Dehumanization Revisited Media Reports on the Refugee Crisis in British Online Newspapers 1 Vesna Lazović Media reports have the power to change the opinions of the readership on any topic simply by repetitively using certain wording and structures. In certain cases, the reader may be prevented from forming an objective stance and unprejudiced perspective, which can prove dangerous if the topic in question is highly sensitive. Ever since the European refugee crisis in 2015, newspaper articles in the UK press have frequently used a range of dehumanizing language. This paper attempts to examine the issue by both quantitatively and qualitatively comparing the structures used to describe refugees, asylum seekers, migrants and immigrants in two consecutive years (2015 and 2016 respectively). Three British online newspapers were selected for this purpose: The Guardian, The Telegraph and The Independent. Through a combination of conceptual metaphor theory, critical discourse analysis (CDA) and corpus linguistics, this paper examines and analyzes the common collocations and metaphors used in media reports to refer to refugees at the outset of the crisis, and then again during the referendum year. The comparison between different periods is crucial, in that it can reveal whether expressions recur over time, especially because the media reports initially provided general information about the refugees such as their number and origins, but later also discussed the impact of the crisis on European countries. This analysis centres around the frequent use of dehumanizing language, which has not weakened over time, despite appeals from the UN Refugee Agency and numerous non-governmental organizations, who have sought to remind the press that words do matter in the migration debate. The findings show 1 This research is part of the project funded by the Slovenian Research Agency (BI- ME/ 21-22-005). AAA - Arbeiten aus Anglistik und Amerikanistik Band 46 (2021) · Heft 2 Gunter Narr Verlag Tübingen DOI 10.2357/ AAA-2021-0011 Vesna Lazović 116 that dehumanization is recurrent and ubiquitous in this context, resulting in the reinforcement of inhumane media treatment of this vulnerable group. 1. Introduction This paper aims to explore the dehumanizing language employed in British online media reports about refugees. In addition to facing extreme difficulties in their everyday lives, refugees rarely have the same opportunities and tend to be treated differently in media reports, which can result in further social exclusion. The European refugee crisis has been the most severe one since the Second World War, so it is crucial to understand the ways readers are acquainted with the situation as well as to acknowledge the discriminatory nature of media discourse. Dehumanization can be realized in different ways, one of which is the use of metaphors. These have the power to structure our perception and understanding, and their repeated use in media discourse can establish prejudice amongst the readership. Inhumane treatment of refugees by the media can mean that negative implications and attitudes towards these people can take root and subsequently affect the readers’ viewpoints. Apart from dehumanizing metaphors, other linguistic strategies can be used for negative constructions in the discourse, including the choice of particular words and frequent collocates, which will be examined in this paper. This study is inspired by the research conducted by Gabrielatos and Baker (2008) and the report written by Allen and Blinder (2013). From the beginning of the European refugee crisis in 2015 onwards, newspaper articles in the British press have made abundant use of dehumanizing language. By analyzing the discourse in three British online newspapers, this paper compares lexical association patterns, exploring reports on refugees both quantitatively and qualitatively. 2. Theoretical Framework In order to examine the role of discourse in creating societal attitudes and beliefs, the language used in online newspaper articles is analyzed within a critical discourse analysis framework (Fairclough 1995, van Dijk 1997, Wodak & Meyer 2009). This approach asserts that significant social changes and the manifestations thereof are embedded and maintained in discourse. In other words, discourse and the mass media play a crucial role both in disseminating potentially prejudiced ideologies and in re/ constructing and re/ creating personal attitudes and ‘knowledges’ (van Dijk 2003). More importantly, the media play a significant ideological role, where certain public problems are deliberately foregrounded and addressed as relevant. Thus language produces, maintains and changes social Dehumanization Revisited 117 relations of power through media discourse, constructing stereotyped assumptions, legitimating dominance and introducing inequality (Fairclough 1989). Critical discourse studies foreground the discursive mechanisms adopted in the realization of ideology and investigate how societal power relations are established and reinforced through language use in the media. In this way, the roles of the social, cultural and cognitive contexts of linguistic usage are revealed. The choice of different linguistic forms is directly linked to the process of producing systems of ideology and power hierarchy, since language itself can manipulate public opinion. The studies of Hartman and Husband (1974) and of van Dijk (1987) confirm that the mass media are a major source of prejudiced knowledge among people who have not yet created their beliefs and attitudes towards certain groups. Furthermore, metaphorical concepts are analyzed within a cognitive linguistic framework established by Lakoff and Johnson (1980) and later further developed by Kövecses (2002). According to them, metaphor is not only a device of the poetic imagination and a rhetorical flourish, but can actually structure our perceptions and understanding. They define metaphor as understanding one conceptual domain - the target domain in terms of another, experientially closer conceptual domain - the source domain. Conceptual metaphors thus structure our thinking and our perception of the world. Finally, corpus linguistics can help us study examples in context and obtain a more objective picture. Following this approach, linguistic phenomena can be described using both qualitative and quantitative methods (Baker 2006). Examples from corpora, especially larger ones, reveal trends and tendencies of lexical and grammatical associations, distributions and frequencies, collocational patterns and so on. Through using text analysis software, corpus analyses can reveal certain concordances and semantic preferences of a word, which might otherwise go unnoticed. 2.1. Dehumanization as Underlying Concept Dehumanization is often defined in dictionaries 2 as the process of depriving people of human qualities, personality or dignity and the process of making human beings seem like objects. Merriam-Webster Online Dictionary also lists the situations when dehumanization occurs, i.e. (a) when a person is subjected to “conditions or treatment that are inhuman or degrading”, (b) when a person is portrayed “in a way that obscures or demeans that person’s humanity or individuality”, or (c) when human involvement is removed or reduced. 2 For example, Longman Dictionary of Contemporary English, Oxford Advanced Learner’s Dictionary, Collins English Dictionary and Merriam-Webster’s Learner’s Dictionary. Vesna Lazović 118 Dehumanization as the concept is understood in this paper as a dangerous rhetorical tool, which can result in people believing that dehumanized groups should not be respected and should not be treated in the same way. Moreover, dehumanization can result in linguistic, semiotic and physical manifestations, e.g. violence or verbal abuse (Stollznow 2008), the lack of empathy or indifferences toward their well-being (Guillard and Harris 2019), but also in negative emotional responses, e.g. hatred, disgust or contempt (Haslam 2006, Utych 2018). Guillard and Harris (2019) claim that dehumanization is a pervasive force in society that can have harmful consequences and emphasize that language can facilitate dehumanization. They state that dehumanizing language can result in negative emotional responses and negative attitudes toward dehumanized groups (i.e. anger and disgust toward immigrants), can increase xenophobia and support for strong national security, and can shape perception. Consequently, such language can generate and reinforce anti-immigrant public opinion and anti-immigrant policies. By being exposed to such media messages, people may develop an implicit bias and the use of dehumanizing language could impact the perception of another group in the long term (Guillard & Harris 2019). Bar-Tal and Hammack (2012: 181) suggest that dehumanization involves categorizing a group as nonhuman with the aim to approve and justify harm toward the dehumanized group. In their opinion, it can be manifested in discourse in different ways by using rhetorical strategies and labels (e.g. subhuman epithets, biological/ zoological labels). Haslam (2006) proposes two forms of dehumanization: (a) animalistic dehumanization, when people are compared to animals and their uniquely human attributes are denied (e.g. the ability to reason, think critically or feel emotions), and (b) mechanistic dehumanization, when people are compared to objects or machines, and their true human nature is denied. There are both cognitive mechanisms and social conditions that enable dehumanization, which is a common social phenomenon (Haslam 2006, Stollznow 2008). The objectification or animalization of a group aims to deny the equal human status of an individual or a group and therefore justify their mistreatment, usually by depriving them of fundamental human rights (Stollznow 2008). When groups are dehumanized, they are excluded from the typical moral consideration given to other human beings (Haslam 2006). Stollznow (2008: 177-178) also underlines that dehumanization is a key word in the discourse of discrimination used by governments, movements and individuals to portray a target as ‘bad’, ‘inferior’ ‘unhuman’, and therefore unworthy of equal respect or protection. She (2008: 191) further regards the use of metaphorical derogatory epithets to be a form of everyday dehumanization and calls it linguistic objectification ( SOMEONE IS SOME- THING ). As a consequence, these dehumanizing terms “can present the enemy as an object that must be handled in a rational and unemotional Dehumanization Revisited 119 manner, absolving the agent of guilt, empathy and social or moral responsibility” (ibid.: 194). Using experimental data, Utych (2018) proved that dehumanization of immigrants through disease metaphors can influence political attitudes toward immigration by causing more negative attitudes toward immigrants. Dehumanizing refugees and immigrants by portraying them as contaminants, viruses, diseases, pollutants can affect political attitudes (Guillard & Harris 2019) and make the public support harsh and punitive action against the dehumanized (Utych 2018). This form of dehumanization denies attributes of affect and cognition to the group that is dehumanized (Tipler & Ruscher 2014), and as such is a powerful tool commonly used in political propaganda. 3. Previous Research on Dehumanizing Metaphors Numerous studies have been conducted on the use of metaphors in news articles about refugees. The media refer to these people using numerous terms, including migrants, refugees, asylum seekers, immigrants, and even illegals, sea arrivals and boat people. Baker et al. (2008) coined the acronym RASIM 3 (refugees, asylum seekers, immigrants and migrants) during the ESRC-funded project Discourses of Refugees and Asylum Seekers in the UK Press 1996-2006 (the RAS project), which combined corpus linguistics and critical discourse analysis. These studies have pinpointed a number of metaphors, mainly with negative implications, in immigration discourse. Santa Ana (1999) analyzed the 1994 anti-immigration political debate in California and identified the dominant conceptual metaphors IMMIGRANTS ARE ANIMALS along with less frequent metaphors IMMIGRANTS ARE WEEDS and IMMIGRANTS ARE COMMODI- TIES . Refaie (2001) discovered several main metaphor themes in Austrian newspaper reports on the Kurdish asylum seekers who arrived in Italy in January 1998, including: water, crime, business/ trade, war, animal, drama/ emergency and weight/ burden. Pickering (2001) singled out the water metaphor in the Australian media reports about refugees entering the country. O’Brien’s study (2003) on the metaphors used in the early 20 th century immigration restriction debate in the United States revealed a list of dehumanizing metaphors, which identified immigrants as natural catastrophes, objects, infectious diseases, conquering hordes, animals, and even waste materials from Europe. Baker and McEnery (2005) examined discourse which framed refugees as packages, invaders, pests or water in British news reports. In addition, Charteris-Black (2006) identified two prevalent immigration metaphors in the 2005 British election campaign: 3 The term refugees and the acronym RASIM will be used throughout this paper for people who are forced to leave their homes in order to escape horrific life conditions. Vesna Lazović 120 NATURAL DISASTER metaphors, which are mostly related to fluids, and CON- TAINER metaphors, both of which “discourage empathy with immigrants by treating them as objectsˮ (ibid.: 569). In both cases, fear and xenophobia towards immigrants is strongly reinforced. Media reports about refugees frequently use water metaphors (floods, tides, flows, etc.). Lee (2007) investigated the media representation of Chinese refugees arriving in California during the years 1850-1890. The common theme in the news of the time was the “trope of inundationˮ; as he explains, unwanted immigrants were described as ʻpouring inʼ, ‘swampingʼ, ʻflooding inʼ, or as coming in ‘tides’ and ‘waves’.(...) The metaphor alludes to an allied and recurrent image typical of this racist discourse, that of the fear of contamination, the terror of being made unclean by the filthy and sick. (Lee 2007) The above-mentioned list of dehumanizing metaphors is not exhaustive, though. Cisneros (2008) focused on an appalling metaphor IMMIGRANT AS A POLLUTANT in news media discourse on immigration, emphasizing the discriminatory nature of such discourse, in which immigrants are constructed as threatening substances and a mobile, toxic threat. Another study by Gabrielatos and Baker (2008) explored the discursive construction of refugees and asylum seekers in the UK press during the period 1996-2005. They noted that negative metaphors of water mass are often used to depict refugees in British newspapers. Similarly, KhosraviNik (2010) claimed that refugees are predominantly depicted in a negative manner, since they are presented via water metaphors and metaphors of natural disasters and labelled as danger, burden and law abusers. In addition, Cunningham-Parmeter’s investigation (2011) focused on legal texts, where the most prevalent metaphors identified were IMMIGRANTS ARE ALIENS , IMMIGRATION IS A FLOOD , and IMMIGRATION IS AN INVASION . Parker (2015) analyzed the representation of refugees and asylum seekers in the UK and Australian print media in the first decade of the 21 st century (2001- 2010). He discovered that metaphors of criminals and water were commonly used in both countries, but in different ways. In Australia, the focus was on border protection, while in the UK, it was on repatriation of refugees. Two more recent research studies have found other instances of metaphor use. Musolff (2017) tracked so-called parasite metaphors in weblogs and discussion forums, where immigrants were depicted as parasites, leeches, or bloodsuckers, whilst Marder (2018) readdressed the implications of the metaphor which links refugees to weapons of mass destruction (first discussed by Greenhill 2010). Table 1 summarizes the negative metaphorical concepts and lists the source domains used in media reports on RASIM and immigration. Dehumanization Revisited 121 Target domains Source domains REFUGEES ASYLUM SEEKERS IMMIGRANTS MIGRANTS ARE ECONOMIC / SOCIAL BURDEN AN UNWANTED PROBLEM SAFETY / SECURITY / ECONOMY / HYGIENE THREAT SUBHUMAN ENTITIES NON - HUMANS ANIMALS INSECTS PARASITES A SOURCE OF INFECTION AND CONTAMINATION INDIGESTIBLE FOOD CONQUERING HORDES INVADERS WASTE MATERIALS POLLUTANTS MATERIAL OBJECTS DISEASED ORGANISMS INFECTIOUS DISEASE COMMODITIES NATURAL DISASTER ( FLOOD , INFESTATION ) TOXIC WASTE WEEDS Table 1. The Metaphorical Concepts Found in Media Reports on RASIM As can be seen from Table 1, reports widely use metaphors of invasion, disease and natural disaster. The common denominator for all of the source domains is a threat, which evokes a negative cognitive concept: that of refugees threatening the integrity, the security, the economy and the wellbeing of society and the nation. Such a narrative can have harmful consequences, since the inhumane discursive treatment of a vulnerable and marginalized group can lead to inhumane or adverse social policies and social treatment (O’Brien 2003: 44). Moreover, if refugees are seen as a threat to the community, harmful policies and public actions against them can be easily justified as measures of self-defense (Brennan 1995, Saxton 2003). 4. Research Methodology The target years analyzed in this paper included the year 2015 and the year 2016. These two consecutive years were chosen because they marked Vesna Lazović 122 the watershed events with far-reaching consequences for the whole continent - the outbreak of the crisis in 2015 and the Brexit referendum in 2016. Although the referendum about the UK’s membership in the EU had been on the table for several years, it was finally held in June 2016 amidst the difficult times the EU was facing while struggling with problems regarding the refugee crisis and the refugee quotas. The historical Brexit decision may have been a surprise, but the UK had always wanted stricter immigration controls and tougher immigration restrictions even for the EU citizens. This research focused on the context surrounding RASIM words in the online newspaper articles from three British quality newspapers: The Guardian, The Telegraph and The Independent. The three online newspapers were chosen because they are the leading national broadsheet newspapers and their role cannot be neglected. According to National Readership Surveys (2017), their reports reached the vast readership. Chart 1 shows the number of average monthly visitors to different newspaper websites in the UK from 2013 to 2016. It is evident that the three online newspapers chosen for the analysis were all ranked among the first five. The other two leading websites are categorized as the tabloid newspapers or the popular press. Chart 1. Newspaper Websites Ranked by Monthly Visitors in the UK from 2013 to 2016 (in Million Visitors). Source: National Readership Surveys, Statista Dehumanization Revisited 123 The website of The Guardian had 8.8 million monthly visitors in 2015 and 7.7 million visitors in 2016. Similarly, 9.1 million visitors accessed the website of The Telegraph on a monthly basis in 2015 and approximately 7.3 million monthly visitors accessed the website in 2016. The Independent reached approximately 5.3 million visitors monthly in 2015 and 4.4 million visitors monthly in 2016. Since the paper tackles the sensitive issue of refugee treatment, it should be noted that political stances of these newspapers are not the same. The Guardian is seen as Britain’s most left-wing newspaper, The Telegraph as predominantly right-wing, while The Independent is viewed as having a center-left political stance (Smith 2017). 2015 2016 2015 and 2016 analysis quantitative quantitative qualitative corpus Timestamped JSI web corpus 2014-2020 English Timestamped JSI web corpus 2014-2020 English representative RASIM examples in Timestamped JSI web corpus 2014-2020 English subcorpus GTI 2015 GTI 2016 GTI 2015 and GTI 2016 approx. no. of words 106,251,726 64,619,222 / methodology Sketch Engine, automatic search Sketch Engine, automatic search examination of concordance lines results word sketches and concordances word sketches and concordances examples in broader context Table 2. Overview of the Methodology The time periods observed were the years 2015 and 2016, and the corpus used for the investigation was the English Timestamped JSI web corpus 2014-2020, which allowed a chronological search to be conducted on the chosen sites. This English corpus consists of news articles gained from their RSS feeds, which is updated on a daily basis (cf. Sketch Engine). Using the text analysis software Sketch Engine, the corpus was further subdivided into two subcorpora: GTI 4 2015 and GTI 2016, which comprised the data found on the official websites of these three newspapers in the years of 2015 and 2016, respectively. The analysis was both quantitative and qualitative in 4 The subcorpora included the articles on the official sites: www.theguardian.com, www.telegraph.co.uk, and www.independent.co.uk, which were found within the chosen timestamped corpus in the respective years. The acronym GTI stands for the names of the newspapers in question (The Guardian, The Telegraph and The Independent). Vesna Lazović 124 nature and consisted of three major methodological steps, summarized in Table 2. It should be noted that each subcorpus included reports on various topics. RASIM words were mostly found in the reports covering the EU refugee crisis, but also immigration worldwide. After the subcorpora were created, the first step was to check the word sketches for RASIM words separately. The results were considered for further discussion if they fulfilled the two criteria: 1) the logDice score, showing the strength of a collocation pair, was equal to or above one, and 2) the raw frequency of their co-occurrences was at least ten for GTI 2015 and at least five for GTI 2016. All of the raw automatic frequency counts were then normed to a basis of a million words of text (pmw - per million words), so that the distribution could be compared across the subcorpora. The threshold value of logDice was set to one so that a wider range of collocation candidates be included in the analysis, primarily because the scores in this context ranged from 0.01 to 6.92 (mostly from 1 to 6). Log- Dice makes the comparison across different corpora easier since it is not affected by the corpus size (Rychlý 2008). Within the second step, additional qualitative analyses were carried out to uncover and examine patterns. To verify the quantitative findings, the concordance lines were further analyzed and thoroughly examined in order to ascertain the surrounding context and identify metaphors and other immediate collocates. Looking at the broader context is crucial since collocation sets can provide “a semantic analysis of a word” (Sinclair 1991: 115- 116) and also reveal information about “the most frequent or salient ideas associated with a word” (Gabrielatos & Baker 2008: 10). 5. Findings and Discussion This paper looks at dehumanizing language used to represent RASIM, primarily by investigating the use of dehumanizing metaphors, i.e. examples of RASIM being compared to non-human entities. In addition, dehumanization can occur more subtly through other linguistic strategies, such as the use of particular wording. For that reason, the paper further investigates collocational patterns and word choice surrounding RASIM. Choosing particular words over others can create different cognitive and emotional responses in individuals, who may not be aware of this process (cf. Utych 2018). Finally, the overuse of quantification to premodify RASIM can also obstruct the process of their identification with other human beings and also the process of readers empathizing with them. Quantifiers in this respect may have a devastating effect, since the focus of the information switches away from real people and their life stories. This part addresses the quantitative findings of both subcorpora - GTI 2015 and GTI 2016, including the frequency of RASIM words, their word sketches, i.e. typical multiword units and concordance listings. As noted Dehumanization Revisited 125 earlier, both subcorpora include the articles found in the TimeStamped JSI web corpus 2014-2020 English for each newspaper and for each year separately. It also includes a qualitative discussion of examples found in both subcorpora and the word choice surrounding RASIM. The first step involved the automatic identification of RASIM words. Table 3 summarizes the frequency and distribution of RASIM words across the newspapers and time periods. Table 3. The Frequency and Distribution of RASIM words Concerning the subcorpora size, GTI 2015 is significantly bigger than GTI 2016. There are approximately 106,251,726 words in the articles found in these three newspapers in 2015 compared to 64,619,222 words found in 2016. However, when the normed frequency counts of RASIM words are cumulatively compared, it is evident that they were used slightly more often in 2016 (1226.05 occurrences pmw 5 ) than in 2015 (1057.16 occurrences pmw). This difference is not of crucial importance given the fact that the crisis was widely reported in both years. 5.1. Analysis of GTI 2015 Since the primary research idea revolved around the question of dehumanizing language, the next step was to investigate word sketches and collocation patterns, thus identifying association patterns possibly connected with RASIM words. The most typical collocates for every lemma in GTI 2015 with the logDice scores are shown in Table 4. 5 The counts were normed in relation to the approximate number of words in the whole corpus (pmw - per million words). Vesna Lazović 126 LEMMA/ PHRASE REFUGEE ASYLUM SEEKER IMMIGRANT MIGRANT Modifiers of a lemma/ phrase nationality: destitute 5.68 nationality: nationality: Syrian 4.91 Mexican 2.24 non-EU 3.89 Eritrean 4.43 Cuban 2.13 Eritrean 3.31 Somali 3.36 status: Burmese 3.22 Rohingya 2.66 undocumented 4.25 EU 2.21 Sudanese 2.36 illegal 2.63 Bangladeshi 2.17 Iraqi 1.63 would-be 1.07 Mediterranean 2.12 Afghan 1.45 other: Rohingya 2.06 numeral: secondgeneration 3.99 Sudanese 1.17 thousand 1.91 first-generation 2.92 status: feelings/ emotions: undocumented 3.41 desperate 2.27 irregular 2.56 vulnerable 1.08 would-be 2.12 status: unemployed 1.59 genuine 1.93 illegal 1.43 skilled 1.16 feelings: desperate 1.99 location: Calais 5.61 Nouns and verbs modified by a lemma/ phrase UNHCR 4.59 solidarity 1.64 resettlement 4.54 crisis 3.37 quota 3.28 influx 3.20 camp 3.08 exodus 2.20 intake 1.60 welcome 1.52 status 1.23 shelter 1.14 Verbs with a lemma/ phrase as an object resettle 6.92 fail 1.23 deport 2.14 resettle 3.54 redistribute 3.40 deport 3.12 relocate 2.50 drown 3.03 drown 1.77 camp 3.00 house 1.71 ferry 2.92 strand 1.37 smuggle 2.78 accept 1.30 rescue 2.52 Dehumanization Revisited 127 Table 4. Word Sketch Results for RASIM with logDice in GTI 2015 Regarding the modifiers, RASIM words are most frequently premodified by adjectives denoting nationality (Syrian, Mexican, Eritrean, Somali), feelings and emotions (desperate, vulnerable), immigration status (illegal, undocumented, irregular, genuine), encampment locations (Calais), but also by quantifiers and numerals, as will be discussed later. It should be noted that negative adjectives prevail when the refugees’ status is described, which is consistent with the hardship and torment they experienced at the time. They could not prove their status and could not reside legally, therefore faced problems obtaining valid visas and getting a job. The only exception is the adjective skilled in the noun phrase skilled migrants. When the context is further examined, it becomes clear that the phrase is used to refer to migrants who have particular skills required by the country (in this case, the UK) and who decide to move in pursuit of better employment conditions, and not to people who have been forced to leave their homes. The Sketch Engine statistics also reveal common patterns and co-occurrences of nouns and verbs. The occurrences of the nouns crisis, camp, UN- HCR, status and shelter, or the verbs arrive, and camp are expected, and refer to the immediate events following the initial crisis. These collocates denote their temporary residence (camp, shelter), their journey (cross, ferry), but also describe how they left their countries (flee, escape) and how they were made to return back (deter, deport). Moreover, the words referring to their treatment during the strenuous journey imply their illegality (smuggle) and their inability to move (strand). This is shown in examples 1 and 2: (1) As many as 100 Britons are believed to have been jailed in France during the past year for attempting to smuggle migrants across the Channel into the UK. [The Guardian, 23 June 2015, document no. 19524991] (2) More than a 1,000 refugees stranded in Hungary took matters into their own hands on Friday and attempted to walk to Austria. [The Telegraph, 4 September 2015, document no. 22375114] Other lexical verbs around RASIM words reveal the topics and events found in the reports in this particular period. Contextual clues suggest that some welcome 1.11 deter 1.62 strand 1.56 strip 1.27 Verbs with a lemma/ phrase as a subject flee 3.73 drown 3.63 drown 2.59 camp 3.61 risk 1.94 risk 3.00 pour 1.58 cross 2.07 arrive 1.52 flee 1.74 escape 1.28 attempt 1.20 flood 1.23 sleep 1.02 cross 1.01 arrive 1.02 Vesna Lazović 128 people managed to cross the Mediterranean and borders, risking their lives in the process, but were eventually accepted and welcomed, while others were not rescued and, unfortunately, lost their lives by drowning. By closely examining examples in context, two metaphors can be observed. First of all, the water metaphor is used to indicate the uncontrollable movement of people. At the same time, it symbolizes the loss of control over immigration (Van der Valk 2000: 234). Through the use of liquid metaphors (e.g. People flood / flow / stream / pour into a country) instead of neutral phrases (e.g. People travel / migrate / move / arrive to/ in a country), refugees are transformed into a mindless, overwhelming and potentially unstoppable mass (Bleasdale 2008). The features of water, such as lack of shape and colour, are hence attributed to humans, which leads to the metaphorical dehumanization of refugees (Kainz 2016). The most frequent water metaphors include the noun influx and the verbs pour and flood, as illustrated in examples 3-5: (3) Together these measures suggest that the EU’s Schengen agreement has effectively broken down, as EU states come up with their own unilateral initiatives to stop, or at least slow down, the refugee influx. [The Guardian, 14 September 2015, document no. 22585639] (4) Despite its generosity, Germany says that it cannot cope with the sheer quantity of refugees now pouring into it territory. [The Independent, 3 September 2015, document no. 22294481] (5) In response, refugees flooded instead into Croatia, which immediately tried to move them back into Hungary and Slovenia, prompting quasi-military manoeuvres from its neighbours. [The Guardian, 19 September 2015, document no. 22030904] The use of the word flood conveys the notion of danger. According to Charteris-Black (2006: 570), water metaphors are particularly powerful and compelling since flooding is connected to a more familiar experience, i.e. the severe floods in Britain in 2001, but also floods which have been increasingly occurring in the country in the last decade. This recall of recent experience can in turn evoke more powerful reactions in the readership. The second metaphor REFUGEES ARE PROPERTY OR COMMODITIES implies the notion of ՙ passiveness ՚ , as they have little or no ability to act or move and are stripped of their benefits and human rights. EU officials make decisions about their future, resettling and relocating them to different places according to the EU’s requirements and criteria. Sometimes, refugees are stranded and cannot move further unless they pay to be smuggled in desperate attempts to reach Europe. It seems they are unable to control their lives and act properly since they largely depend on others. The use of this Dehumanization Revisited 129 metaphor is an example of mechanistic dehumanization, since human nature characteristics including individual agency are denied to others and refugees are seen as object or machines (cf. Haslam 2006). They are in this way represented as passive and inert, lacking the autonomous agency. In the subcorpus RASIM words are frequently used with the noun resettlement and the verbs resettle, redistribute and relocate (examples 6-7): (6) The €50m (£36m) scheme is aimed at redistributing refugees across the continent, using criteria including GDP, population size, unemployment rates and past numbers of asylum seekers and resettled refugees. [The Independent, 14 May 2015, document no. 18279062] (7) On Tuesday, interior ministers also decided to relocate 120,000 refugees among most of the EU states, defying opposition from eastern countries Hungary, the Czech Republic, Romania and Slovakia. [The Telegraph, 25 September 2015, document no. 22932540] The qualitive analysis revealed other verbs as well. Example 8 contains the water metaphor flooding to the region, but also the verb dump: (8) Germany has accused Austria of dumping asylum-seekers near its border to force them to cross into Bavaria, escalating a row between the countries over thousands of refugees flooding to the region. [The Telegraph, 28 October 2015, document no. 24975468] The action of dumping is mainly associated with waste, garbage and trash. The search for other lexical verbs used with RASIM words is, however, beyond the scope of this paper. 5.2. Analysis of GTI 2016 The subcorpus GTI 2016 with approximately 65 million words was much smaller in size, but the findings reveal a striking similarity with the 2015 results, as the comparison shows consistency in terms of semantic preferences and metaphors. Based on the collocation sets, the media hype surrounding the referendum did not significantly influence the treatment of refugees. Table 5 lists the word sketches for RASIM words with the logDice scores in GTI 2016. Vesna Lazović 130 LEMMA/ PHRASE REFUGEE ASYLUM SEEKER IMMIGRANT MIGRANT Modifiers of a lemma/ phrase nationality: destitute 4.63 nationality: nationality: Syrian 4.06 unaccompanied 2.45 non-EU 2.19 non-EU 4.79 Eritrean 2.89 Cuban 1.64 Albanian 3.40 Angolan 2.73 Mexican 1.46 EU 2.86 Somali 2.33 status: status: Ugandan 1.71 undocumented 3.79 irregular 3.07 Sudanese 1.43 illegal 1.86 undocumented 2.53 Afghan 1.27 other: unaccompanied 1.46 numeral: second-generation 3.35 would-be 1.37 thousand 1.25 first-generation 2.94 numeral: feelings/ emotions: thousand 1.10 desperate 1.23 location: status: Calais 4.49 unaccompanied 5.30 location: Calais 4.52 sexual orientation: lgbti 2.08 age: child 1.56 Nouns and verbs modified by a lemma/ phrase resettlement 3.76 taskforce 3.38 crisis 3.00 camp 2.76 influx 2.16 encampment 1.65 quota 1.57 hostel 1.29 Verbs with a lemma/ phrase as an object resettle 6.42 deport 2.86 deport 3.55 deport 3.44 strand 2.18 resettle 2.28 drown 2.05 relocate 2.08 ferry 2.03 help 2.03 camp 1.83 deport 1.94 repel 1.52 deter 1.35 Dehumanization Revisited 131 strand 1.25 smuggle 1.21 Verbs with a lemma/ phrase as a subject flee 3.14 drown 2.12 drown 2.69 queue 2.61 camp 1.96 risk 1.96 board 1.64 arrive 1.13 Table 5. Word Sketch Results for RASIM words with logDice in GTI 2016 To begin with, RASIM words are once again premodified by adjectives denoting nationality (Syrian, Eritrean, Angolan, etc.), feelings and emotions (desperate) and encampment locations (Calais). The reports covered the refugee crises worldwide, discussed the need for skilled non-EU migrants, but also highlighted the problems with immigration status, as shown by the use of the adjectives undocumented, irregular and illegal. In addition, the collocates again denote their temporary residence (camp, hostel), hardship during their perilous journey (smuggle, strand, risk), which started when they had to flee their home country and use different means of transport to reach Europe (ferry, board, arrive), eventually queuing to register and receive food in camps. Unfortunately, not all of them survived trying to cross the Mediterranean in overcrowded boats. The newspapers reported on tragic events of refugees drowning when overcrowded boats capsized. The word sketches also reveal that the reports in 2016 focused more on RASIM deportation and foregrounded the topic of unaccompanied juvenile refugees, as illustrated in examples 9-10: (9) According to the EU police agency Europol, more than 10,000 unaccompanied child refugees have disappeared in Europe in the last two years. [The Independent, 2 April 2016, document no. 33936542] (10) Home Office figures show that the number of unaccompanied child asylum seekers arriving in Britain rose by 56% in the year to 2015, to 3,043. [The Guardian, 20 April 2016, document no. 35100594] In addition, the collocates indicate the recurrent use of two dehumanizing metaphors. The water metaphor is again found in the examples, in particular, with the use of the noun influx (example 11). However, the qualitative analysis of expanded concordances revealed that the nouns wave and flood were also used to refer to refugees in the subcorpus (example 12). (11) The 2.3 mile wire fence is being built by Austria on its border with Slovenia to “control” the refugee influx. [The Telegraph, 1 March 2016, document no. 31736571] Vesna Lazović 132 (12) Greece has been warned that it could be expelled from Europe’s passport-free Schengen zone if it does not manage its borders better and slow the current wave of refugees. [The Independent, 27 January 2016, document no. 29215095] This is in line with the research results of Baker et al. (2008: 287), who noticed that refugees tend to be quantified in terms of water metaphors, being dehumanized in this way, since they are depicted as an out-of-control, agentless, unwanted natural disaster. The metaphor REFUGEES ARE PROPERTY OR COMMODITIES emphasizes yet again the lack of agency and creates a negative framing, in which refugees are not seen as active human participants. RASIM are in the hands of others, who resettle and relocate them (examples 13-14): (13) In September, EU members pledged to relocate 120,000 refugees from Greece and Italy, the frontline states where most new asylum-seekers enter Europe. [The Guardian, 24 May 2016, document no. 37565363] (14) The Prime Minister has promised to resettle 20,000 refugees from Syria in the UK by 2020. [The Independent, 16 January 2016, document no. 28455272] Denying their active involvement and agency is a form of mechanistic dehumanization. Refugees seem not to have any possibility to take control of their own lives. Finally, the verb repel holds a particular semantic preference since it often co-occurs with insects, invaders and intruders. Due to its semantic preference, it could be argued that another dehumanizing metaphor REFU- GEES ARE INSECTS 6 is deployed (example 15): (15) On Monday migrants were repelled with teargas after using a road sign to break open gates. [The Telegraph, 2 March 2016, document no. 31780192] According to Haslam (2006: 257-258), this animalistic form of dehumanization occurs when uniquely human characteristics are denied to people (e.g. civility, moral sensibility and higher cognition). Consequently, they are perceived as uncultured and unintelligent. 5.3. Analysis of Quantification in GTI 2015 and GTI 2016 Quantification is another rhetorical strategy repeatedly used in the representation of RASIM. The use of “quantitative” collocations signals a burden or a problem, particularly those expressed through emotionally charged 6 It could also be seen as the war metaphor, where refugees are enemies and intruders who need to be defeated. Dehumanization Revisited 133 water metaphors (Gabrielatos & Baker 2008: 22), as they warn about the increasing and uncontrollable number of refugees. Luu (2015) argues that such metaphors “rapidly zoom out from a focus on individual humanity and individual stories”, which results in migrants becoming statistics, not people. Along with water metaphors, RASIM were frequently premodified by quantifiers. Transforming human beings into anonymous masses and pure numbers is another instance of dehumanization. The overuse of quantifiers may hide real problems or real people, as seen in example 16: (16) Last month the EU announced plans to relocate 40,000 refugees from Italy and Greece to elsewhere in Europe over the next two years. [The Guardian, 7 June 2015, document no. 18680163] Baker et al. (2008: 287) pointed out that “...about one in five references to refugees and asylum seekers are accompanied by quantification (the Number category).ˮ Their research was based on the analysis of a 140-millionword corpus of British news articles about refugees, asylum seekers, immigrants and migrants in the UK press, from 1996 to 2006. Similarly, the results of this research study reveal an abundance of numerals surrounding RASIM. Table 6 shows the number of occurrences of each RASIM word with a premodifying numeral in both subcorpora (three collocates to the left within the same sentence were included in the search). subcorpus GTI 2015 GTI 2016 raw counts per million raw counts per million refugee 2410 22.68 1822 28.20 asylum seeker 415 3.91 369 5.71 immigrant 302 2.84 238 3.68 migrant 1447 13.62 885 13.70 TOTAL 4574 43.05 3314 51.29 Table 6. The Frequency Counts of RASIM Words Premodified by a Numeral As can be seen, there is an apparent increase per million words, from 43.05 pmw in 2015 to 51.29 pmw in 2016 with regard to the total number of words in each subcorpus. However, there is one similarity when only the total number of RASIM occurrences is taken into consideration: in both years, RASIM words were used with a premodifying numeral in approximately one in ten references (12.44% of all the occurrences in the subcorpus in 2015 and 13.16% in 2016). Example 17 from GTI 2015 can be used to illustrate the point: Vesna Lazović 134 (17) European Union governments have failed to reach agreement on how to divide up the tens of thousands of refugees pouring into southern Italy from north Africa, increasing Rome’s fears that it will be left to cope with the Mediterranean migrant crisis. [The Guardian, 16 June 2015, document no. 19374479] Refugees are treated as property, while the notion of passiveness is highlighted by the use of the phrasal verb divide up. Apart from the water metaphor pouring into southern Italy, the quantifier tens of thousands premodifies the noun refugees. This vague number can influence the reader’s perception of the crisis, since the situation is presented as unstoppable and uncontrollable. The output results contain many examples with premodifiers related to numbers, including more than a million migrants and refugees, 5 million undocumented immigrants, thousands more refugees, hundreds of migrants, the soaring numbers of immigrants, etc. Moreover, the reports extensively commented on the refugee quotas and intake each EU country was prepared to accept. Since the EU countries could not agree, their redistribution caused disagreements (examples 18-19): (18) Cameron limits UK’s Syrian refugee intake ‘to discourage risky journeys’ [The Guardian, 4 September 2015, document no. 22343342] (19) The Government has refused to take part in UN and EU refugee quotas and is instead taking 4,000 Syrian refugees a year from camps near the Syrian conflict zone. [The Independent, 29 January 2016, document no. 29355063] The emphasis on refugee quotas and the overuse of quantifiers in reports may have a devastating effect, since the focus of the information switches away from real people and their life stories (cf. cf. Lazović 2017). As a result, the reader’s empathy can slowly fade away. By turning refugees into faceless numbers, their voices are silenced and rights are suppressed. Yet again, individuals are deprived of their humanity and “lost in the mass sense of these volume terms” (Santa Ana 2002: 73). 6. Conclusion This paper has shown that dehumanization is linguistically constructed in media discourse on immigration with the reinforcement of water metaphors, the overuse of quantification and linguistic objectification. The results mirror the previous studies, which have extensively documented the negative representation of RASIM in the press. Dehumanization Revisited 135 The choice of words in close proximity to RASIM suggest that reports in three British online newspapers do negatively frame the subject of immigration. As Gabrielatos and Baker (2008: 13) emphasized, particular meaning attributes can be arbitrarily associated with RASIM if particular collocates are repeatedly used, esp. in the case of negative constructions. This research study has confirmed that dehumanization is frequently used in the reports on refugees. It is hence important to understand how dehumanizing language operates and how it can lead to serious consequences for dehumanized groups (Utych 2018). Language as a media weapon can construct dominant ideologies, change beliefs and systems and profoundly influence the policies surrounding immigration and the way society treats migrants (cf. Cisneros 2008). When immigration is depicted as threatening, this can arouse feelings of fear and xenophobia, resulting in discrimination towards refugees. Even after numerous appeals by both NGOs and UN experts, the rhetoric on immigration and immigrants has not significantly changed. There are many different examples of harmful metaphors which compare refugees with criminals, animals, water, invaders and aliens. They are seen as a threat, burden and disaster. Owing to their effectiveness as instruments of social control for political organizations, mass media and other institutions, metaphors seem to be deliberately reestablished and reinforced. The majority of research studies on this topic highlight the negative impacts such metaphors can create. They change the way society treats immigrants, discouraging empathy, depriving immigrants of their identity and their rights, reinforcing conscious and subliminal fears, developing and fostering negative social images, and also legitimizing oppressive national practices (cf. Brennan 1995, Santa Anna 2002, Charteris-Black 2006). This list is, unfortunately, not exhaustive. Such inhumane media treatment can also encourage racism, xenophobia, marginalization and discrimination. The results presented in this paper could serve as a foundation for future research, examining the recurrence of themes in longitudinal studies (cf. Baker et al. 2008), comparing the use of metaphors in different languages (cf. Fijavž & Fišer 2020), and also searching for positive metaphors in the discourse on immigration (cf. Santa Ana 1999, Salshour 2016). Salshour (2016), for example, examined the use of liquid metaphors in the representation of immigrants in a New Zealand newspaper and discovered a surprisingly high number of positive representations, where these metaphors were used to portray immigration as beneficial. To conclude, dehumanization can leave deep marks on the ideology of a society, especially if / when used by politicians and journalists and other authority figures, in turn triggering hostility and diminishing interest in helping other people. Therefore, a human perspective should be adopted and more positive and immigrant-affirming metaphors should be used in the media, as words really do matter when issues that affect lives of so Vesna Lazović 136 many people are discussed. The society should strive for more balanced and more objective media reporting. References Allen, William & Scott Blinder (2013). Migration in the News: Portrayals of Immigrants, Migrants, Asylum Seekers and Refugees in National British Newspapers, 2010-2012. Oxford: Compas. Baker, Paul (2006). Using Corpora in Discourse Analysis. London: Continuum. Baker, Paul & Tony McEnery (2005). “A Corpus-Based Approach to Discourses of Refugees and Asylum Seekers in UN and Newspaper Texts.” Journal of Language and Politics 4 (2). 197-226. Baker, Paul, Costas Gabrielatos, Majid KhosraviNik, Michal Krzyzanowski, Tony McEnery & Ruth Wodak (2008). “A Useful Methodological Synergy? Combining Critical Discourse Analysis and Corpus Linguistics to Examine Discourses of Refugees and Asylum Seekers in the UK press.” Discourse & Society 19 (3). 273- 306. Bar-Tal, Daniel & Phillip L. Hammack Jr. (2012). “Conflict, Delegitimization, and Violence.” In: The Oxford Handbook of Intergroup Conflict. Ed. Linda Tropp. Oxford: Oxford University Press. Bleasdale, Lydia K. (2008). “Under Attack: The Metaphoric Threat of Asylum Seekers in Public-Political Discourses.” Web JCLI 14 (1) [online] www.bailii.org/ uk/ other/ journals/ WebJCLI/ 2008/ issue1/ bleasdale1.html. [20 June 2020] Brennan, William (1995). Dehumanizing the Vulnerable: When Word Games Take Lives. Chicago: Loyola University Press. Charteris-Black, Jonathan (2006). “Britain as a Container: Immigration Metaphors in the 2005 Election Campaign.” Discourse & Society 17 (5). 563-581. Cisneros, J. David (2008). “Contaminated Communities: The Metaphor of ‘Immigrant as Pollutant’ in Media Representations of Immigration.” Rhetoric & Public Affairs 11 (4). 569-601. Cunningham-Parmeter, Keith (2011). “Alien Language: Immigration Metaphors and the Jurisprudence of Otherness.” Fordham Law Review 79 (4). 1545-1598. Fairclough, Norman (1989). Language and Power. London: Longman. Fairclough, Norman (1995). Critical Discourse Analysis: The Critical Study of Language. London: Routledge. Fijavž, Zoran & Darja Fišer (2020). “Corpus-Assisted Analysis of Water Flow Metaphors in Slovene Online News Migration Discourse of 2015.” In: Darja Fišer & Philippa Smit (Eds.). The dark side of digital platforms: Linguistic investigations of socially unacceptable online discourse practices. Ljubljana University Press, Faculty of Arts. 56-84. Gabrielatos, Costas & Paul Baker (2008). “Fleeing, Sneaking, Flooding: A Corpus Analysis of Discursive Constructions of Refugees and Asylum Seekers in the UK Press 1996-2005.” Journal of English Linguistics 36 (1). 5-38. Greenhill, Kelly M. (2010). Weapons of Mass Migration: Forced Displacement, Coercion, and Foreign Policy. Ithaca, NY: Cornell University Press. Guillard, Celia & Lasana T. Harris (2019). “The Neuroscience of Dehumanization and its Implications for Political Violence.” In: Predrag Dojčinović (Ed.). Propaganda and International Criminal Law: from Cognition to Criminality. London: Routledge. Dehumanization Revisited 137 Hartmann, Paul & Charles Husband (1974). Racism and the Mass Media. London: Davis-Poynter. Haslam, Nick (2006). “Dehumanization: An Integrative Review.” Personality and Social Psychology Review 10 (3). 252 - 264. Kainz, Lena (2016). “People Can’t Flood, Flow or Stream: Diverting Dominant Media Discourses on Migration.” Oxford Law Faculty. [online] www.law.ox.ac. uk/ research-subject-groups/ centre-criminology/ centreborder-criminologies/ bl og/ 2016/ 02/ people-can’t. [12 August 2020] KhosraviNik, Majid (2010). “The Representation of Refugees, Asylum Seekers and Immigrants in British Newspapers: A Critical Discourse Analysis.” Journal of Language and Politics 9 (1). 1 - 28. Kövecses, Zoltan (2002). Metaphor: A Practical Introduction. Oxford: Oxford University Press. Lakoff, George & Mark Johnson (1980). Metaphors We Live By. Chicago: University of Chicago Press. Lazović, Vesna (2017). “Refugee crisis in terms of language: from empathy to intolerance”. B.A.S./ British and American Studies 23. 197 - 206. Lee, Gregory B. (2007). “Chinese Migrants and the ‘Inundation’ Metaphor. Risk, Representation, Repression: Constructing and Manipulating Fear of the Chinese Other.” East-AsiaNet Workshop: Framing Risk: Hazard Perceptions as a Crucial Factor in Imagining East Asia, Lund, Sweden, June 2007. [online] halshs.archi ves-ouvertes.fr/ halshs-00188544/ document. [30 June 2020] Luu, Chi (2015). Migrants, Refugees, and Expats: How Humanity Comes in Waves. daily.jstor.org/ language-of-migrants-refugees-expats. [5. June 2020] Marder, Lev (2018). “Refugees Are Not Weapons: The “Weapons of Mass Migration” Metaphor and Its Implications.” International Studies Review 20 (4). 576 - 588. Merriam-Webster Online Dictionary: America’s Most-Trusted Online Dictionary. [online] www.merriam-webster.com [15 November 2020] Musolff, Andreas (2017). “Dehumanizing Metaphors in UK Immigrant Debates in Press and Online Media.” Monika Weronika Kopytowska (Ed.). Contemporary Discourses of Hate and Radicalism across Space and Genres. John Benjamins Publishing Company. 41 - 56. National Readership Surveys (2017). “Newspaper Websites Ranked by Monthly Visitors in the United Kingdom (UK) from 2013 to 2016 (in million visitors).” Chart. February 22, 2017. Statista. [online] https: / / www-statista-com.nukweb. nuk.uni-lj.si/ statistics/ 288763/ newspaper-websites-ranked-by-monthly-visitors-united-kingdom-uk/ [30 December 2020] O’Brien, Gerald V. (2003). “Indigestible Food, Conquering Hordes, and Waste Materials: Metaphors of Immigrants and the Early Immigration Restriction Debate in the United States.” Metaphor and Symbol 18 (1). 33 - 47. Parker, Samuel (2015). “Unwanted Invadersʼ: The Representation of Refugees and Asylum Seekers in the UK and Australian Print Media.” eSharp23: Myth and Nation. [online] www.gla.ac.uk/ media/ media_404384_en.pdf. [25 July 2020] Pickering, Sharon (2001). “Common Sense and Original Deviancy: News Discourses and Asylum Seekers in Australia.” Journal of Refugee Studies 14 (2). 169 - 186. Refaie, Elisabeth El. (2001). “Metaphors We Discriminate by: Naturalized Themes in Austrian Newspaper Articles about Asylum Seekers.” Journal of Sociolinguistics 5 (3). 352 - 371. Vesna Lazović 138 Rychlý, Pavel (2008). “A Lexicographer-Friendly Association Score.” In: Proceedings of Recent Advances in Slavonic Natural Language Processing. Ed. Petr Sojka & Aleš Horák. RASLAN, Masaryk University, Brno. 6 - 9. Salahshour, Neda (2016). “Liquid Metaphors as Positive Evaluations: A Corpus-Assisted Discourse Analysis of the Representation of Migrants in a Daily New Zealand Newspaper.” Discourse, Context & Media 13. 73 - 81. Santa Ana, Otto (1999). “Like an Animal I was Treated”: Anti-Immigrant Metaphor in US Public Discourse.” Discourse & Society 10 (2). 191 - 224. Santa Ana, Otto (2002). Brown Tide Rising: Metaphors of Latinos in Contemporary American Public Discourse. Texas: University of Texas Press. Saxton, Alison (2003). “‘I Certainly Don’t Want People Like That Here’: The Discursive Construction of Asylum Seekers.” Media International Australia Incorporating Culture and Policy 109. 109 - 120. Sinclair, John (1991). Corpus, Concordance, Collocation. Oxford: Oxford University Press. Smith, Matthew (2017). “How left or right-wing are the UK’s newspapers? ” YouGov Survey Results. [online] https: / / yougov.co.uk/ topics/ politics/ articles-reports/ 2017/ 03/ 07/ how-left-or-right-wing-are-uks-newspapers. [2 December 2020] Stollznow, Karen (2008). “Dehumanisation in Language and Thought.” Journal of Language and Politics 7 (2). 177-200. Tipler, Caroline & Janet B. Ruscher (2014). “Agency’s Role in Dehumanization: Non-human Metaphors of Out-Groups.” Social & Personality Psychology Compass 8 (5). 214-228. Utych, Stephen M. (2018). “How Dehumanization Influences Attitudes toward Immigrants.” Political Research Quarterly 71 (2). 440-452. van der Valk, Ineke (2000). “Parliamentary Discourse on Immigration and Nationality in France.” In: Ruth Wodak & Teun A. van Dijk (Eds.). Racism at the Top: Parliamentary Discourses on Ethnic Issues in Six European States. Klagenfurt: DRAVA-Verlag. 221-260. van Dijk, Teun A. (1987). Communicating Racism; Ethnic Prejudice in Thought and Talk. Newbury CA: Sage. van Dijk, Teun A. (Ed.) (1997). Discourse as Structure and Process. Discourse Studies: A Multidisciplinary Introduction. Volume I. London: Sage. van Dijk, Teun A. (2003). “The Discourse and Knowledge Interface.” In: Gilbert Weiss & Ruth Wodak (Eds.). Critical Discourse Analysis; Theory and Interdisciplinarity. Palgrave Macmillan. 85-109. Wodak, Ruth & Michael Meyer (Eds.) (2009). Methods for Critical Discourse Analysis. Second revised edition. London: Sage. Online sources Sketch Engine language corpus management and query system, www.sketchengine.eu Timestamped JSI web corpus, www.sketchengine.eu/ jozef-stefan-institute-newsfeed-corpus Vesna Lazović Department of English Faculty of Arts University of Ljubljana