within this category — the word mil [‘thousand’] — is also the fifth strongest collocate of refugiado when all grammatical relations are considered (AF=208; RF=83; logDice score = 9.15). The word sketch of refugiado has also shown a significant tendency to construe refugees as a problem/ source of conflict by means of collocates such as crisis [‘crisis’] (the second top collocate of the query term when all grammatical categories are combined, with AF=570; RF=229; logDice score = 10.44), problema [‘problem’], etc. Furthermore, refugees are frequently discussed in the Spanish press with respect to various institutional and administrative measures, as attested by the salience of words such as cuota [‘quota’] and gestión [‘management’]. Nevertheless, the query term also collocates frequently with words referring to the plight of refugees and to helping them, even though these trends are much less pronounced. Other minor trends in the representation of refugees in the Spanish press include references to the reasons behind their migrations, such as war and political persecution; their religious beliefs (Islam/ Christianity); family relations; age (with refugees being predominantly represented as young/ underage); and group belonging. Similarly, verbs which most frequently collocate with refugiado include, above all, references to movement (see Table 3.4). Refugees are predominantly represented as migrating or being returned (devolver), transported (trasladar), or sent (enviar) somewhere by another actor/ group of actors. In other words, they are often construed as passive and agentless. Less frequent verbs include references to the 115 challenges refugees face during their migrations, including financial constraints (the verb pagar [‘pay’] in collocation with refugiado refers predominantly to the costs of travelling from the Middle East/ Africa to Europe by illegal means, as the analysis of the concordances has shown). There are also two verbs which make reference to the institutional and legal aspects of migrations, including registration and detention of refugees; and two verbs referring to the action of helping them. Table 3.4. Verbs Collocating with Refugiado [‘Refugee’] in the Spanish Subcorpus With respect to the subcorpus of British news, Table 3.5 displays the most statistically significant collocates of the word refugee. First of all, it reveals that refugees are most often described in terms of the geographical spaces they occupy. On the one hand, they are represented with respect to their nationalities as Syrian, Afghan, Rohingya, Sudanese, Pakistani, Palestinian, Iranian (or generally as African). The adjective Syrian is the strongest collocate of the query term when all grammatical categories are combined, with AF=519; RF=197; and logDice score = 10.55. On the other hand, references to places where refugees are accommodated upon their arrival or on their way to Europe are also frequent. These include the word camp — the second strongest collocate of refugee in the focus corpus when all grammatical categories are combined (AF=590; RF=224; logDice score = 10.52). Category Uchodźca as Subject Uchodźca as Object Movement llegar [‘arrive’], huir [‘flee’], salir, regresar [‘leave’], ir [‘go’] devolver [‘return’], trasladar [‘transport’], enviar [‘send’] Plight pagar [‘pay’] varar [‘strand’] Institutional & legal registrar [‘register’], atrapar [‘catch’] Aid recibir [‘receive’] rescatar [‘rescue’] Accommodation acoger [‘accommodate’] 116 Table 3.5. Nouns and Adjectives Collocating with Refugee in the UK Subcorpus The second top category of collocates of refugee includes references to age, with the most salient being the word child (AF=267; RF=101; logDice score = 9.23). This word, used as a modifier, is, in fact, the fifth strongest collocate of the query term when all grammatical categories are combined. Furthermore, refugee has a strong semantic preference for quantification: for example, the word thousand is its sixth strongest collocate when all grammatical categories are combined (AF=168; RF=64; logDice score = 8.98). The analysis of the refugee word sketch has also revealed a significant trend of representing refugees in terms of the binary opposition genuine/real versus bogus/ so-called. Moreover, in the UK press there is a tendency to construe refugees as a source of problem/ conflict and as an issue tackled at the institutional level. Refugees are also frequently represented with respect to their movement, group belonging, religious identity (Muslim), and gender (male or female). Category Collocates Geographical space Syrian, camp, centre, Afghan, Rohingya, Sudanese, Iraqi, Pakistani, Palestinian, African, Iranian, shelter Age child, kid, young, teenage, old, year Numbers thousand, number, more, many, most, few, single Legality genuine, bogus, real, so-called Conflict crisis, situation, problem, issue Institutional & legal deal, quota, scheme, programme, convention, policy, status, agency Collectivity population, group, team Aid help, charity Plight unaccompanied, vulnerable, desperate, lone, war Movement boat, arrival, influx Religion Muslim Gender male, woman 117 The analysis has also detected the presence of a humanitarian aid trend in the portrayal of refugees, as the word help is one of the strongest collocates of the search term both as a modifier and as a verb (see below). Furthermore, there is a number of adjectives which describe the plight of refugees in terms of their vulnerability, such as unaccompanied, lone (both employed most frequently together with child, as in unaccompanied/lone child refugee), vulnerable, and desperate. Verbs which most frequently collocate with refugee (Table 3.6) have a much more limited thematic scope. Two major trends can be observed. On the one hand, verbs which take refugee as object overwhelmingly refer to the action of accommodating refugees performed by European states. On the other hand, when refugee appears as a subject, the verbs tend to focus on movement. Furthermore, there are references to the hardships refugees have to overcome and to helping them. A less significant trend construes refugees as being detained for illegal actions. Table 3.6. Verbs Collocating with Refugee in the UK Subcorpus Category Refugee as Subject Refugee as Object Accommodation live accept, take, welcome, resettle, allow, host, house, integrate, bring, accommodate, admit, stick, receive, let, relocate, keep Movement arrive, flee, come, try (often to cross, enter, reach, etc.), reach, cross, enter, head, leave, use (usually: the route, dinghies, lifejackets), settle, land, attempt (the journey, etc.), go, travel, carry send, bring, return Plight face, seek, die strand, trap, force Aid help, rescue, support, protect Threat detain, stop, include 118 A note should be made with respect to the verb include, which has been placed in the threat category because a concordance analysis has shown that it is usually employed to construe refugees as potentially deceitful with respect to their identities and, consequently, threatening. For example, in excerpt (1.5) it is used to differentiate between refugees and “economic migrants”: (1.5) They include refugees fleeing the turmoil in Syria and Iraq as well as economic migrants from Africa, the Middle East and Asia. (The Sun, 26 September 2017) Excerpt (1.6), in turn, implies, through the use of include, that all refugees are Muslims and establishes a link between the displaced people and terrorist violence (but notice also the use of mitigating strategies, such as “it is believed” and “no one seems sure”, which are aimed at protecting the speaker from accepting the responsibility for this association): (1.6) The perpetrators, it is believed, included several refugees, or if not refugees at least Muslims – at the moment, no one seems sure. (The Guardian, 19 April 2017) 8.1.2. Word Sketches: Imigrant, Inmigrante, Migrant This Section examines the collocational behaviour of the second set of the query terms — imigrant, inmigrante, and migrant. With respect to the Polish corpus, the analysis of the word sketch of imigrant has revealed that its strongest collocate is the adjective nielegalny [‘illegal’] (AF=179, RF=140; logDice score = 10.63). This already suggests a significant discrepancy between how refugees and immigrants are construed in the Polish press as uchodźca collocates with nielegalny only seven times (RF=5; logDice score = 7.21). 119 In a similar way, the second top collocate of imigrant — ekonomiczny [‘economic’] (AF=89; RF=69; logDice score = 9.67) — as well as the synonymous zarobkowy (AF=12; RF=9; logDice score = 8.95) construe some groups of the displaced people as migrating solely for economic reasons. At the same time, refugees are mostly described using the modifiers polityczny [‘political’] (AF=20; RF=15; logDice score = 8.46) and wojenny [‘war’] (AF=13; RF=10; logDice score = 8.4) with respect to the reasons behind their migrations. What is more, the analysis of the concordances of the collocation imigrant + ekonomiczny [‘economic’ + ‘immigrant’] has revealed that the most frequent construction is odróżniać/ oddzielać/ odseparować [‘differentiate’/ ‘separate’] imigrantów ekonomicznych od uchodźców [‘economic immigrants from refugees’]. The third top collocate of imigrant is the word Afryka [‘Africa’] (as in imigrant z Afryki [‘immigrant from Africa’]) (AF=54; RF=42; logDice score = 11.93). This reveals another point of divergence between how refugees and immigrants are construed as the most salient collocates of uchodźca, referring to geographical identities, are Syria (AF=232; RF=181; logDice score = 12.66) and syryjski (AF=48; RF=37; logDice score = 10.15). Furthermore, collocates of uchodźca relating to places are much more numerous and varied (see Table 3.1 above), compared to those of imigrant (Table 4.1). 120 Table 4.1. Nouns and Adjectives Collocating with Imigrant [‘Immigrant’] in the Polish Subcorpus The fourth most salient collocate of imigrant is muzułmański [‘Muslim’] (refugees were construed as Muslim, too). Furthermore, like refugees, immigrants have a semantic preference for quantification as well as movementand conflictrelated items in the Polish subcorpus. However, the overall number of frames used in the construal of immigrants, compared to refugees, is much more limited. Specifically, contrary to uchodźca, among the most salient collocates of imigrant there are no references to humanitarian aid, public attitudes, age, politics, or money. Similarly, verbs collocating with imigrant (Table 4.2) are less varied than those which collocate with uchodźca. There are no references to the legality of immigrants, which shows that, on the one hand, they are not — or scarcely — presented as applying for or receiving permission to stay in Europe. On the other hand, however, this also indicates that the trend of portraying displaced people as terrorists/ criminals in disguise applies to refugees only. Furthermore, in contrast to uchodźca, the word imigrant does not collocate with verbs relating to help. Category Collocates Geographical space Afryka [‘Africa’], Syria, ośrodek [‘centre’], wschód [‘East’], Maroko [‘Morocco’], arabski [‘Arab’], mieszkanie [‘apartment’], państwo, kraj [‘country’], Polska [‘Poland’], granica [‘border’], Europa [‘Europe’] Numbers tys. [‘thousand’], wszystek [‘all’] Incentive ekonomiczny, zarobkowy [‘economic’] Religion muzułmański [‘Muslim’], islamski [‘Islamist’] Legality nielegalny [‘illegal’] Problem problem (z) [‘problem (with)’] Movement łódź z [‘boat (with)’] 121 Table 4.2. Verbs Collocating with Imigrant [‘Immigrant’] in the Polish Subcorpus On the other hand, the analysis of verbs as collocates has also revealed similarities between the representations of groups labelled as refugees and immigrants in the Polish press. First of all, threat-related words are the most frequent collocates of both query terms, which points to a significant tendency of portraying both groups as dangerous. Second of all, the two terms collocate strongly with verbs relating to (not) being allowed to settle in Europe. Third of all, both collocate with verbs of movement. In turn, imigrant, unlike uchodźca, also collocates with the verb deportować [‘deport’]. The analysis of the word sketch of inmigrante (Table 4.3) has shown that the people labeled as immigrants are construed in the Spanish press using similar discursive frames as those who are referred to as refugees. First, movement-related words collocate with inmigrante most frequently. However, they are less numerous than movement-related collocates of refugiado and their thematic distribution follows a different pattern. Specifically, they can be divided into three categories: words denoting the action of arriving at a destination, such as llegada [‘arrival’], which is the third top collocate of the query term when all grammatical relations are combined (AF=180; RF=72; logDice score = 9.62); the twice more frequent references to expulsion, e.g. deportación [‘deportation’] (40% of all words in this Category Imigrant as Subject Imigrant as Object Threat zgwałcić [‘rape’], zabić [‘kill’], molestować [‘molest’], zaatakować [‘attack’] zatrzymać, powstrzymywać [‘stop’] Accommodation przyjmować, przyjąć [‘acommodate’] Movement płynąć [‘swim’], przybywać [‘arrive’] odsyłać [‘send away’], deportować [‘deport’] 128 (1.9) It wants the EU to forge repatriation deals with African nations and share the cost of returning home would-be economic migrants, who currently make up around 60 percent of those arriving by boat. (The Guardian, June 15, 2015) This suggests that becoming a migrant is perceived by the UK press as contingent on being accepted in a host country. The analysis has also revealed that migrant collocates less frequently with adjectives denoting the difficulties faced by the displaced people and words referring to the institutional aspects of their movement than refugee; and not at all with words related to group-belonging, humanitarian aid, or movement. Furthermore, migrant collocates more strongly with Muslim (AF=29; RF= 11; logDice score = 8.5) than refugee (AF=12; RF= 4; logDice score = 7.18). Finally, with respect to gender, it collocates solely with the word male, unlike refugee which collocates both with the adjective male and the noun woman (see Section 8.3.5 for a more detailed discussion). The thematic distribution of verbs collocating with migrant is equally limited as in the case of refugee. They can be subsumed into five main categories. The first category is related to movement, and it is very numerous. In fact, the majority of all verbs denoting actions in which migrants participate are verbs of movement. This is in line with the findings of the analysis of refugee word sketch as well as previous contributions (Berry et al., 2016). The second most populous category are verbs which denote the action of detaining migrants as suspects or perpetrators of illegal acts, and deporting them. Third, migrants are frequently discussed as receivers of accommodation in their host countries. Finally, the two least frequent categories of verbs portray migrants as victims requiring help. 129 Table 4.6. Verbs Collocating with Migrant in the UK Subcorpus 8.2. Wordlists: Context and Topic Indicators This part of the analysis focuses on the most frequent nouns, verbs, and adjectives in the corpus. Their examination is expected to reveal the main discursive themes in the Polish, Spanish, and British press coverage of the “refugee crisis”. To this end, as explained in Section 7.1, lists of lemmas with frequencies over 30 per million were obtained for each of the subcorpora, thus accounting for approximately four fifths of all of the tokens in the corpus (see Table 2.2 for statistical details). Then, in order to identify items that do, in fact, point to themes in discourse, the wordlists were divided, based on their functions in discourse, into words with general meaning, topic indicators, and context indicators, as shown in Figure 3 (lists of concordances were consulted whenever an item’s function was open to question; see Section 7.1 for more details on the categorisation procedure). Verb Categories Migrant as Subject Migrant as Object Movement arrive, try, cross, enter, make (make their way, make the crossing, etc.), reach, come, attempt, flee, land, leave, head, travel, walk, run, go, move, pay (smugglers to be transported), storm, continue (to flow, to arrive) carry, arrive, smuggle, strand, move, charge, ferry Detention stop, find, send, force, deport, return, prevent, catch, detain, hold, arrest, intercept Accommodation live take, allow, house, welcome, give, bring, keep, accept Aid rescue, help, save, pluck Plight die, risk, drown, face 130 Figure 3. Distribution of High-Frequency Words in the Polish (Left), Spanish (Centre), and the UK (Right) Subcorpora As expected (e.g. Scott & Tribble, 2006, p. 23), this has shown that general words prevail significantly and account for 70-80% of all the high-frequency words (depending on the subcorpus). They were not subject to further analysis as they do not point to any particular theme in the corpus. The remaining words, i.e. context and topic indicators, on the other hand, are considered to be indicative of specific discursive themes. In order to facilitate their examination, they were divided into subcategories: context indicators were grouped into ten separate subcategories, whereas topic indicators have pointed to the presence of six main themes in the corpus. The remaining part of this Section thus focuses on the functions of context (Section 8.2.1) and topic indicators (Section 8.2.2) in the three language subcorpora. 8.2.1. Context Indicators With respect to context indicators in the Polish subcorpus (Table 5.1), the most populous category (in terms of types and tokens) includes references to geographical locations (countries, regions, cities, etc.). In order to establish which territories are foregrounded in the press coverage of the refugee crisis, the top twenty types (which 18% 11% 71% general words topic indicators context indicators 17% 11% 72% general words topic indicators context indicators 11% 9% 80% general words topic indicators context indicators 131 account for close to two thirds of all tokens in this subcategory) have been analysed. This has revealed that half of them refer to places in Europe (Polska [‘Poland’] and Europa [‘Europe’] being among the top three context indicators in this category), seven are general words, such as kraj [‘country’] and morze [‘sea’], and only three denote places outside of Europe: Syria, Turcja [‘Turkey’], and Afryka [‘Africa’]. The analysis of references to geographical and linguistic identities has revealed a similar trend: among the top twenty types, which account for almost 80% of all tokens within this category (12,068 tokens), the overwhelming majority are references to European identities (10,040 tokens). This indicates that aspects of the “refugee crisis” which refer to Europe are foregrounded in the corpus, with significantly less attention paid to the circumstances in the regions where migrations originate. This is in line with previous studies, and it indicates, as Van Der Valk (2003, p. 322) notes, that “discourses that are apparently about ‘them’ sometimes actually have ‘us’ as central focus”. Table 5.1. Most Frequent Context Indicators in the Polish Subcorpus Subcategory Types N=540 Tokens AF=164,772 Tokens RF=128,960 Definition Example types Region/ country/ city/ area/ directions 132 47,556 37,240 Continents, regions, countries, wider areas within a country, cities Polska, kraj, Europa, Bruksela Quantification 98 26,602 20,832 Reference to numbers or quantities tys., jeden, liczba, kilka Governance/ politics 80 19,370 15,168 Heads of state, politicians premier, Merkel, PiS Refugees/ migrants 20 18,327 14,352 References to refugees and migrants uchodźca, imigrant 132 Table 5.1. (Cont.) Most Frequent Context Indicators in the Polish Subcorpus Similarly, references to — overwhelmingly European — politicians and heads of state surpass references to refugees and migrants, even though the latter include the query terms (this is also in line with Van Der Valk’s [2003] findings based on her analysis of right-wing parliamentary discourse on immigration in France). Furthermore, types referring to the displaced people are four times less varied than references to politicians. In fact, the displaced people are predominantly referred to using general terms and there are no references to individuals among the list of the top context indicators. Another group of actors who are discussed in the analysed corpus are non-governmental organisations. They are, however, mentioned 24 times less frequently than politicians. This study has also demonstrated that among references to places within cities, the most frequent are obóz [‘camp’] and ośrodek [‘centre’], which indicates that places where refugees are accommodated in Europe (or, possibly, in transit Subcategory Types N=540 Tokens AF=164,772 Tokens RF=128,960 Definition Example types Geographical/ linguistic identity 59 15,278 11,964 Identification of people in terms of their ethnicity, nationality, place of origin or residence, or their language polski, Polak, niemiecki, Syryjczyk Time 52 14,144 11,076 Reference to time points or periods rok, miesiąc, tydzień Gender/ family relations 30 8,323 6,518 Identification of people in terms of their gender, age, and family relations dziecko, rodzina, kobieta Places/ buildings 44 8,305 6,503 Areas within a city, buildings obóz, ośrodek, dom, szkoła Groups/ communities 20 6,063 4,748 Identification of people in terms of their group belonging grupa, organizacja Nongovernment organisations 5 803 628 Non-governmental groups and their representatives fundacja, pozarządowy 133 countries) are significantly foregrounded in the corpus. Other groups of contextindicative words relate to time and identify people in terms of their gender or family relations and group belonging. Finally, the analysis has revealed a high occurrence of references to quantities in the Polish corpus. It is, however, significantly lower than in the Spanish corpus where quantification is the most significant trend (see Table 5.2). Table 5.2. Most Frequent Context Indicators in the Spanish Subcorpus Subcategory Types N=550 Tokens AF=332,089 Tokens RF=133,996 Definition Example types Quantification 142 108,052 43.598 Reference to numbers or quantities uno, mucho, primero Region/ country/ city/ area/ directions 133 76,219 30.754 Continents, regions, countries, wider areas within a country, cities Europa, España, Alemania Time 65 37.236 15.024 Reference to time points or periods año, mes, semana Geographical/ linguistic identity 57 27,423 11.065 Identification of people in terms of their ethnicity, nationality, place of origin or residence, or their language europeo, sirio, alemán, español Refugees/ migrants 15 25.849 10.429 References to refugees and migrants refugiado, inmigrante Governance/ politics 53 20.672 8.341 Heads of state, politicians ministro, Merkel, jefe Places/ buildings 40 13.985 5.642 Areas within a city, buildings centro, campo, lugar, zona Gender/ family relations 21 11.317 4.566 Identification of people in terms of their gender, age, and family relations niño, mujer, mayor, familia Groups/ communities 13 7.586 3.060 Identification of people in terms of their group belonging grupo, miembro, socio Nongovernment organisations 11 3.750 1.513 Non-governmental groups and their representatives UNIR, ONG, oenegé 134 The second top category of context indicators in the Spanish subcorpus includes words referring to geographical locations (countries, regions, cities, etc.). Similarly as in the Polish corpus, the analysis of the top twenty types, which account for over 60% of all tokens within this category (76,219 tokens), has revealed that nonEuropean countries are backgrounded in the corpus. Specifically, nine types refer to places in Europe (with Europa [‘Europe’] and España [‘Spain’] in the top three), eight do not point to any specific location (e.g. país [‘country’]), and the remaining three are Turquía [‘Turkey’], Libia [‘Libya’], and Siria [‘Syria’] (the tenth, twelfth and sixteenth most frequently mentioned context indicators within this category, respectively). On the other hand, among the top twenty types referring to geographical and linguistic identities, references to non-European identities prevail (eleven to eight). The word sirio [‘Syrian’] is the second top context indicator in this category with 2,627 occurrences in the corpus. At the same time, however, it is preceded by the word europeo [‘European’] and followed by four other words which refer to European identities: alemán [‘German’], español [‘Spanish’], italiano [‘Italian’], and griego [‘Greek’]. With the exception of griego, all of them have more than a thousand occurrences in the corpus. In total, even though the types referring to European geographical and linguistic identities are slightly less numerous, they have twice more tokens than types referring to non-European identities (14,142 to 7,367). With respect to (groups of) actors discussed in the Spanish press subcorpus, refugees/ migrants are mentioned most frequently. They are followed by slightly less frequent, but more varied, references to politicians and heads of state. Similarly as in the Polish subset, the displaced people are overwhelmingly referred to using general terms, i.e. collectivised, and references to individuals are scarce — the list of the most 135 frequent context indicators includes only the name of Alan Kurdi (mispronounced as Aylan), a toddler who gained widespread recognition when the picture of his dead body was broadcast in the media. References to non-governmental organisations are close to six times less frequent than references to politicians. This analysis has also revealed that among references to places within cities, by far the most frequent is centro [‘centre’], followed by campo [‘camp’]. This indicates that places where refugees are accommodated in Europe (or, possibly, in transit countries) are foregrounded in the corpus. Finally, there are, among context indicators, references to time, gender, family, and group identity. With respect to context indicators in the UK subcorpus (Table 5.3), similarly as in the case of its Polish counterpart, the most populous category includes references to regions, countries, cities, etc. The top twenty context indicators, accounting for 70% of all the tokens in this category, follow the same trend as in the case of both the Polish and the Spanish subcorpora. Specifically, 11 types refer to places in Europe (with UK, Europe, and London being the top three), seven are general words, such as country or sea, and only two denote places outside of Europe (Syria and Libya). Similarly, with respect to geographical and linguistic identities, when the top twenty types are considered, references to European identities are more frequent both in terms of types (ten to eight) and tokens (7,022 to 4,986). 136 Table 5.3. Most Frequent Context Indicators in the UK Subcorpus Concerning the groups of actors most frequently discussed in the UK press coverage of the “refugee crisis”, the top category comprises references to the displaced people. Nevertheless, it is important to note that, similarly as in the case of the other two subcorpora, the number of specific types of these words is very limited, i.e. refugees are referred to using very general terms and there are no occurrences, among the most frequent context indicators, of references to individuals. Other Subcategory Types N=412 Tokens AF=243,420 Tokens RF=92,745 Definition Example types Region/ country/ city/ area/ directions 109 76,019 28,964 Continents, regions, countries, wider areas within a country, cities UK, Europe, London, country Time 67 40,313 15,360 Reference to time points or periods year, daily, month, week Quantification 35 35,636 13,578 Reference to numbers or quantities number, some, many, more Refugees/ migrants 12 34,839 13,274 References to refugees and migrants migrant, refugee Places/ buildings 33 23,920 9,114 Areas within a city, buildings home, camp, centre, Jungle Gender/ family relations 35 22,758 8,671 Identification of people in terms of their gender, age, and family relations child, man, family, woman, boy Governance/ politics 55 18,486 7,043 Heads of state, politicians May, minister, Merkel, leader Geographical/ linguistic identity 40 15,137 5,767 Identification of people in terms of their ethnicity, nationality, place of origin or residence, or their language Syrian, British, English, French, resident Groups/ communities 9 7,747 2,952 Identification of people in terms of their group belonging group, member, team Nongovernment organisations 9 3,404 1,297 Non-governmental groups and their representatives UNHCR, activist, NGO, Amnesty 137 salient groups of actors include government and non-government officials. References to the former, however, are more than five times more frequent than references to the latter. This analysis has revealed the importance of quantification and the salience of words referring to places — in Europe, above all — where refugees are accommodated: camp, centre, Calais, and Jungle are among the top five types in the category Places/ buildings. Finally, there are also references to time as well as family, gender, and group identity. 8.2.2. Topic Indicators With respect to topic indicators, six main discursive trends have been identified in all three language subcorpora, albeit with some differences in their distribution. Specifically, as can be observed in Figure 4, in all three subcorpora the theme of Safety and security is developed most extensively. The UK press, in particular, put special emphasis on this theme, which accounts for close to half of all topic indicators in this subcorpus. It is followed by movement-related words and references to administrative and political issues. With respect to the Polish and Spanish subcorpora, the latter two categories follow the inverse order. At the same time, in the Spanish subset, references to political and administrative issues are only slightly more common than movement-related words (compared to the Polish subcorpus, where this disproportion is more significant). In fact, the latter are as frequent as in the UK subset (24%). 144 bomb (AF=38; RF=14; logDice score = 9.99). Other groups of collocates of attack refer to racist violence, weapons, and a range of evaluative adjectives. Table 7.2. Nouns and Adjectives Collocating with Attack (as Noun) in the UK Subcorpus Note: Cologne and Eve were included in the Sexual violence category because they refer to a series of sexual attacks which were reported to have been committed in Cologne, Germany on New Year’s Eve 2015. Similarly Paris, Berlin, Westminster, truck, lorry, November, and Christmas all refer to the time, place, and mode of several different terrorist attacks. With respect to sexual attacks, in order to answer the second and third research questions for this part of the analysis — who or what causes the threatening situations and who suffers the consequences? — the concordances of attack + sex have been investigated. This has shown that sexual violence is invariably presented as being committed by refugees/ migrants against European women. These actors are Category Collocates Terrorism terrorist, terror, bomb, Paris, Berlin, chemical, suicide, gas, truck, mass, ISIS, lorry, tube, Westminster, jihadist, November, car, Christmas Evaluation bitter, horrific, violent, horror, shocking, devastating Sexual violence sex, Cologne, Eve Type arson, knife, axe Racism racist, racial Consequences deadly Incentive revenge 145 also attributed the same perpetrator versus victim roles in the extensive coverage of the New Year’s Eve 2015 events in Cologne, Germany. With respect to the terrorism-related items, which constitute the most numerous group of collocates of attack, they should be considered alongside another significant category of items indexing the topic of Safety and security, namely Terrorism/ extremism. A detailed analysis has been undertaken with the aim of revealing who the terrorists are, according to the UK press, and what their relationship with refugees is. In other words: are refugees victims or perpetrator of terror? Hence, a word sketch of terrorist (AF=436; RF=166; ARF=160) was obtained and analysed. This has revealed that terrorists in the focus subcorpus are mostly described in terms of their affinity with organised terrorist movements which are invariably associated with Islam (see Table 7.3). The strongest collocate of terrorist is ISIS (AF=19; RF=7; logDice score = 9.73). What is more, one of the collocates relating to terrorist organisations — Islamic — collocates with State (thus making reference to the radical militant group) only four out of eight times. In the other instances terrorist is modified solely by the word Islamic which denotes a religious belief rather than terrorist affiliation. Another way in which terrorism is associated with Islam is the collocation Muslim + terrorist. 146 Table 7.3. Nouns and Adjectives Collocating with Terrorist in the UK Subcorpus What is more, since Muslim is one of the strongest collocates of both refugee and migrant (see Section 8.1), references to Islam, refugees, and terrorism may be said to form a collocation network, where “even distant collocates can participate in the meaning of a word of interest by sharing a similar conceptual space with this word through the link of associations with other words that directly collocate with the word of interest” (Brezina, 2018b, p. 77). Therefore, the negative prosody of Islam might also be extended to refugees and migrants (see Figure 5). Figure 5. Collocation Network: Terrorism — Islam — Refugees/ Migrants in the UK Subcorpus Category Collocates Terrorist movement ISIS, Islamic, jihadi, State, islamist Religion Islamic, Muslim Potential would-be, potential, suspected Method trained, lone-wolf Place truck, train, tube, market Evaluation sick, hell-bent Age teen 147 Another group of salient collocates of terrorist comprises instances when a person is described as having the potential or being suspected of being or becoming a terrorist. As the analysis of their concordances has shown, these collocations, too, invariably refer to refugees/ migrants, for instance: (1.10) (…) all arrivals would be screened to ensure they did not pose a security threat, amid fears that some would-be terrorists have travelled to Britain posing as refugees. (Daily Mail, December 19, 2015) (1.11) And a majority in each state feels there is a risk that potential terrorists may be hiding among the refugees arriving in Europe. (The Guardian, October 30, 2015) The remaining groups of salient collocates of the query term refer to whether the terrorist acts alone or has been trained by someone and to places where terrorist attacks were committed; there are also evaluative adjectives and one reference to age. In contrast to the UK subcorpus, Polish press used general threat-related words more frequently than words referring specifically to physical violence and damage to property, as indicated in Table 7.4. Apart from words which make reference to danger in general, such as problem, kryzys [‘crisis’], etc., it includes a considerable number of words (17; tokens: N = 2,871) which refer to threat to safety (but do not specify the type of threat), such as bezpieczeństwo [‘safety’] or grozić [‘threaten’], as well as conflict words (7; tokens: N = 905), e.g. spór [‘argument’] and przeciwnik [‘opponent’]. 148 Table 7.4. Topic Indicators Indexing the Topic of Threat in the Polish Subcorpus One word, however, points more precisely to the nature of threat that the “refugee crisis” is supposed to present: islamizacja [‘islamisation’] . Islamisation is 29 the outcome of Arab expansion in the form of conversion to Islam and the adoption of Muslim culture by the conquered peoples (Peacock, 2017, p. 1). In the corpus, it is presented as the result of migrations and a major threat for Poland, Europe, and the Subcategory Types N=268 Tokens AF=35,701 Tokens RF=27,957 Definition Example types Danger 57 8,457 6,781 References to danger or problem in general problem, kryzys, konflikt Violence and damage 60 6,843 5,487 References to physical violence and damage in property atak, walka, sprawca, gwałt Response to threat 27 4,450 3,484 References to actions intended to minimise or eliminate perceived threat policja, ochrona, śledztwo Extremism/ terrorism 13 2,438 1,909 References to extreme views and terrorism zamach, terrorysta War 13 2,417 1,892 References to military conflict wojna, armia, wojskowy Legal response to crime 18 2,051 1,606 References to punishment kara, więzienie Protest 18 1,811 1,418 References to actions intended to convey indignation marsz, protest, demonstracja Suffering and death 18 1,770 1,386 References to physical suffering and death śmierć, zginąć Exclusion/ inclusion 20 1,521 1,191 References to exclusion and inclusion based on race, religion, etc. prześladować, rasistowski, ksenofobiczny Psychological suffering 8 1,321 1,034 Fear and other negative outcomes on mental health strach, obawa, straszyć Legality 9 1,183 926 Legal status nielegalny, przemyt Health 7 764 598 References to health and illness choroba, zdrowie It has been included in the Danger subcategory as it does not fit in any of the more specific 29 categories and it has been considered impractical to create a separate category for only one word. 149 Western world in general, which is bound to be annihilated (nadchodzi koniec ich świata [‘the end of their world is coming’], Rzeczpospolita, January 16, 2016). Interestingly, however, almost half of the occurrences of the word islamizacja are found in the only liberal newspaper in the subcorpus: Gazeta Wyborcza. While it is often employed in the critical coverage of anti-immigration manifestations and political movements (when words of other actors are provided in reported speech), the frequent association of immigration with such a negatively evaluated phenomenon might nevertheless contribute to the formation of a negative discourse prosody of migration-related words. With respect to references to damage and physical violence in the Polish subcorpus, the overwhelming majority refer to violent crime (43 types, 5,613 tokens), including 6 explicit references to sexual violence (820 tokens), such as the words gwałt [‘rape’ as noun] (the fourth most frequent word in this subcategory, with 257 occurrences and 201 occurrences per million words) and zgwałcić [‘rape’ as verb] (the eight most frequent word in this subcategory, with 217 occurrences and 169 occurrences per million words). A random sample of 100 concordances of gwałt 30 has shown that 86% of them construe refugees/ migrants as the perpetrators. What is more, rape is sometimes represented as ingrained in refugees’ culture (conflated with religion), for example: (1.16) Imigranci: gwałty na dzieciach to element kultury (…) Zatrzymani mężczyźni to imigranci pochodzący z Somalii. Podczas przesłuchań wszyscy zgodnie deklarowali, że są wyznania muzułmańskiego. Swoje gwałty usprawiedliwiali kulturą, w której byli wychowani. Twierdzili, że gwałty na nastolatkach to coś akceptowalnego przez ich społeczeństwo. [‘Immigrants: raping children is an element of culture (…) The detained men are immigrants from Somalia. During the interrogation they unanimously declared to be of Here and elsewhere, ‘random’ means that the ‘shuffle’ function on the Sketch Engine interface was 30 used to select the concordance lines. 150 Muslim religion. They justified the rapes with the culture in which they had been brought up. They claimed that raping teenagers is acceptable in their society.’] (Fakt, December 10, 2015) Thus, refugees’ culture and religion are construed as completely incompatible with the European values and, most importantly, threatening. Furthermore, the word sketch of the top type in the subcategory of Violence and damage — atak [‘attack’] (AF=593; RF=464; RF=464) — has revealed that it collocates most saliently with words relating to sexual violence and terrorism (however, there are also two collocates relating to racist violence; see Table 7.5). With respect to the former, the third most salient collocate of atak is tło (AF=14; RF=11; logDice score = 9.29) which is predominantly (79%) used in the construction atak na tle seksualnym [‘sexual attack’], the fifth — the adjective seksualny [‘sexual’] (AF=10; RF=8; logDice score = 8.73), and the sixth — Kolonia [‘Cologne’] (AF=10; RF=8; logDice score = 8.71) in reference to the 2015 New Year’s Eve sexual attacks against women. All three collocations invariably construe refugees as perpetrators, as the analysis of the concordances has shown. Table 7.5. Nouns and Adjectives Collocating with Atak [‘Attack’] in the Polish Subcorpus Category Collocates Terrorism terrorystyczny [‘terrorist’], bombowy [‘bomb’], masowy [‘mass’], samobójczy [‘suicide’], (w) Paryż [‘(in) Paris’] Sexual violence seksualny [‘sexual’], (na) kobieta [‘(on) woman’], tło [‘background’], (w) Kolonia [‘(in) Cologne’] Racism rasistowski [‘racist’], ośrodek [‘centre’] Evaluation brutalny [‘brutal’] 151 Collocates of atak referring to terrorism are slightly more numerous, and its top collocate is the adjective terrorystyczny [‘terrorist’] (AF=66; RF=52; logDice score = 11.31). This should be considered alongside the fourth most frequent trend in the subcorpus, namely the extremism/ terrorism frame. Therefore, similarly as in the case of the UK subcorpus, the word terrorysta [‘terrorist’] (AF=462; RF=370; ARF=209) is examined in more detail. Its word sketch has shown that there are two main frames of representation of terrorists in the Polish subcorpus: one refers to a person’s intentions of carrying out a terrorist attack, the other to their religious/ geographical identity. The former is lexicalised using words such as potencjalny [‘potential’] — the second most salient collocate of the query term (AF=16; RF=13; logDice score = 11.64) — niedoszły [‘would-be’], and domniemany [‘alleged’]. As the analysis of the concordances of terrorysta + potencjalny has revealed, they almost invariably link migrations with terrorist threat, for instance: (1.17) (…) ostrzega się przed muzułmańskimi uchodźcami jako potencjalnymi terrorystami [‘(…) there are warnings against Muslim refugees as potential terrorists’] (Gazeta Wyborcza, September 16, 2015). While all of them come from the only liberal newspaper in the corpus, Gazeta Wyborcza, which only uses them in reported speech and in all cases refutes the damaging association, its frequent use might nevertheless contribute to the formation of a negative discourse prosody of migration-related words. The other group of modifiers of terrorysta comprises words which associate terrorist activities with religious beliefs. This group includes the strongest collocate of the query term, i.e. islamski (AF=28; RF=22; logDice score = 10.65), as well as other words referring to Islam: islamistyczny [‘islamic’] and muzułmański [‘Muslim’]. Such collocations carry a strong negative semantic prosody of Islam (see Baker et al., 152 2013). Because muzułmański and islamski are also some of the strongest collocates of both uchodźca and imigrant (see Section 8.1), references to Islam, refugees, and terrorism might, again, be said to form a collocation network, which results in the negative prosody being extended to refugees and migrants (see Figure 7). Figure 7. Collocation Network: Terrorism — Islam — Refugees/ Immigrants in the Polish Subcorpus In the Spanish press, similarly to its Polish counterpart, references to danger or problem in general prevail within the larger category of Safety and security (see Table 7.6). They are followed by references to political and institutional solutions to events construed as problematic. References to physical violence and damage to property, in turn, are much less frequent. Specifically, they are almost twice less numerous than words indexing danger in general. This indicates that the Spanish press focused on discussion of danger in general as well as solutions to eliminate it rather than specific violent events, unlike its Polish and British counterparts. 153 Table 7.6. Topic Indicators Indexing the Topic of Threat in the Spanish Subcorpus Another difference between the Spanish and the other two subcorpora concerns references to damage and physical violence: it contains much less items relating to sexual violence (only two words: violar and violación [‘rape’ as verb and noun]). Similarly, ataque [‘attack’], which is the second most frequent word in this subcategory (AF=625; RF=252), does not collocate with any words referring to sexual Subcategory Types N=281 Tokens AF=73,543 Tokens RF=29,674 Definition Example types Danger 95 24,510 9.889 References to danger or problem in general crisis, problema Response to threat 39 13.827 5.579 References to actions intended to minimise or eliminate perceived threat control, guardacostas, enfrentar Violence and damage 43 9,430 3.804 References to physical violence and damage in property ataque, golpe, violencia, herir Legal response to crime 27 5.761 2.324 References to punishment denunciar, acusar, juez Suffering and death 20 5.584 2.253 References to physical suffering and death morir, sufrir, víctima Legality 13 3.967 1.600 Legal status legal, illegal War 5 2.830 1.141 References to military conflict guerra, militar Extremism/ terrorism 13 2.475 998 References to extreme views and terrorism terrorista, atentado Psychological suffering 8 1.984 800 Fear and other negative outcomes on mental health miedo, temer Health 6 1.284 518 References to health and illness médico, sanitario Exclusion/ inclusion 8 1.158 467 References to exclusion and inclusion based on race, religion, etc. persecución, odio, racista Protest 4 659 265 References to actions intended to convey indignation protesta, manifestar 256 modest efforts (bez wątpienia nie jest wysiłkiem ogromnym [‘without a doubt, it does not require tremendous effort’]). In the Spanish press Europe is mostly evaluated as incapable (incapaz, incapacidad), retrograde (retrógrada), increasingly ineffective (cada vez menos efectivo), of increasingly questionable capacity (capacidad (…) se ve cada vez más cuestionada), paralysed (paralizada), alarmingly decadent (alarmante decadencia), chaotic (descoordinación), unwilling (falta de voluntad política [‘lack of political will’]), apathetic (inacción), betraying itself (una Europa solidaria se está traicionando [‘a solidary Europe is betraying itself’]), as well as lacking empathy and solidarity (falta de empatia y solidaridad). The newspapers published in Spain also mentioned frequently the “deficiencies” of the European asylum systems (deficiencias estructurales [‘structural deficiencies’], “deficiente” sistema de asilo [‘deficient asylum system’]). Similarly, solutions proposed by European institutions were qualified as insufficient (soluciones fáciles para un problema complejo [‘easy solutions for a complex problem’]), absolutely unacceptable from the ethical point of view (absolutamente rechazable desde el punto de vista ético), objects of the utmost caution (objeto de todas las cautelas), incomprehensible (incomprensible), a shame for democracy (un bochorno para la democracia), a great fraud (gran fraude), a disgrace (vergüenza), unacceptable (inaceptable), a failure (fracaso), as well as “too late, too slow, and chaotic” (Tarde, muy lentamente y de forma descoordinada, la Unión Europea empieza a reaccionar [‘The European Union begins to act but too late, too slowly, and chaotically’]). 257 Negative evaluations of the political response towards migrations were least frequent in the UK sample. Furthermore, unlike Polish and, above all, Spanish press, its criticism focused on the home asylum system and the local government’s response. Specifically, with respect to the former, the Home Office staff were referred to as being under pressure, making poor decisions, as well as “undertrained, overworked and operating in a ‘constant state of crisis’” in The Guardian’s December 25, 2017 article; whereas its immigration policy was described as “controversial” and “setting a worrying precedent”. With respect to the latter, Daily Mirror, in its September 3, 2015 news report, quoted a TV presenter claiming that Britain’s attitude to refugees was “selfish. Inhuman. Indifferent. Uncharitable. Unchristian. Unpleasant”. With respect to positive evaluations of the European response to the “refugee crisis”, very few instances have been identified in the analysed sample. In the Polish and Spanish subcorpora they all referred to Germany and its political leader, Angela Merkel. Specifically, Gazeta Wyborcza (September 3, 2015) referred to the former as an “outstanding exception” (wybitny wyjątek) to the otherwise largely antiimmigrant European attitudes. Spanish newspapers described Angela Merkel’s leadership as brave and exemplary (valiente liderazgo y ejemplo), unique (solo emergió un lider digno de ese nombre [‘there has been only one leader worthy of this name’]), at the vanguard (a la vanguardia), and impeccable (impecable). In the UK sample there were no instances of positive evaluation of institutional response other than one text which praised the country’s own “proud history of hosting, supporting and protecting those in need, including some of the most vulnerable children affected by the migration crisis” (The Guardian, October 8, 2017). 258 Finally, the analysis of discourse markers was less useful in detecting the humanitarian aid stance, because they rarely linked the arguments to the conclusion. This is, obviously, related to the fact that the humanitarian aid stance was expressed explicitly much less frequently in the analysed sample. Nevertheless, the analysis has shown a limited use of markers expressing causality, e.g.: (3.3) Por eso es urgente que Europa aborde una política de asilo y de gestión de la migración que no sea meramente reactiva, reparta las cargas entre todos los países y garantice el respeto a los derechos humanos. [‘This is why Europe urgently needs to implement an asylum and migration policy, which would be not just reactive, to distribute the load among all countries, and to guarantee that human rights are respected.’] (El País, June 12, 2018); as well as sequence, e.g.: (3.4) In the wake of the heart wrenching images of the bodies of Syrian children washed up on a Turkish beach many Britons have begun to spread messages on social media telling refugees they are welcome in the UK. (Daily Mirror, September 3, 2015) Humanitarian aid was also sometimes represented as a preferable alternative to securitising approaches, for instance: (3.5) En vez de acoger y asistir a las víctimas de estos conflictos, se construyen muros para que no entren. [‘Instead of accommodating and assisting the victims of these conflicts, walls are constructed so that they cannot pass.’] (El Mundo, December 20, 2017) (3.6) Rather than looking at their own interests only, nations will have to consider their moral responsibilities to those who are suffering. (Daily Mirror, November 21, 2015) and as the most important first step in the management of the “refugee crisis”: (3.7) El Gobierno español se alinea con quienes consideran que ante todo hay que dar una respuesta humanitaria a las emergencias (…). [‘The government agrees with those who 259 think that, above all, a humanitarian response to the emergencies is necessary (…).’] (El País, June 12, 2018) 9.2.2. Arguments and Topoi Argumentation supporting the humanitarian aid claim has been revealed to be much less extensive than argumentation supporting the inverse conclusion. First of all, less individual arguments have been identified. Second of all, the articles in the sample usually focused on developing one argument rather than combining a variety of arguments, as was the case with texts advancing the opposite thesis. As a result, a significant number of the identified arguments were only found in one or two separate texts in the sample. Third of all, these arguments were often largely based on examples. For instance, a text aiming at construing refugees as willing/ able to integrate and contribute to the host society would typically present the life story of a specific person who managed to start a family, open a business, etc. in their country of destination. Finally, a significant number of the identified arguments were negative arguments aimed at providing counter-responses to prejudicial and often racist anti-immigration discourses, for instance by arguing that refugees are not terrorists (Argument H7). The arguments for helping refugees can be divided loosely into three groups (see Table 14). The first group of arguments construes humanitarian aid in terms of an ethical or legal obligation (Arguments H1 and H2). They are followed by the most numerous group of arguments employing different strategies of positive (or neutral) presentation of refugees (Arguments H3 through H9). The most frequent and extensively developed argument in this group represents refugees as victims 260 (Argument H3). Argument H4 focuses on the perceived vulnerability of refugee women and children. The remaining arguments which form part of this subgroup are, as has already been mentioned, negative arguments (e.g. refugees are not terrorists), where refugees are often conflated with Muslims. The last group comprises arguments which refer to Europe’s response to the “refugee crisis”: its resources and capability to help (Arguments H10 through H12), the potential benefits for Europe of accommodating refugees (Arguments H13 and H14) as well as the Europeans’ attitudes towards them (Arguments H15 and H16). Finally, the last argument aims at normalising migrations by construing them as a universal and unavoidable phenomenon in human history. Table 14. Arguments For the Humanitarian Aid Stance № Argument Poland UK Spain H1 Helping refugees is Europe’s ethical responsibility. x x x H1a Others helped us when we were in need. x H1b Helping refugees is Europe’s religious responsibility. x H2 Europe has a legal obligation to help refugees. x x H3 Refugees are victims. x x x H4 There are many refugee women and children. x x H5 Not all refugees are Muslims. Muslims are not a homogenous group. x H6 Refugees will not islamise Europe. x H7 Refugees are not terrorists. x x x H7a Victims of terrorist violence are not terrorists. x H7b Displaced people do not have the resources to carry out a terrorist attack. x H7c ISIS does not equate with Islam. x 261 Table 14. (Cont.) Arguments For the Humanitarian Aid Stance Argument H1: Helping refugees is Europe’s ethical responsibility. This argument construes helping refugees in terms of a moral or social obligation in general (examples 3.8 through 3.10) or makes reference to specific moral values which are said to form Europe’s bedrock, such as solidarity, compassion, generosity, etc. (examples 3.11 and 3.12). It can be seen as a counterargument to Argument S22. (3.8) Tenemos el deber moral y político de acoger a los refugiados sirios [‘We have a moral and political duty to accommodate Syrian refugees’]. (La Vanguardia, September 4, 2015); (3.9) Rather than looking at their own interests only, nations will have to consider their moral responsibilities to those who are suffering. (Daily Mirror, November 21, 2015) H7d Other European nations have lived in coexistence with Muslims for many years. x H8 Anti-immigrant discourses lead to Muslim radicalisation. x H9 Anti-immigrant discourses are false. x H10 Not so many refugees require help. x x x H11 Differentiating between refugees and migrants is possible. x H12 Europe is capable of helping refugees. x x x H13 Europe benefits economically from migrant workforce. x H14 Refugees have the potential of becoming good/ prominent citizens. x x x H15 People support humanitarian aid. x x x H16 Even conservative politicians support the refugee cause. x H17 Migrations are a universal issue. x x x 262 (3.10) Ochrona uchodźców to zarówno moralny, jak i prawny obowiązek. [‘Protecting refugees is both a moral and a legal obligation.’] (Gazeta Wyborcza, September 3, 2015) 40 (3.11) La solidaridad alemana con los refugiados es un ejemplo de lo que debería ser Europa [‘German solidarity with refugees is an example of what should Europe be’]. (El Mundo, September 4, 2015) (3.12) This is the crisis of our times, and how we respond to it is a test of our values, our spirit, our ingenuity, our generosity. (The Guardian, November 24, 2015) On the other hand, however, the instantiations of this argument were often accompanied by counter-arguments, which decreased their argumentative force. For example, the fragment presented in excerpt 3.9 was immediately followed by the stipulation that the European nations do not have a moral responsibility towards all the displaced people but only those who fit some unspecified set of criteria (see Argument S8 as well as Section 8.1 for a more detailed discussion of how refugeeness is often contested in the analysed news reports): (3.13) Proper control measures should be in place to make sure that the refugees are genuine and not posing as refugees. (Daily Mirror, November 21, 2015) Similarly, the Polish liberal daily Gazeta Wyborcza explained that refugee protection (see excerpt 3.10), in spite of good intentions, is not, in fact, easy: (3.14) To nie jest łatwe zadanie, ale nie jest niemożliwe. [‘This is not an easy task but it is not impossible.’] (Gazeta Wyborcza, September 3, 2015) Its British counterpart, on the other hand, while admitting that Europe “has a moral duty to help Syrians”, also emphasised that this responsibility should be shared: (3.15) If Europe has a moral duty to help Syrians, then the Gulf certainly does too. (The Guardian, September 4, 2015) See Argument H2. 40 263 This, in consequence, shifts the attention towards those other agents who are, moreover, assumed to be avoiding the duty imposed on them. Subargument H1a: Others helped us when we were in need. In the Polish sample, there were two additional subarguments which specified why helping refugees is Europe’s moral responsibility. The first one, by comparing the plight of the displaced people to the historical experience of migrations of Poles, argued that Poles should help refugees out of gratitude for the help they received in similar circumstances in the past. This argument is also related to Argument H19 which postulates that there will always be refugees in need of help by pointing at the universal nature of the refugee experience which is not limited neither temporally, nor spatially. The argument was observed in the following excerpt: (3.16) To, że nie chcemy przyjmować imigrantów, jest przejawem niewdzięczności z naszej strony. Przecież kiedyś, będąc w potrzebie, mogliśmy liczyć na to, że inne państwa pozwolą Polakom zamieszkać na swoim terytorium! [‘The fact that we don’t want to accommodate immigrants speaks to our ingratitude. In the past, when we were in need, we could count on other countries to let Poles live in their territories!’] (Super Express, January 17, 2018) Subargument H1b: Helping refugees is Europe’s religious responsibility. The second subargument observed in the Polish sample which points at Europe’s moral duty to accommodate refugees appeals to the principles of Christian faith. It was used as a counter-discourse in response to the anti-immigration rhetoric which sees the obligation to help refugees as limited to Christians (see, for instance, 264 Argument 3 of the preceding section). Thus, excerpt 3.17 points out that this interpretation goes against the teachings of Jesus: (3.17) Zasada: "Pomagać tylko chrześcijanom" jest niechrześcijańska, gdyż zaprzecza Ewangelii Jezusowej. [‘The principle to “only help Christians” is un-Christian because it contradicts the Gospel of Jesus.’] (Gazeta Wyborcza, November 14, 2015) The other fragment where this subargument was observed (3.18) emphasises the perceived dissonance between the proclaimed religious beliefs of Poles and their 41 attitudes towards out-groups: (3.18) Określamy się jako naród katolicki, a jesteśmy jedną z najbardziej agresywnych społeczności. Narodem wierzącym, a nierespektującym głosu Kościoła. [‘We consider ourselves to be a Catholic nation, but we are one of the most aggressive societies. A nation of believers who do not respect the voice of the Church.’] (Super Express, January 17, 2018) It is therefore a strategy of negative self-presentation, which aims at pointing out the inconsistencies between the beliefs of the group and their actions as well as stressing the necessity to follow the Church’s guidelines with respect to refugees. Argument H2: Europe has a legal obligation to help refugees. Several texts in the Polish and Spanish samples discussed Europe’s legal obligation to help refugees either in general terms or with respect to (usually breaches of) specific international agreements regarding the protection of refugees (in contrast, see Argument S18 for a discussion of discourses presenting antiimmigrant international agreements as legally justified). With respect to the former, El Mundo (June 12, 2018) called it “a moral and judicial duty” (deber moral y jurídico), whereas Gazeta Wyborcza (November 14, 2015) mentioned the right to seek asylum of specific national groups: It assumes that the majority of Poles are Catholic, possibly as suggested by the results of the last 41 census (preceding the publication of this text) when 88% of Poles declared to be Catholic (Główny Urząd Statystyczny [Statistics Poland], 2011). 265 (3.19) Syryjczykom z Jordanii, Libanu i Turcji prawo azylu w UE więc przysługuje. [‘Syrians from Jordan, Libya, and Turkey therefore have the right to asylum in the EU.’] With respect to the latter, El Mundo referred to the fact that the “refugee quotas” had not been reached as vergüenza [‘shame’] and gran fraude [‘big fraud’] (September 26, 2017); whereas Rzeczpospolita warned of legal (as well as financial) consequences of rejecting them: (3.20) Rządy będą do tego (przyjmowania uchodźców) zmuszone wyrokami europejskich sądów i wizją kar finansowych. [‘European courts’ decisions and the prospect of fines will obligate the governments (to accommodate refugees).’] (September 3, 2015) Similarly, securitisation measures were sometimes qualified as illegal, for instance: (3.21) (…) vías jurídicamente dudosas para salir del apuro [‘(…) legally doubtful ways of getting out of this predicament’]. (El País, June 11, 2018) This, however, does not necessarily and unequivocally point at the humanitarian aid solution. Furthermore, the movement of the displaced people towards Europe was in the above-quoted fragment referred to as apuro — a difficult and urgent situation for the European Union rather than for the migrants. Therefore, it is possible that sometimes the aim of this type of reasoning is to express concern for Europe which may suffer harmful consequences as a result of the migrations. Argument H3: Refugees are victims. This argument construes refugees as victims in need of humanitarian assistance: first, with respect to the different adversities they suffer in their countries of origin. Second, to a lesser extent, hardships the displaced people suffer during their journeys are mentioned (including the treatment in refugee detention centres). The third group of the instantiations of this argument focuses on the negative outcomes on refugees’ health, including those resulting in death. Due to space 272 Argument H6: Refugees will not islamise Europe. The second argument in this set addresses the fears that immigration from the Middle East and Africa leads to the islamisation of Europe: (3.43) Czy uchodźcy wprowadzą w Czechach prawo szariatu, czym jesteśmy straszeni każdego prawie dnia? Szariat to prawo religijne i obowiązuje tylko muzułmanów. Żydzi i chrześcijanie, nawet w krajach islamskich, nie byli nigdy do szariatu zobowiązywani. Postępowali zgodnie z własnym prawem. Jest więc absurdem bać się, że kilka tysięcy uchodźców - nawet gdyby wszyscy byli prawowiernymi muzułmanami - spowoduje, że nasz kraj stanie się krajem islamskiego prawa religijnego. [‘Will refugees introduce the sharia law in the Czech Republic, as we are warned almost every day? Sharia is a religious law and affects only Muslims. Jews and Christians, even in Islamic countries, were never obliged to adhere to the sharia. They followed their own laws. So it is absurd to fear that a few thousand refugees — even if they were all pious Muslims — would change our country’s laws into Islamic religious laws.’] (Gazeta Wyborcza, November 14, 2015) It supports its claim by mentioning refugee figures — which are considered to be insufficient to lead to islamisation — and builds on the preceding argument: Muslims are not a homogenous group (specifically: not all Muslims are devoutly religious). It also uses a subargument: Jews and Christians living in Islam-controlled countries were never forced to adhere to sharia law. It concludes by referring to the fear of islamisation as “absurd”, thus presenting anti-immigration discourses as unreasonable and irrational. Argument H7: Refugees are not terrorists. Texts in all three samples attempted to present counter-arguments to one of the most harmful misconceptions about Muslim refugees: that they are (potential) terrorists (see Argument S1). With regard to the UK sample, The Guardian declared its alignment with the views of the US President Barack Obama who said that 273 (3.44) (…) in the aftermath of the Paris attacks, we must not start to equate the issue of refugees with terrorism. (November 24, 2015) Similarly, in the Spanish subset, there was only one instantiation of this argument, also provided in relation to a politician’s speech: (3.45) Juncker, pidió ayer a los europeos "no confundir" terroristas con refugiados, quienes, precisamente, "huyen de la filosofía y la mentalidad que inspiran" esos actos (…) y recalcó que "son criminales, no refugiados” [‘Yesterday Juncker asked the Europeans “not to mistake” terrorists for refugees who, in fact, “flee the philosophy and the mentality which inspire” these acts (…) and emphasised that “they are criminals, not refugees”’]. (La Vanguardia, November 16, 2015) It is, however, important to notice how the non-defining relative clause referring to refugees makes the final part of the sentence somewhat ambiguous. With respect to the Polish sample, this argument was, again, found above all in the text quoted in the two preceding sections (Arguments H5 and H6), published in Gazeta Wyborcza on November 14, 2015 (following a series of terrorist attacks in Paris). Another instantiation of this argument has also been observed in a different article from the same newspaper (November 16, 2015). Moreover, in both texts it was supported by a series of subarguments which are discussed below. Subargument H7a: Victims of terrorist violence are not terrorists. (3.46) Po pierwsze, setki tysięcy uchodźców uciekają od bomb, są więc oni jak najdalsi od pomysłu, żeby tam, gdzie znaleźli schronienie, podkładać bomby. [‘First of all, hundreds of thousands of refugees are fleeing bombs so they are as unlikely as possible to plant bombs where they found shelter.’] (November 16, 2015) This subargument purports that people who are fleeing from terrorist violence are, naturally, not involved in it. Therefore, they are “as unlikely as possible” to engage in violent activities in places where they found safety. 274 Subargument H7b: Displaced people do not have the resources necessary to carry out a terrorist attack. (3.47) Po drugie, ludzie, którzy rozpaczliwie przedzierają się do Europy na małych łodziach, często po drodze tonąc, nie byliby w stanie zorganizować i przeprowadzić skoordynowanej serii ataków i zamachów bombowych. [‘Second of all, people desperately trying to reach Europe on small boats and often drowning would not be able to organise and carry out a coordinated series of bomb attacks.’] (November 16, 2015) As can be observed in fragment 3.46, carrying out a terrorist attack is presented as a sophisticated and difficult operation, whereas refugees lack the necessary resources as attested by their means of transportation which does not even guarantee safe passage to Europe. Therefore, the text implies that since refugees do not even have the means to be able to travel safely, they are hardly in a position to be involved in more complex activities, such as terrorism. Subargument H7c: ISIS does not equate with Islam. This fragment compares the activities of ISIS to other historical religiouslymotivated crimes in order to draw the conclusion that committing violent acts in the name of religion does not reflect on religion itself: (3.48) Państwo Islamskie (…) ma z islamem raczej tyle wspólnego, ile mieli irlandzcy terroryści czy ci, którzy palili czarownice, z Ewangelią. [‘ISIS is related with Islam in the same way that Irish terrorists or those who burned witches had to do with the Gospel.’] (November 14, 2015) 275 Subargument H7d: Other European nations have lived in coexistence with Muslims for many years. Finally, the last subargument aimed at convincing the readers that (Muslim) refugees are not terrorists appeals to the purportedly peaceful coexistence of Islam and other religions in Western countries: (3.49) W Niemczech, Anglii, Francji od dziesiątków lat żyją miliony muzułmanów. Miejscowi obywatele ich znają, wiedzą, że olbrzymia większość z nich nie ma nic wspólnego z okropnościami islamskiego terroryzmu. [‘Millions of Muslims have been living in Germany, England, and France for several decades. The locals know them and are aware that the vast majority have nothing to do with the atrocities of Islamist terrorism.’] (November 14, 2015) Nevertheless, similarly to fragments 2.33 and 3.50, it also presents Muslims as not belonging to the general population as the phrase miejscowi obywatele [‘local citizens’] refers to anyone but Muslims (even though the latter are also represented as occupying the same geographical space). Argument H8: Anti-immigrant discourses lead to Muslim radicalisation. Several texts in the Polish sample analysed appeal for more positive attitudes towards refugees by claiming that hostility may lead to their radicalisation. Crucially, however, this argument is based on the very prejudicial topos that Muslims are prone to becoming radicalised which may, in fact, lead to the adverse conclusion (see Argument S1d). The following example was retrieved from a tabloid newspaper; the same argument has, however, been observed in the liberal Gazeta Wyborcza as well: (3.50) Im bardziej będą one [nastroje ksenofobiczne] radykalne, tym bardziej radykalni staną się żyjący w Europie muzułmanie, dla których niechęć wobec nich wyrażana przez Europejczyków, podkreślanie przez nich, że muzułmanie są obcymi, to pchnięcie wyznawców islamu w ręce radykałów i pogłębienie problemów, które z fundamentalizmem 276 islamskimi mamy już dzisiaj. [‘The more radical [xenophobic attitudes] are, the more radical the Muslims living in Europe will become. Europeans’ hostility towards them and stressing that Muslims are different pushes people of Islamic faith towards the hands of radicals and worsens the problem we already have with Islamist fundamentalism.’] (Super Express, November 15, 2015) Furthermore, despite what the text seemingly advocates, it also employs the us versus them distinction between (Christian) Europeans and “Muslims living in Europe” (temporarily, outside of the remit of the society) (see also excerpts 2.63 and 2.43). Argument H9: Anti-immigrant discourses are counterfactual. This argument is another example of counter-discourse. It does not, however, provide any actual reasoning (other than simply stating that anti-immigrant rhetoric is false). It has been observed uniquely in the Spanish subset: (3.51) Cuando el ministro italiano del Interior, Matteo Salvini, proclama que "se acabó la buena vida de los inmigrantes", no solo está fomentando una imagen distorsionada de la realidad, sino que alienta la persecución [‘When the Italian Minister of Interior Affairs Matteo Salvini claims that “good life for immigrants is over”, he is not only promoting a distorted image of the reality, but also encouraging persecution’]. (El País, June 12, 2018) Argument H10: Not so many refugees require help. Migration figures have also been found to be used in support of humanitarian aid measures, albeit differently and less frequently than in texts advocating for securitising approaches (see Argument S13). Specifically, when numbers were provided to argue for helping refugees, they were referred to as small, e.g.: (3.52) una estrategia (tímida en valores absolutos, pionera en relativos) [‘a strategy (timid in absolute terms, pioneer in relative terms)’]. (El Mundo, September 4, 2015) (3.53) To oznacza, że 26 innych krajów UE zajmuje się łącznie niewiele ponad 180 tys. wniosków azylowych, co bez wątpienia nie jest wysiłkiem ogromnym. [‘This means that the 277 other 26 EU countries are responsible for little more than 180 thousand asylum applications which, doubtless, is not a huge effort.’] (Gazeta Wyborcza, September 3, 2015) As a result, these refugee figures could be presented as “managable”, as in excerpt 3.53, which postulates that processing of 180 thousand applications is a relatively easy task, or excerpt 3.54 from El País: (3.54) Ahora son menos de 200.000 los que llegan al año. La cifra sigue siendo importante, pero es mucho más manejable [‘Now less than 200.000 arrive every year. The number remains significant but it is much more manageable’]. (June 12, 2018) Similarly, Yvette Cooper, a politician quoted in The Guardian, used statistical information to advance the view that from the perspective of a smaller geographical area accommodating a significant number of refugees in the UK could hardly be considered radical demographic change: (3.55) If every area in the UK took just 10 families, we could offer sanctuary to 10,000 refugees. (September 3, 2015) This, along with Argument S13, shows how statistical information can easily be manipulated and used by the media to reach a variety of different or even opposing conclusions. Argument H11: Differentiating between refugees and migrants is possible. One Polish newspaper perpetuated the distinction between migrants and refugees in order to appeal for helping the latter. It therefore presented humanitarian aid as contingent on a set of prejudicial criteria in the same way Argument S8 uses these categories in order to advocate for rejecting those who are labelled as “immigrants”. (3.56) Ochojska uspokaja też, że są odpowiednie procedury, które pozwalają odróżnić tzw. imigrantów ekonomicznych od tych, którzy są zmuszeni uciekać ze swoich krajów przed wojną i głodem. [‘Ochojska eases the fears by explaining that there are procedures which 278 allow us for differentiating the so-called economic immigrants from those who are forced to flee war and hunger in their countries.’] (Super Express, September 4, 2015) Nevertheless, it also conflated the two — supposedly separate — categories, as it included people fleeing poverty in the “deserving” group, while it remained difficult to ascertain whom is included in the category of “economic immigrants”. This, ultimately, made the distinction and the criteria upon which it was based fuzzy and unclear, thus potentially placing all migrants within the undeserving category. Argument H12: Europe is capable of helping refugees. This argument, in response to anti-immigration discourses (see, above all, Argument S12), presents European countries as being prepared financially and organisationally to help refugees. With respect to the former, it was used in two Polish texts, both published in the same tabloid newspaper, Super Express, for instance: (3.57) Unia Europejska chce, by Polska przyjęła 10 tys. uciekinierów z ogarniętych wojną rejonów, tymczasem rządzący nie wypowiadają się o tym pomyśle przychylnie, twierdząc, że naszego kraju nie stać na przyjęcie tak dużej rzeszy ludzi. Janina Ochojska, założycielka Polskiej Akcji Humanitarnej, twierdzi, że taki argument jest bzdurny, bo Unia na każdego przyjętego uchodźcę płaci 10 tys. euro. [‘The European Union wants Poland to accommodate ten thousand refugees from war-torn regions but the government is not keen on this idea: they claim that our country cannot afford to accommodate such a huge mass of people. Janina Ochojska, founder of Polish Humanitarian Action, claims that this argument is absurd because the EU pays 10 thousand euros per refugee.’] (September 4, 2015) The argument was thus formed in response to the inverse claims pronounced by Polish politicians (referred to as “absurd”). In the other two subsets this argument was only mentioned as part of the statement pronounced by an English cardinal of the Roman Catholic Church, Vincent Nichols: 279 (3.58) «Estamos dejando morir a la gente y recogiendo cadáveres en las playas, cuando Europa es un sitio rico, que debería ser capaz de poner en marcha una respuesta a corto plazo» [‘“We are letting people die and collecting corpses from beaches even though Europe is a rich place which should be able to respond quickly”]. (El Mundo, September 4, 2015) (3.59) “Europe is such a wealthy place (…) What is screaming out is the human tragedy of this problem, to which we can be more generous”. (The Guardian, September 3, 2015) Notice also the use of another argument — refugees are victims in need of help — which further supports the conclusion. The latter, on the other hand, was used only in the Spanish sample — for instance in the above-quoted excerpt 2.60 which, apart from mentioning economic resources, also refers to the capabilities to respond quickly and efficaciously to an urgent situation. However, it has been observed mostly in reference to the arrival of Aquarius in Valencia, as in the following example: (3.60) Cuando el barco atraque en Valencia encontrará una red preparada para recibir a los rescatados porque esta y otras muchas ciudades españolas tienen desde hace tiempo dispositivos de acogida que apenas se han utilizado [‘When the ship docks in Valencia, it will be met with a network prepared to receive the rescued people because this and many other Spanish cities have had mechanisms of reception in place, which have rarely been used’]. (El País, June 12, 2018) Argument H13: Europe benefits economically from migrant workforce. One text in the analysed sample (Rzeczpospolita, September 12, 2016) suggested that Europe needs immigrants to provide workforce: (3.61) Najnowsze badania wskazują, że w Polsce nasila się niechęć do imigrantów, choć bez nich nasza gospodarka nie może się już obejść. [‘Recent opinion polls indicate that hostility towards immigrants in Poland increases even though our economy cannot do without them.’] However, it is important to note that the text implied that it only considered guest workers from Eastern Europe to be useful workers, whereas the “refugee crisis” is 280 only discussed as the purported reason for the increasingly hostile anti-immigrant public attitudes. Argument H14: Refugees have the potential of becoming good/ prominent citizens. This argument employs strategies of positive other-presentation in order to construe refugees as having the potential of becoming good or even prominent European citizens (contrary to counter-discourses, as observed in Arguments S1 through S3). It thus portrays them as similar or even morally superior to Christians. It is usually the sole focus of a given article which describes a success story of a particular person. The UK press seems to have been particularly interested in these stories, based on the analysed sample. For instance, on April 10, 2018 The Guardian published an article titled “The Afghan cop who became a German nurse and other refugee success stories”, whereas the Daily Mirror discussed the case of a refugee Olympic swimmer (June 10, 2018) and a Cambridge University PhD student “smuggled into Dover in a lorry” (October 24, 2017). With respect to the Spanish sample, one article (El País, June 30, 2016) related the success story of a refugee in the UK. While it seemed to suggest that it is not the only refugee success story, the argument is nevertheless fallacious as it is based uniquely on this example. Furthermore, it is construed as insufficient: (3.62) Las historias de éxito e integración como la de Salah no logran apaciguar los ánimos en Dover [‘Stories of success and integration, like the story of Salah, are not enough to appease the moods in Dover’]. Another Spanish article, published in La Vanguardia (June 3, 2018), construed refugees as not only able to succeed in Europe but even superior to Europeans. The article focuses on people of immigrant background who committed particularly courageous and laudable acts and refers to them as “guardian angels” (ángeles 281 guardianes). It suggested that immigrants are particularly likely to become heroes, thus committing the fallacy of false generalisation: (3.63) Y los inmigrantes ganan por goleada en esa lista de héroes [‘And immigrants win by a landslide in this list of heroes’]. Finally, in the Polish sample this idea was only expressed in one text but, contrary to the other two subsets, in a more general way: (3.64) Islam zobowiązuje uchodźców do nieislamskich krajów do ścisłego przestrzegania miejscowego prawa i zwyczajów. Zdecydowana większość muzułmanów żyjących w Europie miejscowe prawo uznaje i zachowuje. (…) Wśród tych, którzy są naprawdę gorliwymi, wierzącymi muzułmanami, wielu zachowuje przede wszystkim te etyczne zasady islamu, które są podobne do chrześcijańskich. Swoim przestrzeganiem zasad etycznych - takich jak uczciwość w handlu, solidarność z biednymi czy troska o rodzinę - mogliby zawstydzić niejednego chrześcijanina. [‘Islam obliges refugees to non-Muslim countries to strictly comply with local laws and customs. The vast majority of Muslims living in Europe respects and follows the local law. (…) Among those who are truly devout Muslims, many follow only those ethical rules of Islam which are similar to the Christian ones. Many Christians would pale in comparison to their adherence of ethical rules such as honesty in trade, solidarity with the poor, or family caregiving.’] (Gazeta Wyborcza, November 14, 2015) At the same time, the strength of this argument is greater: on the one hand, it acts at the level of the topos which usually underlies discourses on the purported unwillingness of Muslim immigrants to adhere to the “Western” laws by claiming the exact opposite. On the other hand, it applies the argument to “the vast majority of Muslims” rather than an individual. Nevertheless, it was the only instantiation of this argument in the Polish sample. 288 Table 15.1. Metaphors of Immigration As Table 15.1 shows, the Spanish press was most likely to employ metaphors in the representation of refugees and immigrants, followed by the UK, and the Polish newspapers. The displaced people were, above all, construed as commodities and threatening natural water phenomena. COMMODITY metaphors were particularly frequent in the Spanish subcorpus (where they accounted for 60% of all the detected metaphors, and had the highest number of lexicalisations) and relatively unlikely to be used by the Polish press. Instead, the Polish press had a preference for WATER metaphors which accounted for 60% of all metaphors identified in this subcorpus (even though the number of their lexicalisations is limited). The UK press was, in turn, unique in its preference for ANIMAL and GUEST metaphors. Finally, the least common metaphorical mapping represented refugees and immigrants as invaders. All of these metaphors were employed, albeit to different extents, regardless of the stance advocated by a given text, i.e. both in texts for and against accommodating refugees in Europe. Each identified mapping is discussed in a separate Section below. Metaphor Spain UK Poland Total COMMODITY AF= 3,671 (RF=1,481) AF=1,878 (RF=716) AF=300 (RF=235) 5.849 LIQUID AF=1,917 (RF=774) AF=1,003 (RF=382) AF=1,158 (RF=907) 4.078 ANIMAL AF=294 (RF=119) AF=1,282 (RF=488) AF=106 (RF=88) 1.682 (UNWANTED) GUESTS AF=94 (RF=38) AF=86 (RF=33) AF=162 (RF=127) 1.228 INVADERS AF=151 (RF=61) AF=972 (RF=370) AF=151 (RF=118) 388 Total AF=6,127 (RF=2,472) AF=5,221 (RF=1,989) AF=1,877 (RF=1,469) 289 10.1. Commodity Metaphors The MIGRANTS ARE A COMMODITY metaphor represents governments as countries and politics as business, whereas migrants and refugees are construed as objects of these business dealings. It has received relatively little attention in previous research despite its widespread use: for example, Arcimaviciene and Baglama (2018) found that OBJECT and COMMODITY metaphors accounted for a quarter of all the metaphors used in the 57 UK and US media reports they examined. Montagut and Moragas-Fernández (2020) noted the use of OBJECT metaphors in Spanish newspapers, whereas Petersson & Kainz (2017) observe that they are among the most common metaphors in Swedish and German news coverage of immigration. Taylor (2021b) has shown that this group of metaphors dominated in terms of continuity based on her investigation of a corpus of UK press spanning from 1900 to 2018. Similarly, this investigation also indicates a significant preponderance of the Spanish, British, and, to a lesser extent, Polish press to discuss migrants using terms associated with economic goods. The largest number of COMMODITY metaphors has been identified in the Spanish subcorpus, with 3,671 separate instantiations (RF=1,481) distributed over 17 different lexical items (see Table 15.2). The most frequently used item was traficante [‘trafficker’], which represents refugees as objects of illegal business dealings, followed by cuota [‘quota’] which, in its basic meaning, refers to obligations regarding quantities of goods for import or export (but also, in a more general sense, parts or portions of something, according to DRAE). Similarly, refugees are also frequently construed as being returned (devolver, devolución, exportar) to a country they came from (often without having recourse to legal advice or being able to petition for 290 asylum, as the frequent use of devolución al caliente [‘blanket return’] suggests) and as being distributed over a certain territory (repartir, distribuir, reparto, distribución). Table 15.2. COMMODITY Metaphors in the Spanish Subcorpus Metaphor Example(s) traficante [‘trafficker’] (AF=737; RF=297) (4.1) Es una travesía peligrosa conocida para traficantes que tienen su industria bien engrasada (…). [‘It’s a dangerous route, of which traffickers, who know their business well, are aware (…).’] (El Mundo, August 18, 2017) cuota [‘quota’] (AF=691; RF=279) (4.2) ¿Por qué la resistencia de España a acoger las cuotas de refugiados que se le proponían? [‘Why is Spain resisting accommodating the proposed cuotas of refugees?’] (ABC, October 26, 2015) devolver [‘return’] (AF=503; RF=203) (4.3) Todos los que se devolvieron a Turquía eran nacionales de otros países distintos a Siria. [‘Those who were returned to Turkey were nationals of countries other than Syria.’] (ABC, April 5, 2016) reparto [‘distribution’] (AF=457; RF=184) (4.4) Bruselas activó el reparto de refugiados a finales de 2015 como solución temporal a la crisis. [‘Brussels launched the distribution of refugees at the end of 2015 as a temporary solution to the crisis.’] (El País, June 14, 2017) devolución [‘return’] (AF=425; RF=171) (4.5) Además, Nieto volvió a defender las devoluciones en caliente. [‘Moreover, Nieto once again defended returns at the border.’] (El País, March 8, 2017) repartir [‘distribute’] (AF=281; RF=113) (4.6) La última propuesta para repartir la carga migratoria no prevé asignar cuotas obligatorias a los países. [‘The last proposal to distribute the migratory load does not include obligatory quotas.’] (La Vanguardia, June 6, 2018) cargar [‘load’] (AF=137; RF=55) (4.7) Un barco pesquero cargado con unos 700 inmigrantes y refugiados naufragó en el momento en que iban a ser rescatados. [‘A fishing boat loaded with 700 immigrants and refugees sank when they were about to be rescued.’] (ABC, August 6, 2015) distribuir [‘distribute’] (AF=123; RF=50) (4.8) Necesitamos un mecanismo claro para distribuir la carga. [‘We need a clear mechanism to distribute the load.’] (ABC, April 22, 2015) distribución [‘distribution’] (AF=90; RF=36) (4.9). El primer ministro eslovaco, Robert Fico, votó contra la distribución de asilados. [‘Slovakian Prime Minister Robert Fico voted against the distribution of asylum seekers.’] (El País, November 16, 2015) traficar [‘traffic’] (AF=89; RF=36) (4.10) Terminaremos con las mafias que trafican con seres humanos. Esos traficantes con la miseria humana son los principales culpables. [‘We will end the mafias which traffic with humans. The traffickers of human misery are to blame.’] (ABC, February 3, 2018) 291 Table 15.2. (Cont.) COMMODITY Metaphors in the Spanish Subcorpus In the UK subcorpus the COMMODITY metaphor is instantiated, above all, using the word deal in reference to the March 16 agreement between the European Union and Turkey regarding the control of migrations. The following two top items are references to human trafficking and, similarly to deal, also have hundreds of occurrences (See Table 15.3). The analysis has also detected a number of items which conceptualise refugees in terms of goods to be distributed using nouns such as distribution, quota, share, load, cargo, and batch; and verbs such as distribute, load, trade, import, and exchange. In one text, refugee children were described as “datestamped” in the same manner that a commodity’s shelf life may be determined by the Metaphor Example(s) carga [‘cargo’, ‘load’] )AF=86; RF=35) (4.11) Lo hacen con su lancha fueraborda, muy parecida en sus fundamentos a las que usan los traficantes de seres humanos para llevar su carga de migrantes, pero la suya es mucho más sólida y normalmente mucho menos sobrecargada. [‘They use an outboard engine, similarly to human traffickers carrying their loads of migrants, but their boat is much more solid and usually not so overloaded.’] (La Vanguardia, November 15, 2016) sobrecargar [‘overload’] (AF=34; RF=14) (4.12) Ambos consideran que el sistema de ayudas no debe sobrecargarse admitiendo migrantes que no dan y sólo «se llevan». [‘Both say that the aid system should not be overloaded by migrants who never give, only take.’] (El Mundo, October 16, 2017) neto [‘net’] (AF=11; RF=4) (4.13) (…) terminar con los flujos netos de una región a otra (…). [(…) end net flows from one region to another (…).’] (La Vanguardia, July 25, 2016) exportar [‘export’] (AF=4; RF=2) (4.15) Cuando Europa 'exportaba' a los refugiados. [‘When Europe was “exporting” refugees.’] (El Mundo, October 28, 2015) lote [‘batch’] (AF=2; RF=1) (4.16) “Oí cómo negociaban por un lote de ocho sudaneses, a 220 euros el esclavo”. [‘“I heard them negotiating over a batch of eight Sudanese, 220 euros a slave”.’] (La Vanguardia, March 25, 2018) comercio [‘trade’] (AF=2; RF=1) (4.17) (…) las fuerzas de seguridad tailandesas contra esta red de comercio humano. [‘(…) Thai security forces against the human trade network’] (El Mundo, May 12, 2015) importar [‘import’] (AF=1; RF=0) (4.18) "Hemos importado al país personas totalmente embrutecidas", ha afirmado, por su parte, el ministro del Interior de Berlín, Frank Henkel. [‘“We have imported to the country completely desentised people”, said German Minister of Interior Affairs, Frank Henkel.’] (El País, July 27, 2016) 292 manufacturer. Despite the use of hedging devices, this metaphorical expression might be particularly effective in contributing to the construal of refugees as dehumanised and homogenous items for trade because it appears to be less conventionalised than other COMMODITY metaphors, and is, therefore, more likely to attract the reader’s attention, and invoke its literal meaning (Deignan, 2005, p. 30). The presence of this and other innovative or non-conventional metaphorical expressions also points to the underlying conceptual metaphor’s productivity. In sum, 1,878 separate instantiations (RF=716) of the COMMODITY metaphor, spread over 16 different lexical items, have been detected in the UK subcorpus. Table 15.3. COMMODITY Metaphors in the UK Subcorpus Metaphor Example(s) deal (AF=578; RF=220) (4.19) This is the squalid reality for tens of thousands of migrants after leaders including ex-PM David Cameron signed a deal with Turkey to pull up the drawbridge. (Daily Mirror, December 12, 2018) trafficker (AF=336; RF=128) (4.20) They also mean human traffickers can operate with impunity amid the lawlessness, raking in [pounds] 670 million a year from their sordid trade. (Daily Mail, April 21, 2015) trafficking (AF=269; RF=102) (4.21) Smugglers make £1million a day trafficking migrants from Bangladesh across Med. (The Sun, January 1, 2017) quota (AF=225; RF=86) (4.22) The quota clash is holding up progress on other EU asylum reforms (…). (The Guardian, June 18, 2018) net (migration) (AF=115; RF=44) (4.23) Net migration has been too high and it needs to come down. (Daily Mail, October 7, 2015) share (AF= 113; RF=43) (4.24) The move would increase pressure on Britain and countries in Eastern Europe that have so far been reluctant to take part in schemes, such as quotas, to share the burden caused by the influx. (Daily Mail, September 2, 2015) traffic (AF=57; RF=22) (4.25) It is rare to meet a trafficked woman who has not been faced with this choice. (The Guardian, February 1, 2018) distribution (AF=46; RF=18) (4.26) Europe is facing "a challenge that would affect generations" and which requires the introduction of a fair system of distribution of migrants across the EU. (The Guardian, August 23, 2015) distribute (AF=41; RF=17) (4.27) Most countries would see an increase in asylum seekers if they were distributed proportionally. (The Guardian, June 26, 2017) 293 Table 15.3. (Cont.) COMMODITY Metaphors in the UK Subcorpus The lowest number of COMMODITY metaphors has been detected in the Polish subset of the corpus: 300 separate instantiations (RF=235) distributed over 11 different lexical items. The most common among them was, by far, kwota [‘quota’]: in fact, it had over ten times more occurrences than the second top item, dystrybucja [‘distribution’]. In general, this metaphor referred mostly to transporting refugees and was lexicalised using items such as dystrybucja, przeładować [‘overload’], eksport [‘export’], importować [‘import’], załadować [‘load’], eksportować [‘export’], and ładunek [‘load’]. As can be observed in Table 15.4, more than half of the lexical items expressing the COMMODITY metaphor in this subcorpus were used less than ten times per a million words. Metaphor Example(s) load (AF=24; RF=9) (4.28) As third load of 'child refugees' arrives from Calais as official figures reveal TWO THIRDS of those who had ages assessed were found to be adults. (The Sun, October 19, 2016) (4.29) Dozens of migrants are loaded on to trucks in Agadez, Niger. (Daily Mail, June 26, 2018) cargo (AF=23; RF=9) (4.30) (…) with smugglers reassuring their cargo that there is little risk of drowning since EU vessels will pick them up. (Daily Mail, September 15, 2016) trade (AF=23; RF=9) (4.31) (…) those claiming to be expeople smugglers or involved in the illegal migrant trade were lined up and selected for EU funding to launch businesses of their own choice. (Daily Mail, September 18, 2017) batch (AF=12; RF=5) (4.32) The home secretary, Amber Rudd, may have wrongly raised expectations about the first batch of Calais refugees when she said last week that children aged 12 and under would have been given priority. (The Guardian, October 19, 2016) import (AF=9; RF=3) (4.33) Not one preening luvvie questions the sanity of Germany importing more than a million refugees in just one year when we are confronted by Islamic terror. (The Sun, September 25, 2016) exchange (AF=6; RF=2) (4.34) How many people would be exchanged? (The Guardian, March 8, 2016) date stamped (AF=1; RF=0) (4.35) Practical as well as ethical considerations show that the idea cannot work; children are not "date stamped”. (The Guardian, October 19, 2016) 294 Table 15.4. COMMODITY Metaphors in the Polish Subcorpus Metaphor Example(s) kwota [‘quota’] (AF=233; RF=182) (4.36) (…) Platforma Obywatelska, ustami Grzegorza Schetyny, wyraziła sprzeciw przeciwko przyjmowaniu kwot uchodźców, jakie narzucać miałaby UE (…). [‘ (…) Grzegorz Schetyna, on behalf of Platforma Obywatelska, objected to accepting refugee quotas imposed by the EU (…).’] (Super Express, May 12, 2017) dystrybucja [‘distribution’] (AF=21; RF=16) (4.37) Każdy kraj Unii (…) będzie musiał przyjmować imigrantów; ba, będzie miał swoje "obowiązkowe kwoty" w ramach systemu "dystrybucji imigrantów". [‘Every EU member (…) will have to accommodate immigrants; better yet, each will have “obligatory quotas” based on the system of “immigrant distribution”.’] (Rzeczpospolita, May 13, 2015) przeładować [‘reload’, ‘overload'] (AF=13; RF=10) (4.38) Ale ponad 60 tys. udało się przeładować z chybotliwych łódek albo wyciągnąć z wody. [‘But more than 60 thousand could be reloaded from the wobbly boats or removed from the water.’] (Rzeczpospolita, November 23, 2016) (4.39) Dzięki temu przeładowane uchodźcami łodzie są zatrzymywane (…). [‘As a result, the boats overloaded with refugees are stopped (…).’] (Gazeta Wyborcza, July 10, 2015) handel [‘trade’] (AF=12; RF=9) (4.40) Śledztwo w tej sprawie wykazało jednak, że za śmierć pięciu osób odpowiadają dwaj Syryjczycy trudniący się handlem ludźmi. [‘The investigation showed that two Syrian human traffickers were responsible for the deaths of five people.’] (Fakt, March 4, 2016) handlarz [‘trafficker’] (AF=6; RF=5) (4.41) Czy może padły ofiarą handlarzy żywym towarem? [‘Were they victims of traffickers of human cargo?’] (Fakt, April 8, 2016) eksport [‘export’] (AF=6; RF=5) (4.42) W tym roku umowę o "eksporcie" azylantów podpisały Izrael i Rwanda. [‘This year, Israel and Rwanda signed a deal to “export” asylum seekers.’] (Gazeta Wyborcza, August 4, 2015) (migracja) netto [‘net (migration)’] (AF=3; RF=2) (4.43) Migracja netto do Wielkiej Brytanii (…) odnotowała największy spadek rok po głosowaniu. [‘Net migration to Great Britain (…) reached a record low a year after the vote.’] (Gazeta Wyborcza, December 19, 2018) importować [‘import’] (AF=2; RF=2) (4.44) Na szczęście są Kluby "Gazety Polskiej", które w sytuacji chwilowych braków na rynku lokalnym mogą zorganizować wyprawę do Budapesztu, by importować stamtąd, jeśli nawet nie samych uchodźców, to przynajmniej antyimigrancką histerię. [‘Luckily, there are the clubs of “Gazeta Polska” which, in case of temporary market shortages, can organise a trip do Budapest to import, if not refugees themselves, at least some anti-immigrant histeria.’] (Gazeta Wyborcza, March 17, 2016) załadować [‘load’] (AF=2; RF=2) (4.45) Policjanci najpierw załadowali koczujących na dworcu Keleti imigrantów do pociągów, obiecując im, że pojadą do Niemiec. [‘The police loaded the immigrants camping at the Keleti station onto trains and promised them they would go to Germany.’] (Gazeta Wyborcza, September 14, 2015) 295 Table 15.4. (Cont.) COMMODITY Metaphors in the Polish Subcorpus The widespread use of COMMODITY metaphors, as Arcimaviciene and Baglama argue (2018, pp. 5-6): heavily contributes to the creation of social reality based on the mythical narrative that migration is not related to people, their lives and fate, but is rather a process based on the exchange of commodified relations between countries or governments. In other words, it desentises the audiences to the human aspect of migrations. This leads to the suppression of positive emotions such as empathy or compassion in the recipients of such messages which, in turn, may decrease their opposition to securitising measures. Furthermore, by dehumanising migrants, COMMODITY metaphors legitimise the anti-immigrant approach and behaviour as morally acceptable. The COMMODITY metaphor also promotes a pragmatic vision of migrations represented in terms of profit versus loss. This causes the decisions taken with regard to migrants to be dictated by self-interest and market-like concerns of supply and demand. Based on this logic, migrants are acceptable when they contribute to the economic well-being of the European citizens (and the decisions on where to locate Metaphor Example(s) eksportować [‘export’] (AF=1; RF=1) (4.46) Według mediów i aktywistów Izrael już od 2015 r. "eksportuje" tam niechcianych przybyszy na podstawie podpisanych wcześniej specjalnych umów (Uganda temu zaprzecza). [‘According to the media and activists, Israel has been “exporting” unwanted migrants there since 2015, following agreements signed beforehand (Uganda denies this).’] (Gazeta Wyborcza, January 30, 2018) ładunek [‘load’] (AF=1; RF=1) (4.47) Imigrantów wiozą płotki - zwykle bezrobotni imigranci z Włoch, których przemytnicza mafia zatrudniła jako kierowców i wysłała z "ładunkiem" do Niemiec. [‘The trucks with immigrants are driven by small fish — usually jobless immigrants from Italy, hired by smuggling mafias to get the “load” to Germany.’] (Gazeta Wyborcza, August 28, 2015) 296 this resource most usefully are the latter’s prerogative, as the common use of words such as reparto/ distribution/ dystrybucja imply) and rejected when this representation is coupled with the topos of economic burden (as can be observed in, for instance, excerpts 4.12 and 4.24). 10.2. Water Metaphors In line with previous contributions which have shown a significant media tendency to liken migrants to liquids (e.g. Nguyen & McCallum, 2016; Petersson & Kainz, 2017; Soto-Almela & Alcaraz-Mármol, 2019; Akşak, 2020; Van Der Valk, 2003; Pickering, 2001), the second most frequent metaphor in the present corpus has water as its source domain . It was lexicalised using a wide range of items, including some less 45 entrenched metaphorical expressions, which points to its productivity and impact on migration discourses. With respect to the different subcorpora, 1,158 WATER metaphors have been detected in the Polish subset (RF=907). The most common WATER metaphor was realised using the word napływ [‘influx’], followed closely by fala [‘wave’]. Other most salient instantiations of this metaphor were verbs: napływać, napłynąć [‘flow’] and zalewać, zalać [‘flood’] (followed by its noun-form: zalew [‘flood’]). At the same time, the majority of the remaining WATER metaphors occurred less than ten times each in the analysed material, as can be observed in Table 15.5 below. Both MIGRANTS ARE WATER and COUNTRY IS A CONTAINER metaphors were considered 45 jointly here due to the conceptual link between them: the source domain of water “activates both disaster and container scenarios” (Charteris-Black, 2006, p. 569). 297 Table 15.5. WATER Metaphors in the Polish Subcorpus Metaphor Example(s) napływ [‘influx’] (AF=462; RF=362) (4.48) Od kilku lat zmagamy się z masowym napływem uchodźców... [‘We’ve been struggling with the mass influx of refugees for several years…’] (Rzeczpospolita, May 31, 2017). fala [‘wave’] (AF=428; RF=335) (4.49) Wszystko wskazuje więc na to, że Europę zaleje kolejna fala imigrantów [‘So it sure seems that Europe will be flooded by another wave of immigrants’] (Fakt, September 28, 2015). napływać, napłynąć [‘flow’](AF=117; RF=91) (4.50) (…) z tym krajem dzieje się coś złego, od kiedy zaczęli napływać tam imigranci ekonomiczni [‘(…) something bad has been happening to this country ever since economic immigrants started to inflow’] (Fakt, March 23, 2016). zalewać, zalać [‘flood’] (AF=76; RF=59) (4.51) Imigranci zalewają całą Europę [‘Immigrants are flooding all Europe’] (Super Express, August 6, 2016). zalew [‘flood’] (AF=33; RF=25) (4.52) Imigranci to zalew Europy śmieciem ludzkim [‘Immigrants are a flood of human waste’] (politician Janusz Korwin Mikke quoted in Fakt, May 27, 2016). strumień [‘stream’] (AF=13; RF=10) (4.53) Ciągle rośnie cena powstrzymania strumienia imigrantów płynących w stronę Grecji [‘The price of stopping the stream of immigrants flowing towards Greece is rising constantly’] (Rzeczpospolita, March 8, 2016). potok [‘stream’] (AF=9; RF=7) (4.54) Potok uchodźców wydaje się nie mieć końca [‘The stream of refugees seems to never end’] (Gazeta Wyborcza, November 16, 2015). wchłonąć, wchłaniać [‘absorb’] (AF=8; RF=6) (4.55) (…) państwo socjalne nie jest w stanie wchłonąć milionów Afrykanów, którzy chcą przybyć do Europy [‘(…) a welfare state cannot absorb the millions of Africans who want to come to Europe’] (a tweet quoted in Gazeta Wyborcza, October 6, 2018). rozlać [‘spill over’] (AF=3; RF=2) (4.56) Tymczasem za chwilę uchodźcy rozleją się po całym kontynencie [‘Soon, refugees will spill over the whole continent’] (Gazeta Wyborcza, August 12, 2015). absorbować [‘absorb’] (AF=2; RF=2) (4.57) Polska nie ma możliwości, aby absorbować dużą grupę emigrantów [‘Poland does not have the capacity to absorb a large group of emigrants’] (Witold Waszczykowski, quoted in Fakt, May 31, 2017). potop [‘deluge’] (AF=2; RF=2) (4.58) (…) straszyć Polaków muzułmańskim potopem [‘(…) frighten Poles with a Muslim deluge’] (Gazeta Wyborcza, June 14, 2017). lawina [‘avalanche’] (AF=2; RF=2) (4.59) Tyle że w sierpniu do UE ruszyła lawina imigrantów [‘But in August an avalanche of immigrants started’] (Rzeczpospolita, September 17, 2015). tsunami (AF=1; RF=1) (4.60) I w końcu przed trzema laty na Niemcy ruszyło istne tsunami uchodźców [‘Finally, three years ago, a real tsunami of refugees started heading for Germany’] (Gazeta Wyborcza, May 20, 2018). wzbierać [‘well up’] (AF=1; RF=1) (4.61) Fala będzie wzbierać [‘The wave will well up’] (Rzeczpospolita, June 2, 2017). przelewać [‘slop over’] (AF=1; RF=1) (4.62) Austriacy wychodzą z siebie, by pomóc przelewającym się przez kraj uchodźcom, ale obcym nie ufają [‘Austrians go over and beyond to help refugees slopping over the country but they don’t trust the foreigners’] (Gazeta Wyborcza, September 12, 2015). 304 (4.114) Niektórzy mówią, że fala migrantów jest zbyt duża, by ją zatrzymać. Ale ona jest zbyt duża, żeby jej nie zatrzymywać. [‘Some say the wave of migrants is too big to be stopped. But it is too big not to be stopped.’] (Gazeta Wyborcza, December 11, 2015) Furthermore, this association of water with threat is not limited solely to its movement but also to its use as a means of transportation. Specifically, the WATER metaphor often argues that refugees transport (or are used to do so by an external actor) some sort of danger to Europe, most frequently terrorism: (4.116) Trzeba być idiotą, żeby nie przewidzieć elementów terrorystycznych w całej tej fali [‘One has to be an idiot not to see a terrorist element in all this wave’] (Fakt, June 13, 2016). (4.117) (Europa) ni siquiera ha intentado filtrar a los tan peligrosos yihadistas camuflados dentro de la avalancha. [‘Europe hasn’t even tried to filter out the dangerous jihadists camouflaged as part of the avalanche.’] (ABC, December 22, 2016) Furthermore, the movement of water may be related to another metaphor: CHANGES ARE MOVEMENTS which, in turn, entails that lack of control over movement is lack of control over (social) change (Charteris-Black, 2006, p. 572). This metaphor might therefore appeal particularly to right-wing political forces which traditionally tend to resist social change (Charteris-Black, 2006, p. 579). However, this does not mean that dehumanising WATER metaphors are always chosen consciously and with a prejudicial intention. For example, their frequency in discourses surrounding refugees has sometimes been associated with the fact that refugees in many cases actually come across the sea (El Refaie, 2001) or the fact that on drone footage taken from high altitudes individuals are, in fact, undistinguishable and therefore look “like a dehumanised, steadily moving mass” (Törmä, 2017, p. 21). What is more, in the present study numerous liquid metaphors have been identified in texts which advocated for a more humanitarian attitude towards refugees (similar results have been reported by KhosraviNik [2009; 2014], 305 Nguyen and McCallum [2016], Salahshour [2016], and Taylor [2021b]). Nevertheless, dehumanising linguistic strategies, despite the potentially positive intentions lying behind their use (e.g. of conveying the urgency to act on behalf of the migrants), still reduce an individual to a homogenous part of a collective and erase their agency. As KhosraviNik (2009, p. 19) notes, this discursive tendency “may be best described as victimisation rather than positive representation”. What is more, the speaker’s intent may fail to correspond to the readers’ reception depending on how they were primed by previous discursive exposures (Taylor, 2021b, p. 4). 10.3. Animal Metaphors Apart from conceptualisations of refugees through inanimate natural phenomena, the analysis has also revealed the presence of metaphors which draw on the animal world, especially in the UK subcorpus. The ANIMAL metaphor has been discussed extensively in previous research (e.g. Mujagić & Berberović, 2019; Taylor, 2021b). Studies indicate that it is particularly effective in eliciting disgust reactions (Marshall & Shapiro, 2018). Therefore, in spite of the relatively limited lexicalisation of this metaphor in the corpus under the present investigation, its impact on the readers should not be underestimated. By far the most frequent ANIMAL metaphor used with respect to the displaced people employs the word jungle/ jungla/ dżungla to denote refugee settlements, most prominently Camp de la Lande near Calais, France. As can be observed in Table 15.8, this tendency is particularly salient in the UK subcorpus, which can be attributed to external factors: the Camp de la Lande was created to accommodate refugees trying to get to the UK from the continent via the Eurotunnel, therefore its 306 management was of primordial importance for the security of the UK border. In all three subcorpora the word jungle/ jungla/ dżungla emphasises the foreignness of the refugee camp (there are no jungles in Europe), but most importantly, the jungle stands for a place that is uncivilised, disorderly, and menacing (Törmä, 2017, p. 26). Therefore, as the relationship between the refugees and the camp where they live is metonymical in nature, the use of the word jungle implicates that its inhabitants are equally unfamiliar, foreign, savage, and dangerous as the physical space they occupy (Törmä, 2017, p. 26). Table 15.8. ANIMAL Metaphors in the Corpus Metaphor Example(s) UK Subcorpus (Calais) jungle (AF=1,168; RF= 445) (4.118) Bouchart refused, saying it puts locals "again under the threat of the creation of another Jungle". The original Jungle, home to up to 9,000 migrants, was levelled nine months ago. (The Sun, August 6, 2017) swarm (AF= 54; RF=20) (4.119) Migrants are using new tactics to ambush drivers, including stealing motorway signs and placing them on roads to force unsuspecting truckers to stop before swarming around the lorries. (Daily Mail, December 23, 2015) (4.120) David Cameron (…) has said that the government will leave no stone unturned as it responds to what he described as the " swarm of people" from Calais. (The Guardian, August 4, 2015) flock (AF=20; RF=8) (4.121) They are designed as places to return those rescued at sea but EU leaders admit they could become a 'pull factor', meaning asylum seekers would flock to them. (Daily Mail, February 11, 2017) hunt (AF=19; RF=7) (4.122) Last week neo-Nazis hunted 20 migrants through the streets of the eastern town of Bautzen shouting "F*** off from our Nazi district!”. (The Sun, September 20, 2016) cockroach (AF=11; RF=4) (4.123) While media and politicians perceive them as cockroaches or statistics, the crisis will only escalate. (The Guardian, June 1, 2015) stampede (AF=6; RF=2) (4.124) Mrs Merkel said she accepted some responsibility for the stampede of refugees across Europe last summer. (The Sun, September 1, 2016) horse (AF=4; RF=2) (4.125) ‘The British people want their government to be able to control numbers, not shut the gate when the horse has already bolted.' (Daily Mail, January 23, 2016) 307 Table 15.8. (Cont.) ANIMAL Metaphors in the Corpus Note, with respect to the word swarm, that it includes its uses both in the noun and the verb form. Metaphor Example(s) Polish Subcorpus dżungla (w Calais) [‘(Calais) jungle’] (AF=106; RF=83) (4.126) Miejscowi aktywiści nie mają złudzeń - "dżungla" się odrodzi. [‘Local activists have no illusions: the “jungle” will regenerate.’] (Gazeta Wyborcza, October 25, 2016) chmara [‘swarm’] (AF=3; RF=2) (4.127) “Chmara przeprawiająca się przez morze" – tak premier Wielkiej Brytanii, David Cameron nazwał zwiększający się napływ uchodźców z innych krajów. [‘“A swarm crossing the sea” — is how the United Kingdom’s prime minister David Cameron referred to the increasing influx of refugees from other countries.’] (Fakt, August 3, 2015) szarańcza [‘locust’] (AF=3; RF=2) (4.128) Jedni pokazują, że południowe Węgry po wizycie emigrantów wyglądają jak pole po przejściu szarańczy [‘Some point out that southern Hungary, after a visit from emigrants, looks like a field plagued by locust’] (Rzeczpospolita, September 12, 2015). Spanish Subcorpus jungla (de Calais) [‘(Calais) jungle’] (AF=250; RF=100) (4.129) Otra jungla se instala en París. [‘Another jungle is being set up in Paris.’] (El País, October 29, 2016) enjambre [‘swarm’] (AF=23; RF=9) (4.130) Cameron se ganó la condena unánime al referirse a los inmigrantes como una «plaga» o un « enjambre ». [‘Cameron was unanimously condemned for referring to the immigrants as “plague” or “swarm”.’] (El Mundo, September 4, 2015) caza [‘hunt’] (AF=12; RF=5) (4.131) La detective Corrigan también cree que la turba no salió a la caza del inmigrante. [‘Detective Corrigan does not believe the mob left to hunt the immigrant, either.’] (ABC, April 9, 2017) manada [‘pack’] (AF=4; RF=2) (4.132) Me entristecen como al que más los ataques que esos lobos solitarios o en pequeñas manadas lanzan contra personas que, por lo general, les han acogido con generosidad y altruismo. [‘It makes me sad to see these lone wolves or small packs attack people who, in general, welcomed them with generosity and altruism.’] (ABC, July 27, 2016) estampida [‘stampede’] (AF=3; RF=1) (4.133) La noche anterior, en la explanada del check point se produjo una estampida cuando algunos sirios empezaron a discutir con la policía. [‘The previous night on the esplanade of the check point there was a stampede when some Syrians started to argue with the police.’] (El Mundo, August 28, 2015) cazar [‘hunt’] (AF=2; RF=1) (4.134) A la mayoría (…) los caza la policía francesa y los devuelve en caliente. [‘The majority (…) are hunted by the French police and returned immediately.’] (El País, October 25, 2018) 308 As Table 15.8 above demonstrates, refugees are also — albeit to a much lesser extent — likened to animals by means of container words which normally designate groups of animals as well as verbs denoting the movement of animals or referring to actions of which animals are typically objects, such as hunting. It should, nevertheless, be noted that different distancing strategies are often employed with respect to the ANIMAL metaphors, especially in the Polish and the UK subcorpora. In the latter, for example, out of the 54 concordances of the word swarm, as many as 31 times (RF=12) the word is placed inside commas. Among the remaining 23 concordances, 11 (RF=4) include quotes, and two — a negation, as in: they are not a swarm, thus contesting the image of refugees as a swarm of insects. Similarly, cockroach is mostly used in reported speech, e.g.: (4.135) While media and politicians perceive them as cockroaches or statistics, the crisis will only escalate. (The Guardian, January 1, 2015) In the Spanish subcorpus, on the other hand, ANIMAL metaphors are rarely accompanied by distancing strategies or approached critically. For example, the noun caza [‘hunt’] was placed inside quotes five times (out of the twelve occurrences) but counter-discourses were not provided. A similar tendency has been observed for the words estampida [‘stampede’] and manada [‘pack’]. The main exception is the word enjambre [‘swarm’] which was mostly used to quote a David Cameron speech, as the example 4.131 shows (similarly as in the other two subcorpora). The MIGRANTS ARE ANIMALS metaphors dehumanise and pseudospeciate the displaced people by presenting them as different, separate, and distant, which, arguably, emotionalises and desensitises the audiences as well as legitimises prejudicial and securitisation actions. Even though the analysed newspapers sometimes partially reject this dehumanising metaphor, its presence in the corpus 309 does attest to the existence of this discursive trend within the wider discursive community. Furthermore, it legitimises it by reproducing explicitly racist language which might lead to confirming and strengthening some of the already existing racist attitudes and beliefs (Santa Ana, 1999; van Dijk, 2008, p. 55). 10.4. (Unwanted) Guests Metaphors The REFUGEES ARE (UNWANTED) GUESTS metaphor has rarely been discussed in previous research: according to Taylor (2021b, p. 12) this may be due to the fact that it is seen as less problematic than other common metaphors of migration. It builds on the metaphor COUNTRY IS HOME and its underlying basis are expectations about who rightly belongs in particular geographical spaces. It therefore conveys (racist) assumptions about group membership and strengthens the us versus them division in a particularly effective way by comparing the “insiders” to family members sharing a history, culture, language, needs, aspirations, etc., whereas the “outsiders” are, by default, foreign and distant. The former are also the ones who get to decide who should be allowed “in” and who should stay “outside”, especially considering the limited capacity of the “home”. As a consequence, the metaphor also implies that the “home” should be protected against external threats which, as Nguyen and McCallum stress (2016, p. 169), “has profound implications because security is a basic human need that significantly drives people’s decision-making in response to a perceived ‘security problem’”. In fact, in the corpus under this investigation, refugees are often represented as unwanted or uninvited guests (this idea is often lexicalised explicitly using items such as unwanted/ nieproszony) who, furthermore, abuse the hospitality 310 by overstaying their welcome and causing various troubles (see, for example, excerpts 4.163, 4.174, and 4.180 in Table 15.9). Table 15.9. GUEST Metaphors in the Corpus Metaphor Example(s) UK Subcorpus welcome (AF=700; RF=267) (4.161) Similarly, only a majority in Germany (69%) and Denmark (53%) feel their country has the economic means necessary to welcome refugees. (The Guardian, October 30, 2015) host (AF= 179; RF=68) (4.162) (…) the government-run temporary migrant holding centre (CETI), which is overflowing at twice its capacity hosting a mix of economic migrants and Syrian asylum seekers. (The Guardian, May 10, 2017) hospitality (AF=28; RF=11) (4.163) It is sickening to see how our country's hospitality has been abused. (Daily Mail, January 3, 2018) invite (AF=23; RF=9) (4.164) Merkel invited the Third World to move to Germany. (The Sun, November 25, 2017) overstay (AF=23; RF=9) (4.165) The figure for illegal migrants - mainly people overstaying their visas - adds half as much again to official net migration. (Daily Mail, June 16, 2017) guest (AF=12; RF=5) (4.166) The main topic of conversation is the "guests". In October the town had been informed by the Finnish Red Cross that the local hotel, the Strandbo, closed for the off-season, would become a temporary home for 100 refugees. (The Guardian, January 23, 2016) invitation (AF=4; RF=2) (4.167) From Berlin to Brussels it is clear: there cannot be an openended invitation to the miserable millions of Europe's southern and eastern periphery. (The Guardian, October 30, 2017) uninvited (AF=3; RF=1) (4.168) But Europe has continued to try and push the thousands of uninvited migrants who try to reach European shores further and further away from the continent. (The Guardian, June 5, 2018) Polish Subcorpus zaprosić, zapraszać [‘invite’] (AF=62; RF=49) (4.169) Dlaczego Angela Merkel zaprosiła uchodźców? [‘Why did Angela Merkel invite refugees?’] (Fakt, May 15, 2016) gość [‘guest’] (AF=30; RF=23) (4.170) Lęk przed gościem to niepokój, że przemieni nasz dom, życie. [‘The fear of one’s guest is the anxiety that they might transform our home, our life.’] (Gazeta Wyborcza, April 23, 2016) gościć [‘host’] (AF=29; RF=22) (4.171) W ciągu ostatnich siedmiu lat koszt goszczenia Syryjczyków przekroczył 2 mld dol. rocznie. [‘In the last seven years the cost of hosting Syrians reached more than two billion dollars yearly.’] (Gazeta Wyborcza, January 31, 2018) 311 Table 15.9. (Cont.) GUEST Metaphors in the Corpus Metaphor Example(s) gospodarz [‘host’] (AF=19; RF=15) (4.172) O tym, jak zachowują się imigranci wobec gospodarzy, o ich niechęci do integracji, o kosztach, o naruszeniach praw człowieka w imigranckich gettach napisano setki książek i tysiące artykułów, z których nie wyciągnięto żadnych wniosków. [‘Hundreds of books and thousands of articles have been written on immigrants’ behaviour towards their hosts, their unwillingness to integrate, the cost, and on the human rights’ breaches in immigrant ghettos, but no lessons were drawn.’] (Rzeczpospolita, January 16, 2016) zaproszenie [‘invitation’] (AF=18; RF=14) (4.173) Czy zatem zaproszenie imigrantów nie może wywołać w przyszłości takich kłopotów jak we Francji czy Niemczech? [‘Isn’t it possible, then, that inviting immigrants might cause the same problems it did in France and Germany?’] (Rzeczpospolita, June 1, 2016) nieproszony [‘uninvited’] (AF=4; RF=3) (4.174) Trzeba być idiotą, aby okazać nieproszonym gościom tyle uwagi. Co gorsza, końca nie widać. [‘One must be na idiot to pay so much attention to unwanted guests. What’s worse, there is no end in sight.’] (Fakt, June 13, 2016) Spanish Subcorpus hospitalidad [‘hospitality’] (AF=28; RF=11) (4.175) (…) un millón de refugiados se había beneficiado de la hospitalidad alemana. [‘(…) a million of refugees benefitted from Germany’s hospitality.’] (El Mundo, September 26, 2017) huésped [‘guest’] (AF=24; RF=10) (4.176) «Son huéspedes a nuestra costa y encima molestan». [‘“They are guests on our coasts and, what is more, they are troublesome.”’] (ABC, August 26, 2015) invitación [‘invitation’] (AF=16; RF=6) (4.177) "Esa política -dice Kurzes interpretada como una invitación por parte de potenciales emigrantes africanos (…)”. [‘This policy, says Kurz, is interpreted as an invitation by potential African emigrants (…).’] (La Vanguardia, October 10, 2016) anfitrión [‘host’] (AF=9; RF=4) (4.178) Los inmigrantes se adaptaban a la sociedad anfitriona. [‘Immigrants complied with the host society.’] (ABC, March 5, 2018) invitar [‘invite’] (AF=7; RF=3) (4.179) (…) cuando Merkel hizo su famoso llamamiento público en el que invitaba a cientos de miles de refugiados a Alemania, ¿cuál era su legitimación democrática? [‘(…) when Merkel made her famous appeal to the public where she invited hundreds of thousands refugees to Germany, what was her democratic legitimacy?’] (El Mundo, May 10, 2016) invitado [‘guest’] (AF=5; RF=2) (4.180) Al principio éramos más tolerantes, pero ahora sabemos que nuestros invitados van a quedarse para siempre y nuestra hospitalidad no da más de sí. [‘At the beginning we were more tolerant but now we know that our guests are going to stay forever and our hospitality is limited.’] (La Vanguardia, June 4, 2017) hospedar [‘host’] (AF=5; RF=2) (4.181) «Nuestras puertas no están cerradas, pero por el momento no hay necesidad de hospedar en nuestro territorio a estas personas», aseguró ayer Suleyman Tapsiz. [‘“Our doors are not closed but right now there is no need to host those people in our territory”, said Suleyman Tapsiz yesterday.’] (ABC, February 7, 2016) 312 With respect to the problems that the “guests” are represented as causing, the purported cultural differences feature prominently in the analysed texts (see also Section 9.1). This is based on the topos that a host has the right to impose rules and norms of behaviour on their guests and the latter are obliged to comply. For example, excerpt 4.178, using the nostalgia fallacy, asserts that in the past immigrants knew to comply with the host society’s norms, contrary to immigrants now. Similarly, excerpt 4.170 implies that a guest exercising an impact on the life of her host is undesirable and elicits fear; instead, it suggests that the acceptable, non-threatening situation would require the guest to adapt fully to the host’s lifestyle. Other problems the presence of guests engenders are related to the costs of hosting them (e.g. excerpt 4.171) as well as their potentially harmful intentions as is implied, for example, in fragment 4.173, where refugees are associated with terrorism and violence. 10.5. War Metaphors Finally, this highly pejorative metaphor takes war as its source domain (see Table 15.10). Specifically, in this type of metaphor refugees are likened to invaders and their movement is represented as an attack against the European borders. Importantly, as Pickering (2001, p. 174) notes, war metaphors tend to construe a black-and-white world, where only one side has the moral high ground, whereas the other is demonised. The boundaries between them are also drawn clearly: in the case of migrations, they correspond to the actual national and geographical borders (Pickering, 2001, p. 174). 313 Table 15.10. WAR Metaphors in the Corpus Metaphor Example(s) UK Subcorpus invasion (AF= 24; RF=9) (4.136) A vast influx of aliens could cause patriotic feelings of foreign invasion, with unforeseeable results. (Daily Mail, September 25, 2015) horde (AF=19; RF=7) (4.137) In the besieged port there is just one tiny window, framing harassed officials, at which the desperate hordes must queue (Daily Mirror, September 6, 2015). Trojan horse (AF=11; RF=4) (4.138) In yesterday's Mail, MPs warned that the Akrotiri base could be used as a 'Trojan horse' by migrants seeking a back door into Britain. (Daily Mail, October 24, 2015) invade (AF=8; RF=3) (4.138) It wasn't hard to realise why migrants are invading the Channel Tunnel terminal on a nightly basis. (Daily Mirror, July 30, 2015) combat (AF=8; RF=3) (4.139) Lorry drivers in Calais tooling up with tasers and CS gas to combat desperate migrants Drivers are risking prosecution if caught with the weapons in England - but feel it is necessary with refugees becoming increasingly desperate to cross the border. (Daily Mirror, November 1, 2015) battle (AF=8; RF=3) (4.140) Tory MPs accused her of "raising the white flag" on the battle against illegal migrants. (The Sun, May 17, 2018) fight (AF=5; RF=2) (4.141) (…) leaving European governments without viable Libyan partners in their fight against migration. (The Guardian, November 28, 2016) defender (AF=3; RF=1) (4.142) On the one hand, demagogues such as Hungary's Viktor Orbán have positioned themselves as defenders of a Christian European civilisation, enacting anti-migrant policies to protect Europe from being overrun by Muslim hordes. (The Guardian, June 5, 2018) Polish Subcorpus bronić [‘defend’] (AF=55; RF=43) (4.143) Jeśli chcemy bronić Europy przed napływem migrantów, musimy też bronić Bliskiego Wschodu. [‘If we want to defend Europe from the influx of migrants, we have to defend the Middle East, too.’] (Gazeta Wyborcza, September 18, 2015) inwazja [‘invasion’] (AF=48; RF=37) (4.144) (…) a zwłaszcza prezydentowi Zemanowi, który o uchodźcach mówi jako o "zorganizowanej inwazji". [‘(…) especially president Zeman who refers to refugees as an “organised invasion”.’] (Gazeta Wyborcza, February 6, 2016) horda [‘horde’] (AF=21; RF=16) (4.145) “Patrząc na te hordy atakujące policję w Budapeszcie, na granicy Austrii, w Grecji, naprawdę trudno jest się tam dopatrzyć prawdziwych uchodźców”. [‘“Looking at the hordes attacking the police in Budapest, at the Austrian border, in Czech Republic, it is really difficult to see any real refugees.”’] (Gazeta Wyborcza, September 19, 2015) najazd [‘invasion’] (AF=10; RF=7) (4.146) W tym roku Berlin spodziewa się najazdu 800 tys. imigrantów starających się o azyl. [‘This year Berlin is expecting an invasion of 800 thousand immigrants seeking asylum.’] (Rzeczpospolita, August 25, 2015) 320 11. Summary of the Results This Section sums up the results of the three analytical procedures employed as part of this study: the quantitative analysis of wordlists and collocations, the analysis of argumentation, and the identification and examination of metaphors employed to represent migrants and refugees. The following Chapter discusses these results and concludes the study. The part of the study which employed corpus linguistics methods had two stages: the analysis of word sketches of the main topical words (refugee and migrant in the UK subcorpus, uchodźca and imigrant in the Polish subcorpus, and refugiado and inmigrante in the Spanish subcorpus), and the examination of the most frequent nouns, verbs, and adjectives in the corpus. With respect to the former, it showed both differences and similarities in how the displaced people were portrayed in the three countries of the analysis, and depending on the label attached to them (refugee versus migrant). The most common collocates concerned their geographical identities, numbers, movement, accommodation, reasons for migrating, legal status, religious beliefs, family and group relations, ages, and gender. Some of them pointed to discussions of the varying attitudes and reactions towards the displaced people in Europe, including both humanitarian aid and hostility. With respect to the latter, the analysis of wordlists has demonstrated a strong focus on Europeans and places in Europe in the coverage of the „refugee crisis”, on the topic of safety and security and, to a lesser extent, on political and administrative issues, movement, humanitarian aid, religion, and the economy. Here, too, differences between the three subcorpora could be observed. 321 The analysis of argumentation has shown that arguments against accommodating refugees were significantly more numerous and developed much more extensively (and in a larger number of texts) than the arguments supporting the inverse conclusion. The former stance was also expressed explicitly more frequently or suggested by means of evaluation. Twenty-two different arguments were employed to support it. The majority construe refugees as embodiments of various forms of threat, while others question refugees’ identities and motivations, mention the financial aspect of accommodating refugees, migration figures, and popular attitudes, and present securitising solutions as beneficial to the displaced people. Many of them are fallacious. The seventeen arguments for helping refugees, in turn, could be grouped into three main categories: those that construe humanitarian aid in terms of an ethical or legal obligation, those that present refugees positively and/or victimise them, and those that refer to Europe’s capability to help and the benefits of immigration. Finally, metaphor analysis has revealed five main frames of metaphorical representation of refugees and immigrants. The most common of them employ commodities and liquids as their source domains. They are followed by animal, (unwanted) guest, and invader metaphors. In other words, the most frequent metaphors employed with respect to the displaced people dehumanise them and present them as threat. 322 12. Discussion The present Chapter concludes the investigation into British, Polish, and Spanish press representations of refugees and migrants during the 2015-2018 “refugee crisis”. It is organised in the following manner. First, a few remarks are made regarding the framing of the migrations per se (Section 12.1). While the study focused on the construal of its main actors, the examination of the corpus nevertheless brought important insights into how their movement itself was problematised. This has, in turn, revealed important ideological assumptions which contributed to the construal of refugees and migrants, and which were likely to have impacted on their inclusion into European societies. Section 12.2 comments on the naming strategies employed with respect to the people on the move, whereas the main trends in their construal are discussed in Section 12.3. Section 12.4 focuses on the cross-linguistic comparison by demonstrating how the press in the three different European countries converged and diverged in their representations. Section 12.5 discusses the potential benefits of employing methodological triangulation to the present research by comparing the results of the two strands of the investigation — the analysis of argumentation and the quantitative examination of the corpus. Finally, Section 12.6 concludes this Chapter. 12.1. Migrations as a Crisis The 2015-2018 migrations from Africa and Asia to Europe were predominantly labelled, using a conventionalised metaphor, a crisis. The word crisis, in its basic meaning, refers to the turning point of a disease, signalling either recovery or death. However, from the early 17th century it came to be used to denote “a time of great 323 danger, difficulty or doubt when problems must be solved or important decisions must be made" (OALD). Therefore, its use with respect to migrations contributes to construing them as a potential threat with grave and far-reaching consequences, requiring urgent attention. Importantly, as the analysis has demonstrated, it was the European communities, rather than the displaced people, who came to be represented as the principal victims of this crisis. First, the source of the crisis was located in Europe or at the Mediterranean Sea even though the precarious situation of many refugees, the result of, for instance, the war in Syria, predated their migrations to Europe. Furthermore, only a small minority actually undertook the sea journey, whereas 85% remained in neighbouring countries (UNHCR, 2019). Second, scarce attention was paid to the factors which forced migrants to risk the extremely perilous journey, especially in the UK and the Polish samples. And even when some context was provided, it invariably concealed historical or present-day role of Europe in shaping the geopolitical and economic realities of the refugeeproducing regions. As a result, the Middle East and Africa could be seen as “a space that breeds war, violence, persecution, and poverty” (Kirtsoglou & Tsimouris [2016, p. 8]) whilst Europe was portrayed as developed, superior, and suffering the consequences of destabilisation in the global South. Third, the solutions were largely limited to the European context: above all, border securitisation measures prevailed. And even when humanitarian aid was considered and the label “crisis” was used to draw attention to the deaths at sea, the fact that the movement of the displaced people was conceptualised as a one-off and short-lived crisis event effectively sidelined “every day systemic issues such as poverty, vulnerability, and environmental fragility” (McAdam, 2014, p. 3, emphasis in 324 the original) which might have led to deprioritising more comprehensive and longerlasting solutions. As McAdam (2014, p. 2) stresses, “hazards do not become disasters (or crises) for individuals or communities unless they face a certain degree of socioeconomic vulnerability”. The eradication of the adversities these societies meet in their everyday lives would therefore help mitigate the effects of critical emergency situations. Finally, the findings of the present study can be seen in line with previous contributions which have shown that migrations tend not to be problematised in public discourses as long as migrants are not headed towards Europe (or another Western country) in spite of the fact that these events might possibly be described more adequately as crises. Two examples are provided by Sager (2018, p. 591): The Dadaab refugee camp is rarely discussed as a migration crisis, despite ongoing violence in Somalia that prevents voluntary repatriation, violence in the camp itself, and the limited prospects for resettlement for most of its residents. Finally, though the cancelation of the Deferred Action for Childhood Arrivals in the United States is a crisis for the 800,000 Dreamers, as well as for their families and communities, the media has not chosen to represent it in these terms. A similar point is made by De Genova and colleagues (2016, p. 13) who also stress the fact that camps on the borders of conflict zones in Turkey, the Middle East, and Africa have much bigger populations than any refugee camp located on the European continent. Similarly, Maddaloni and Moffa (2018, p. 604) discuss naming strategies of migrations from wealthier countries to Southern Europe (e.g. “lifestyle migration”) versus movements in the opposite direction. 325 Referring to the recent migrations as a crisis thus serves important ideologically-motivated functions. It helps maintain the hierarchical relations between the global North and South while obscuring Europe’s agenda. In fact, some scholars argue that crisis narratives are a means by which institutions and governments can “claim rights to stewardship over land and resources they do not own. By generating and appealing to crisis narratives, technical experts and managers assert rights as “stakeholders” in the land and resources they say are under crisis” (Roe, 1995, p. 1066). It also allows Europe to evade its humanitarian responsibilities, especially long-term, by presenting migrations as a disruptive anomaly rather than a logical and expected outcome of the given geopolitical, economic, and environmental circumstances. Furthermore, its alarmism may trump considered and constructive responses grounded in empirical research. As Dempster and Hargrave (2017, p. 20) emphasise, “the language of ‘crisis’ used around recent movements of refugees and migrants into Europe may have projected a sense of vast scale and unmanageability”. Therefore, it plays into the narrative which construes refugees as a threat. It also provides the governments with a useful distraction — or even scapegoat — from some of the more challenging tasks such as, for instance, dealing with terrorism. Finally, it aligns with — and, arguably, fans the flames of — far-right movements thus serving important political goals. 326 12.2. Hierarchisation and Conflation Concerning the labeling of the displaced people, it has been observed that the analysed press tended to employ one of two terms: refugee or migrant (UK); uchodźca [‘refugee’] or imigrant [‘immigrant’] (Poland); and refugiado [‘refugee’] or inmigrante [‘immigrant’] (Spain). However, while the newspapers in Poland and Spain had a clear preference for the terms uchodźca and refugiado (respectively), it is interesting to note that the British news outlets used both terms to the same extent. Therefore, the study set out, first, to investigate whether the words in these pairs were used differently or interchangeably regardless of their dictionary definitions. It has revealed a much more complex linguistic reality: while the two terms had, in fact, differing roles and functions, suggesting a categorisation of groups of migrants with attitudinal implications, they were also conflated using a variety of linguistic strategies. Specifically, the analysis has demonstrated that the word imigrant/ inmigrante/ migrant was the more likely choice when a given utterance/ text aimed at obscuring or delegitimising the rights of the people of concern, whereas refugees were represented as more deserving of humanitarian aid. This was reflected, above all, in the frequent references to immigrants as illegal and economic as both types of construal are considered to be strong predictors of negative public attitudes (e.g. Blinder et al., 2011; Goodwin et al., 2017; Meidert & Rapp, 2019 ). Apart from these 46 two modifiers, which were (among) the top collocates of imigrant/ inmigrante/ migrant, a range of other strategies were employed which presented them as lacking authorisation to remain in Europe (for example, they were less frequently construed as applying for or receiving asylum) as well as motivated mostly by financial But see Lawlor and Tolley [2017] who show that in the Canadian context, the inverse may be true. 46 327 concerns. At the same time, refugees were referred to as political and war (in the Polish and Spanish samples but not in the UK sample) . Similarly, immigrants were 47 more often portrayed as Muslim, which is another strong predictor of negative public attitudes (Heath & Richards, 2016; Wike et al., 2016), especially considering the very strong negative discourse prosody of Islam in the focus corpus (similar results were reported, among others, by Baker et al. [2013] and Clarke et al. [2021] who examined representations of Islam in the UK media). Finally, immigrants were represented less frequently as children and, therefore, less likely to be associated with vulnerability and the need for protection. Both refugees and migrants were often discussed in terms of their geographical identities, which has been shown to influence the tone of reporting on displaced people (Lawlor & Tolley, 2017). This may, at least partially, be interpreted as a discursive deracialisation strategy where nationality serves as a proxy for race, therefore protecting the speaker from potential accusations of racism (Goodman & Burke, 2010, p. 112). However, there was a significant divergence between the purported nationalities of people labeled refugees versus immigrants/ migrants. Specifically, refugees tended to be represented as coming from countries which have for a long time been associated with military conflict and political destabilisation such as Syria, above all, but also Afghanistan and Iraq. Immigrants, on the other hand, were predominantly described as African (and Subsaharan in the Spanish sample) despite the fact that Africans constituted a minority of migrants travelling to Europe during that period. The vagueness of this denomination , the low level of 48 Which may also be problematic: as Durand-Delacre (2022, p. 3399) writes, “social, economic, 47 political, cultural and environmental factors all play a role in decisions to move and (…) any attempt to isolate one over the others necessarily misses a crucial part of the picture”. Berry et al. (2015, p. 7) also note that the newspaper coverage of the “refugee crisis” tends to rely on 48 “vague geographical descriptions”, especially in Spain and the UK. 328 knowledge of the European publics on the socio-political situation in specific African countries as well as the stereotypisation of Africa and its inhabitants (e.g. Keim & Somerville, 2018) thus helped to delegitimise the claims of “African immigrants” for asylum. Moreover, the analysis has demonstrated a mismatch between the amount of attention refugees of particular geographical identities received and their actual numbers. In particular, the newspapers in Poland, Spain, and the UK were preoccupied with migrants who crossed the borders using illegal means rather than those who were most numerous. In other words, the numbers of people crossing the European borders with proper documentation, while significantly higher, were deemed to be much less newsworthy than the apparent urgency to stop those who risked their lives by undertaking “illegal” sea journeys (similar results are reported by Van Der Valk [2003, p. 323]). The study has also shown that the “deserving” group was often represented as being less numerous than the “undeserving" even though research indicates that nine-in-ten in refugee camps in Europe during the “refugee crisis” were forced to migrate due to violent conflict, persecution, and human rights abuse (Crawley & Skleparis, 2017). This implied a recognition of the right to seek asylum, at the same time limiting it to just a small proportion of the displaced people. Furthermore, the analysed press sometimes explicitly discussed the possibility of differentiating between the two groups, including in texts with a purportedly pro-immigration agenda. 329 This — typically two-step — hierarchisation has already been observed in previous analyses of immigration-related discourses (e.g. Goodman & Speer, 2007; Taylor, 2014; Berry et al., 2015; Efe, 2017; Goodman et al., 2017). For instance, Taylor (2014) has demonstrated that the word refugee was used in the UK press with respect to displaced people elsewhere in the world whereas for refugees in the country alternative, less legitimating terms were preferred. Similarly, Howden (2016) reports that Hungarian journalists were instructed to avoid using the word refugee altogether during the recent “refugee crisis”, but instead to adopt a “foreign-sounding imposition from Latin”: migran [‘migrant’]. Goodman and colleagues (2017), noted that the use of the word refugee in the UK press increased immediately after the publication of the photo of Alan Kurdi, thus stressing the vulnerability and the need for protection of the people involved. However, it was soon replaced with the term migrant. Finally, based on a study of 1,500 images of the “refugee crisis”, Maneri (2021) noticed that the label changed alongside with skin colour and gender of the portrayed: “refugees” tended to be light-skinned and female, whereas large groups of black men were consistently described as “migrants” or “clandestini”. Moreover, one can also reflect on the naming choices that were not made, which can also inform the interpretation of findings. While the present study has shown that the displaced people were discursively grouped based on the documentation they possessed or not, the country of origin, or the reasons and aims for migrating, etc., other types of categorisations could have been used — but were not. For example, following Goodman and Speer (2007, p. 180), it can be argued that asylum-seekers could equally accurately be categorised “in terms of those who have fled a country in which the British army is involved and those who have not, or in terms of those who have come from ex-British colonies and those who have not”. In 336 European men (with the exception of elite actors; see below) being significantly backgrounded (a similar observation has been made by Gray and Franck, 2019). Finally, all three subcorpora converged in the portrayal of women, above all, with respect to family relations and care for children, whereas men were described predominantly as young and single, therefore unburdened by the responsibilities of family life (similar results were reported for an analysis of the words girl and boy in a general-language corpus [Stefanowitsch, 2020, p. 258]). These findings could be viewed in line with research on the representations of Muslim women in the Western media (e.g. Al-Hejin, 2015; Karimullah, 2020), which has noted than they were often represented as passive, submissive, and under threat. This shows that in discourses within the cultural and religious environment of Islam Muslim women are presented as victims; however, when Muslims enter in contact with the Western world, the Europeans’ well-being becomes the main focus, as this analysis has demonstrated. In both contexts, however, male violence remains the constant. It may be argued to serve important, albeit different, functions in both types of discourses. While in the context of majority Muslim countries, the aim may be to rationalise Western intervention or military aggression (Karimullah, 2020; Rettberg & Gajjala, 2016, p. 179); when migrations to Europe are concerned, the good cause of fighting gender-based violence serves as justification for deterrence measures even though they are ultimately applied regardless of gender. In both cases, however, the media utilise feminism — understood as the adherence to Western norms rather than the right of free choice as the problematisation of the veil, observed both in this study and in previous contributions (e.g. Baker et al., 2013; AlHejin, 2015; Karimullah, 2020; Clarke et al., 2021; Freedman et al., 2022), has 337 demonstrated — to safeguard Europe’s right to secure the borders . Furthermore, 52 with respect to migrants who manage to get to Europe, this gender-based distinction between victimised and threatening refugees has potentially far-reaching consequences: as Busetta et al.’s research (2021) has shown, male refugees in Europe are much more vulnerable than women: for instance, they are twice as likely to suffer from poor health and have a much lesser chance of obtaining legal right to stay (in Italy where the study was conducted). This investigation has also pointed to the trend of representing the displaced people as a burden on the European economies, even though it was less salient than previous contributions have suggested (e.g. Greussing & Boomgaarden, 2017; Holzberg et al., 2018). This is another strong predictor of exclusionary attitudes (Hercowitz-Amir & Raijman, 2020), especially in high-income democracies (Bansak et al., 2016). Furthermore, it has been shown to exert a strong impact on how asylum policies are shaped (Mayblin, 2019). This type of construal is based on the assumption that the extent to which a person deserves to be given shelter should be measured in terms of their potential to contribute to the host country’s economic development. Meanwhile, the moral and legal obligation to provide refuge are backgrounded because the question of usefulness/ profitably versus financial burden takes precedence. What allowed for this discourse to emerge during the “refugee crisis” was, above all, the media’s tendency to represent the displaced people as “economic immigrants”; but it could also be observed, for instance, in the topos of “asylum-shopping” — in spite of the fact that research does not provide much evidence to support the claim that state aid acts as a significant pull factor (Mayblin, In a similar fashion, Ging (2017), for instance, notices how the threat of sexual aggression at the 52 hands of Muslims is evoked to discredit feminist movements in manosphere discourses. Similar results are also reported by Törnberg & Törnberg (2016). 338 2019). It can also be considered in line with the COMMODITY metaphor which also promotes a market-driven, pragmatic vision of migrations by conceiving of refugees in terms of economic assets or risks. Apart from the three main types of threat, which were also discussed in the literature (the threat to security, safety, and the economy), the analysis has also revealed a tendency to construe refugees as causing conflict and disruption in national and international politics, including the rise of populism. This is another one of the many instantiations of the paradoxical representation of refugees as both agentless and particularly potent: as, in fact, possessing similar power as the main political actors of impacting on diplomatic relations, policy, and social governance. This, effectively, switched the blame for decisions made at a high elite level to people most affected by their consequences (whereas the European citizens were construed as the “true” victims). The construal of refugees as responsible for the rise of populism is particularly striking as populism is largely based on creating and reinforcing the binary us-versus-them division thus, in fact, capitalising on the “uncontrolled” movement of the displaced people. Finally, refugees were represented as source of societal conflict and as disrupting the everyday lives of European citizens. While metaphorical representations of refugees served a range of different discursive functions, the threat theme was one of their key elements. The analysis has revealed five dominant metaphors employed with respect to the displaced people. The most frequent were dehumanising metaphors which likened them to objects, liquids, and animals, albeit there was some cross-linguistic variation (see also Section 12.3). These metaphors assume that the displaced people have no volition and no agency and, therefore, desentise the audiences to the human aspect of migrations, 339 which leads to the suppression of positive emotions such as empathy or compassion. Furthermore, by dehumanising migrants, they legitimise anti-immigrant and securitisation approaches and behaviours as morally acceptable. Furthermore, and more specifically with respect to the particular dehumanising metaphors, the REFUGEES ARE A COMMODITY metaphor construes migrations in terms of loss and benefit or supply and demand for “host” countries. These market-like concerns are therefore suggested as being crucial in the decision regarding the “return” or “distribution” of migrants — conceived of as economic goods — across Europe. The REFUGEES ARE LIQUID metaphor, on the other hand, carries negative discourse prosodies of large numbers, danger, and lack of control. However, it is the ANIMAL metaphor that is arguably the most hostile and racist among the identified dehumanising metaphors. In consequence, while its use was limited, compared to the two most frequent dehumanising metaphors, its effect on the receivers of media messages should not be underestimated. It presents the displaced people as different, separate, and distant; as, in fact, outside the human realm. Even if distancing strategies were employed, as was often the case in the analysed material, it legitimises racist language which might lead to confirming and strengthening some of the already existing racist attitudes and beliefs (Santa Ana, 1999; van Dijk, 2008, p. 55). It can also be considered as particularly effective in eliciting negative emotions, such as disgust, based on empirical findings (Marshall & Shapiro, 2018). Furthermore, it is unlikely that it may be used unconsciously, contrary to the largely conventionalised WATER and COMMODITY metaphors, as attested by the frequent use of distancing devices. 340 The metaphors which did not dehumanise refugees but, instead, utilised the myth of moral authority (Arcimaviciene & Baglama, 2018), and presented them as dangerous invaders and unwanted guests, were less frequent. The MIGRANTS ARE INVADERS metaphor contributed to construing the displaced people as dangerous and ill-intentioned. Furthermore, because it draws on several elements from the target domain — refugees, like soldiers in an army, were numerous and mostly male; like armies, they moved towards their goal — it might be particularly effective in providing “more concrete visual imagery to enhance perceptions of threat about immigrants” (Marlow, 2015, p. 269). The GUEST metaphor, albeit more subtle, also bestows on immigrants a lower status, deprives them of rights, and frequently presents them as a nuisance or even threat to the “host” society. Because it builds on the COUNTRY IS HOME metaphor, it is particularly effective in conveying ideas about the “moving power” (Allen et al., 2017, p. 222) and the right to belong in a particular place. Specifically, it is based upon a distinction between insiders (family members and friends) and those who can only be admitted conditionally (guests) or who should be prevented from entering (unwanted guests). In other words, it implies that the “home” should be protected against external threats which, as Nguyen and McCallum stress (2016, p. 169), “has profound implications because security is a basic human need that significantly drives people’s decision-making in response to a perceived ‘security problem’”. The figure of a threatening refugee was also tightly linked to their representations as deceitful. Specifically, the displaced people were often construed as hiding their malevolent intentions which were purported to really have motivated them to migrate. However, this type of construal encompassed other aspects of refugee identity as well. A range of specific discursive strategies were used by the 341 press to construe refugees as deceiving with regard to their ages, identities, motivations, and intentions. Specifically, refugees were often reported to pose as children; as lying about their countries of origin or presenting false documentation; as not being truthful about their religious beliefs (for instance, converting to Christianity to boost their chances of being allowed to remain in Europe rather than out of true belief); and as pretending to have other reasons to flee their home countries whereas their true objective was presented as financial gain or even terrorist activity. In fact, it was one of the strongest trends in the corpus under investigation which, however, has not been examined systematically up to date (with the exception of Zawadzka-Paluektau, forthcoming c). It was also instantiated by references to refugees as bogus or so-called versus real or genuine in the UK press; and prawdziwy [‘real’] in the Polish press, as discussed in Section 12.2. The effectiveness of this discursive strategy aimed at delegitimising the rights of migrants can be attested by poll results indicating that many Europeans (including Poles, Spaniards, and Britons), indeed, doubted that refugees were really refugees (Ipsos Mori, 2018). Strikingly, the second most significant topic (following the threat theme) in all the three language subcorpora related to European political and administrative issues, which suggests that refugees were portrayed as something to be regulated and controlled through policy. This, on the one hand, further corroborates the finding, discussed above, regarding the representation of the movement of the displaced people as a crisis for Europe (rather than for the migrants). On the other hand, it switches the attention from the border to the political arena and makes it a political, rather than humanitarian, issue. In other words, discourse dominated by policymakers and other officials inevitably ignores the human consequences of these 342 policies. What is more, and especially in light of scarce mentions of non-government organisations or other alternative actors, it may imply that the management of the situation belongs firmly to the political sphere. In other words, it puts trust on governments and institutions to respond adequately regardless of the fact that their main concern is to seek public legitimation in order to remain in power. In a similar vain, the investigation has shown that the Spanish, Polish, and British press prioritised the opinions and beliefs of politicians at the expense of migrant voices (similar findings were reported by Berry et al., 2015; Georgiou, 2017; among others). The former were, thus, presented as the only valid and trustworthy actors to speak about the migrations and, in consequence, to act. They were often quoted directly which may, on the one hand, result from time and money constraints as well as the perceived reliability of elite sources. On the other hand, however, it may reflect the journalists’ attempt at appearing neutral and objective and guarding themselves from being held personally responsible for the expressions of hostility, prejudice, or racism, while conveying ideologically-loaded messages. The latter, in turn, were devoid of a platform to narrate themselves their stories and, therefore, could not be recognised as human agents (Couldry, 2008). This may be argued to lead to their marginalisation and exclusion from the political sphere as they were not allowed to claim their rights. It also legitimises decision-making on their behalf and assumes a homogeneity of their experiences. Scholars also point to evidence indicating that the exclusion of migrant voices contributes to negative public opinion and growing prejudice (Crawley et al., 2016). 343 The present study has also revealed a strong semantic preference of terms denoting refugees for quantification. Specifically, they were often presented as too numerous to be allowed to settle in Europe or in a given country (as attested, for example, by discussions surrounding the “refugee quotas”). This pseudodemographic argument thus utilises the fear of overpopulation referred to as “Malthusian paranoia” by Vollmer and Karakayali (2018), and the fear that the majority could turn into a minority unless a targeted minority disappears (Vermeersch, 2019). Furthermore, references to migration figures were also used to support several anti-immigration arguments. Most notably, as mentioned above, within the hierarchisation of refugees into deserving and undeserving the former were presented as much less numerous than the latter. Likewise, numbers were provided to claim (or imply) that a proportion of migrants were not whom they pretended to be, and that they held hostile intentions, for example to commit terrorist acts. The investigation has also uncovered a significant number of metaphors — such as water, commodity, and invasion metaphors — which presented refugees as too numerous and, consequently, out of control and dangerous. Conversely, counter-discourses claimed that migration figures were not exceedingly high, thus also building on the same misconception. The analysis has also revealed a tendency to collectivise refugees by presenting them as homogenous masses of people using a variety of different linguistic strategies such as metaphors, quantification, items denoting group membership, etc. The tendency to portray refugees as homogenous collectives rather than individuals contributes to their perception as different and alien: a Them rather than an Us (Sapolsky, 2017). 344 Finally, the study has revealed the presence of the theme of humanitarian aid. However, the analysis has shown it to be a minor trend. Therefore, it probably had little potential to influence the audiences, especially considering the fact that media effects are effective only when the evaluative direction is consistent (Meltzer et al., 2021). The Spanish press in particular advocated for helping refugees, but the argumentation was weak and limited in all three samples. Nevertheless, the study has uncovered a number of arguments used in support of helping refugees other than the victimisation tendency which has been the main or sole focus of previous studies into positive representations of refugees (e.g. Don & Lee, 2014; Cole, 2017). These arguments were often based on abstract notions of ethical/ religious/ legal responsibility (similar results were reported by Chouliaraki et al., 2017) as well as providing counter-arguments to anti-immigrant discourses, which might have, in fact, further supported the harmful associations of refugees with terrorism, economic burden, or violence by the sheer force of repetition. Thus, the apparent counterarguments can be argued to have actually fed into the dominant logic of (un)deservingness. Furthermore, as Dempster and Hargrave (2017) point out, they were unlikely to resonate beyond those who were already supportive. The victimisation trend was the most extensively developed argument which, while its objective may have been to raise awareness of the adversities faced by the displaced people, and although it may inspire individuals’ sympathies, risks portraying refugees as passive and dependent. This might lead to their perception as incapable to contribute to the society, which is considered to be key to successful integration (Bansak et al., 2017; Greenbank, 2017). What is more, studies have shown that the victimisation of migrants may be seen by the audiences as an attempt at manipulation and exploitation which might, consequently, lead to lack of response 345 and engagement (Bruna Seu, 2010). Furthermore, the lack of experiential overlap between refugees and the European audiences was shown to be an obstacle to eliciting empathetic reactions; in fact, its effect was even greater than the perception of geographical distance and ethic or cultural differences (Huiberts & Joye, 2018). Victimisation also presents asylum as an act of charity rather than a legal humanitarian obligation (Pickering, 2001, p. 177). Finally, previous contributions have indicated that the switch to victimisation in newspaper reports on refugees may be accounted for by a change of the government’s agenda rather than reflect a genuine concern about the plight of the displaced people (Don & Lee, 2014). On the other end of the spectrum were the “success stories” which, in turn, might have contributed to creating unrealistic expectations about the role of migrants in a society. They could also be perceived as “exceptions to the rule”. At the same time, this study has shown that the strategies which scholars have pointed out as potentially effective in shaping positive attitudes towards refugees and migrants — such as emphasising shared values and presenting refugees as economic innovators or political protagonists (Cole, 2017; Dempster & Hargrave, 2017) — were largely absent from the press reporting on the “refugee crisis”. In fact, there were no references to whom the displaced people actually were — their identities were mostly construed in terms of their movement to Europe and those aspects which conditioned their acceptance into the host societies. Similarly, the context of their plight was largely absent from the media reporting, as mentioned above, especially with respect to the role of Europe in shaping past and current global relations which should, arguably, be the starting point of alternative discourses on the “refugee crisis”. Finally, as Franquet Dos Santos Silva et al. (2018) argue, in lieu of either construing refugees as threatening and foreign collectives or putting their vulnerability on 352 sample. Furthermore, its preferred LIQUID metaphor accounted for the majority (60%) of all metaphors detected in this subcorpus. It was also least likely to employ ANIMAL and COMMODITY metaphors. At the same time, however, it was more likely than the other two subcorpora to construe refugees as invaders. 12.5. Methodological Triangulation: Evaluation and Limitations The present study employed a triangulation of corpus-assisted, mainly quantitative analysis with a more traditional qualitative investigation of a sample of the corpus with a focus on argumentation. This Section aims at evaluating the results obtained through the combination of the two analytical procedures with respect to the present study. Specifically, it will discuss the main costs and benefits of each approach by focusing on convergent and complementary findings. The comparison between the two analytical approaches did not uncover any discordant findings. Above all, it is important to note that the two types of analysis provided many shared findings and, therefore, contributed to the formation of a largely consistent frame of representation of refugees in the Spanish, Polish, and British newspapers. Most importantly, they have both revealed a significant focus on safety and security and the construal of (especially male) refugees as threat. They have also shown that terrorist and sexual violence were considered to be most newsworthy, and that both were often linked to Islam which, in general, has been shown to display a very negative discourse prosody. Both types of analysis have also revealed: the focus on legal, administrative, and institutional matters (most notably, the tendency to refer to the displaced people as illegal); the arbitrary hierarchisation of migrants and the conflation of the terms refugee and migrant (uchodźca and imigrant; refugiado and 353 inmigrante); the topoi of economic and cultural threat; the construal of refugees as deceitful; victimisation of children and women; and the presence of metaphors. Finally, both analyses have pointed towards humanitarian aid discourses within the analysed newspapers. At the same time, however, the two methodologies contributed findings which would not have been observed if only one of the methodologies were applied, but which can be considered as complementary to the wider picture. With respect to the analysis of argumentation, it allowed for uncovering the different ways in which the association between refugees and terrorism was made in the analysed newspapers, for example through the discussion on fake passports. It was also more effective in revealing the themes of cultural differences leading to harmful outcomes and of the effect of migrations on Europeans’ everyday lives. It is possible that these findings were not observed to such an extent as a result of the quantitative analysis because they are not associated directly with (many) specific linguistic items (in a way that, for instance, terrorism is associated with words such as terrorist, bombing, attack, or names of terrorist organisations). Moreover, it has uncovered a wider range of linguistic strategies of positive self-presentation than the quantitative strand did, which, although it was not the objective of the present study, also had a potential to shape attitudes and responses towards the displaced people. The analysis of argumentation also gave more scrutiny to humanitarian aid discourses. On the one hand, it has demonstrated them to be much more prominent than they appeared based on the quantitative results. On the other hand, it was more effective in identifying counter-arguments to the most widespread anti-immigrant discourses. 354 Finally, the qualitative analysis has revealed the extent to which securitisation and humanitarian aid discourses were mixed in the analysed newspapers by showing that, while many texts clearly opted for one of the two conclusions, arguments or topoi relating to the other were also present in the majority of them. In other words, it provided important evidence to the claim that discourses surrounding the 2015-2018 “refugee crisis” were to a large extent part of the humanitarian securitisation framework (Chouliaraki et al., 2017). This could not have been observed based on the quantitive analysis which took the corpus — rather than a text — as a unit of analysis. On the other hand, the use of corpus tools allowed for providing strong evidence of frequency and salience of other findings which also helped complement the wider picture of the representation of refugees. This was possible because corpus linguistics methods are much more efficient in detecting trends which gain salience only when considered collectively due to the cumulative nature of language. For example, the way how refugees were represented in terms of their geographical identities could only be evaluated critically when enough such choices were observed. Furthermore, it would not be possible to contrast real-life migration figures regarding the different nationalities if quantitative evidence of their presence in the analysed discourses were not available. A similar argument can be made regarding the divergent construal of refugees versus immigrants: it was only through the quantitative analysis of collocational patterns that these differences could be fully appreciated. 355 While both types of the analysis have shown the salience of Islam in the representation of the displaced people, corpus methods were also able to point out the frequent discussions of Christianity in the analysed discourses. This has, in turn, revealed significant differences between the construal of the two religions which has shown that the press in Poland, Spain, and the UK did not harbour bias towards religion in general. Other major trends that have only been uncovered as a result of the quantitative analysis include the significant focus on the movement of the displaced people and their accommodation in Europe; the role of gender in their construal; the foregrounding of European places as well as European actors and their voices; and the theme of health threat. Finally, with respect to metaphors, the use of quantitative methods helped put some of the frequently discussed mappings into perspective. In particular, the analysis has shown that the ANIMAL metaphors, discussed extensively in previous contributions, accounted for only a small fraction of all metaphors in the analysed sample (with the regrettable exception of the UK subcorpus), whereas metaphors which have received hardly any scholarly attention — such as the COMMODITY metaphor and, to a lesser extent, the (UNWANTED) GUEST metaphor — were among the most widespread. To sum up, the use of quantitative methods on a large dataset helped claim a fuller coverage of representations of refugees and provided more definite evidence of the generalisability of the findings. On the one hand, it was useful in identifying repetitive lexical patterns indicating more subtle but pervasive ideological representations such as, for example, the widely divergent representations of Christianity and Islam. Such trends might have been easily missed in a qualitative investigation even if it was conducted on the entire dataset. On the other hand, the 356 use of corpus linguistics tools allowed for detecting minor or marginalised discourses such as the association between refugees and health threat in the Polish subcorpus. Nevertheless, a singular corpus linguistics approach to discourse analysis would also suffer from weaknesses. The abundant amount of frequency-based information may make it challenging to identify truly significant findings and the analyst might be tempted to just “count what is easy to count” ((Stubbs & Gebirg 1993, p. 78). This limits the potential for explanation and interpretation for the patterns that emerge. Furthermore, even when significant findings are identified, examining each expanded concordance line is time-consuming and, therefore, often unfeasible. This way, alternative uses of a given item, for example in counterdiscourses, might be missed. Finally, corpus approaches are less suited to identifying discourses that do not follow a given linguistic pattern such as, for instance, the way that immigrants were sometimes construed as (potential) exemplary citizens, as well as more subtle (or even implicit) expressions of ideological discourses such as topoi. The qualitative analysis, on the other hand, is more suited to exploring complex discourses in more depth. This, as Baker and Levon (2015, p. 232) point out, “enables the analyst to discover subtle, and perhaps unexpected, patterns of socially meaningful language use, and to situate those patterns within a broader social, historical and ideological context”. For instance, in the context of the present study, the argumentation analysis has uncovered how securitisation and humanitarian discourses coalesce in the analysed press to the extent that traditionally antiimmigrant discourses may be employed apparently in support of humanitarian aid measures: for example by claiming that if refugees are not provided assistance, they might turn to religious extremism. Finally, focusing on a smaller sample allows for a more detailed examination of the emerging discursive patterns. For example, while 357 both analytical procedures have demonstrated that refugees were often construed as sources of conflict, the qualitative analysis has indicated its areas more specifically: from personal conflict between different elite actors (e.g. the Pope versus a politician), through societal conflict, to political and diplomatic frictions. It was also more suited to indicating specific linguistic devices such as, for instance, the use of even to present a given opinion as object of wide consensus which does not elicit the slightest protest. A potential shortcoming of employing qualitative analysis to a selection of texts, however, is that the findings cannot be generalised beyond the usually small sample. Furthermore, the text selection also influences on the results. For example, the present study focused on articles published following critical moments during the “refugee crisis” which might have marginalised discourses produced in “normal” circumstances (even though the sample did not contain solely texts regarding specific extraordinary events). Finally, qualitative analysis was arguably influenced by the perspective of the author (while it should be noted that corpus linguistics does not guarantee scientific neutrality either, it is arguably more analytically systematic). In conclusion, the multi-method approach has proven useful to identifying the dominant as well as marginal discourses surrounding the displaced people by allowing the researcher to mediate between a micro and a macro approaches to language and counterbalancing their respective advantages and weaknesses, thus enhancing their reliability and validity. However, the use of methodological triangulation should not mean that the shortcomings of each approach are disregarded simply because they might be compensated by the other approach’s strengths. In fact, „triangulation is only as strong as the study’s underlying theoretical, methodological, and analytical paradigms and the researchers' skills and [Document text truncated for crawler view.]