Analysis of Online Comments to News Reports

Abstract

Having outlined the common methodological perspective that C.O.N.T.A.C.T. partners adopted for both research strands of the project, it is now time to turn to a general discussion of the results obtained. To this end, this chapter will focus on the analysis of the comments corpora that were compiled at the first stage of our investigation; through the application of different techniques and against the background of various theoretical standpoints, the following sections touch on topics of central importance for the discourse-analytic discussion of hate speech, broadly construed.

Having outlined the common methodological perspective that C.O.N.T.A.C.T. partners adopted for both research strands of the project, it is now time to turn to a general discussion of the results obtained. To this end, this chapter will focus on the analysis of the comments corpora that were compiled at the first stage of our investigation; through the application of different techniques and against the background of various theoretical standpoints, the following sections touch on topics of central importance for the discourse-analytic discussion of hate speech, broadly construed. More specifically, Sect. 3.1 discusses categorisation in the context of Othering and its use as a means of defending one’s identity against the perceived threat posed by minority groups in the Italian setting, with Sect. 3.2 building up on the topic of categorisation by zooming in on comments related to the LGBTIQ community in Lithuania and discussing stereotyping as another strategy for the expression of hate and discrimination. Moving on to the issue of xenophobia, Sect. 3.3 explores the discursive dynamics of Polish online “patriotism” and its interface with fear-mongering and incitement to hatred, while, remaining on the topic, Sect. 3.4 highlights the use of conceptual metaphors in comments related to migrants in Cyprus. Finally, turning to the discussion of indirectness in discriminatory discourse, Sect. 3.5 focuses on implicitness as a commonly used way of signalling an unfavourable stance towards minorities in Malta, and Sect. 3.6 examines the intricate ways in which constructed and fictive dialogue are used to legitimise xenophobic and homophobic discourse in the Danish context.

3.1 Categorisation and Defence Strategies

• Ernesto Russo and Pablo Bernardino Tempesta

Categorisation is a fundamental human cognitive process which allows us to recognise and understand reality, by grouping its objects into categories depending on some meaningful criterion (Cohen and Claire Lefebvre 2005). When it comes to the specific cognitive process of social categorisation, which divides individuals into social groups (Allport 1979), it is typically undertaken on the basis of common and shared characteristics of a group of people, as, for example, nationality, gender, age, skin colour, religion, etc. This enables us to view the relevant people more as members of a specific social group rather than as mere individuals.

In this respect, categorisation plays a key role in the process of stereotype-forming and, as Mazzara discusses, by extension, prejudice-forming too:

It is evident how the concept of stereotype is extremely connected with prejudice, to such an extent that it is both confused and associated with it. It is possible to claim that a stereotype is the cognitive core of a prejudice, a set of information and beliefs related to a particular category of objects [i.e. social groups etc.] elaborated into a unique, coherent, stable image able to uphold and to create a prejudice against them. In other words, the stereotype is able to funnel the evaluation of data into a prejudice.

(Mazzara 1997: 72, translation our own)

It follows then that mentally categorising individuals and/or behaviours into more generic groupings paves the way for the shaping of mental beliefs, which are in turn known as stereotypes, and which are sometimes formed on the basis of personal (often hostile and harmful) opinions, called prejudice.

This process of generalisation gives rise to a mechanism of contrast in which one tends to group together all those people with alike characteristics that one considers to be incompatible with one’s own worldview (also known as Weltanschauung). Through this latter process, which is generally known as Othering, a social group becomes (mentally) classified as not belonging to the individual’s in-group by means of a clear opposition (in terms of a characteristic like gender, nationality, religion, etc.). This often takes the form of vilification and “denies the Other those defining characteristics of the ‘Same’, [such as] reason, dignity, love, pride, heroism, nobility, and ultimately any entitlement to human rights” (Gabriel 2008: 213).

Connecting cognitive categorisation and stereotype-forming processes to the development of hate manifestations (both in verbal and physical forms) towards determined social groups, Bennett’s Developmental Model of Intercultural Sensitivity (1993) offers a comprehensive account of the role that defence instincts play here. As Castiglioni summarises (2005: 18–20), when a cognitive defence strategy is activated, only the in-group is favourably considered and assumed to stand above the other(s) in terms of intelligence, civilisation, historic roots, etc. Everything else that forms part of the out-group is condemned because of fear, which makes those in mental defence mode perceive themselves as being besieged. In this setting, defence often takes the form of denigration where others are represented in a negative way, and attributed undesired characteristics through simplifications based on limited knowledge (stereotypes). Denigration mainly takes the form of verbal hostility against different cultures, but there are also many cases in which people masquerade their aggressiveness as defence, by underlining the ‘dangers’ posed by an ethnic or religious group.

Such cognitive defence strategies seem to arise from the fear of not being able to maintain one’s self-schema, as well as from the need to counter the anxiety of confronting one’s weaknesses and tackle anything opposing one’s worldview. With this in mind, we will now focus on a few examples of comments that we have collected from Italian media as a way of showcasing this defensive approach to hate speech.

• (1) Domani la nostra città sarà per l’ennesima volta un deprimente palcoscenico di qualche migliaio di frustrati, vittime di aberrazioni della natura. ¹

  Tomorrow our city will again become a depressing stage for a thousand or so frustrated people, victims of nature’s perversion.

The homophobic comment above is found in an article reporting the unfolding of the 2015 gay pride parade in Milan with the presence of the city mayor. It was made by two city councillors, members of the Italian regionalist party Northern League (famous for its xenophobic positions). Besides the intention to attack the opposing party, this comment reveals their hostility towards the LGBT community as a whole, which they characterise as ‘frustrated’ as well as comprising ‘victims of nature’s perversion,’ hence highlighting their declassification through denigration. The use of the possessive adjective ‘our’ in ‘our city’ aims to stress the identification of a common good (the city) and emphasises their feeling of being threatened and besieged by the LGBT community which is evidently not considered to belong to the councillors’ in-group (Othering). Thus the defence mechanism manifests itself with the neat opposition that the councilmen build between themselves (and their audience) and the LGBT community with all its characteristics.

The following comment also belongs to the same article and constitutes another interesting example of a homophobic statement:

• (2) Che palle che ci fanno questi gay pride e i relativi componenti e pure i politici che gli accodano per i voti, pisapia docet. Ovviamente ognuno di noi deve esprimere la sua sessualità nel letto con chi più gradisce, contento lui/lei contenti tutti, ma non vedo perché devono fare queste RIDICOLE BUFFONATE e SONO ANCORA PIU’ BUFFONI COLORO CHE AUTORIZZANO A FARLE. ²

  What a load of bullshit these gay parades, their affiliates and also the politicians who join them to get more votes, Pisapia is the first of them. Anyone should be able to obviously express their sexuality in bed with whom they want, but I really don’t understand why they should be doing this RIDICULOUS NONSENSE and THOSE GIVING THEM PERMISSION TO DO IT ARE EVEN MORE RIDICULOUS.

Here, the reader commenting on the article categorises the LGBT community as engaging in nonsensical activities. In order to understand the defence mechanism at play here, we need to focus on the event described in the article, the gay parade, which the reader seems to be hostile to (despite him/her stating that everyone is free to privately express their sexuality). This hostility manifests itself not only in relation to LGBT community members but also to anyone supporting them (e.g. the mayor and city council). This extension towards anyone related to this social group represents a mechanism of generalisation juxtaposing ‘I, myself’ from ‘them’, thus highlighting the fear of an attack to one’s own identity. On the one hand, the commenter seems to be in favour of sexual freedom as a commonsense principle for all individuals, but then on the other, ironically labels gay parades as ‘ridiculous nonsense’.

We would like now to go through an example of a xenophobic comment taken from the Italian weekly news magazine L’espresso, addressed against the Roma community, which is often discriminated against in Italy:

• (3) Quando vedrò un ROM onesto nevicherà il 15 di Agosto! E’ l’ora di farsi sentire, di far capire a questa feccia che prima vengono i diritti degli onesti cittadini e poi i loro. Non se ne può più di vivere col terrore che ti vengano a svaligiare casa, causandoti molti danni per pochi euro di refurtiva. Basta!!! Che se ne tornino nei Balcani, devono capire che l’Italia deve essere un paese deromizzato. ³

  When I see an honest Roma person it will be snowing on the 15th August! It’s time to raise the voice. The time has come to make this scum understand that the rights of honest citizens come first and then theirs follow. Enough with living with the fear of burglars who cause lots of damage for just a few euros of loot. Enough!!! Let them go back to the Balkans, they need to understand that Italy has to be deromanised!

More so than the previous examples, this comment shows how the phenomenon of stereotype-forming is deeply rooted in society. With a strong emphasis on the use of (cynical) sarcasm, the reader underlines how being honest and being a Roma person is contradictory and essentially ‘as odd as snow in mid-August’. The reader’s cognitive process follows the line of advocating a common fight against a foe (‘them’) with a clear defence strategy of ‘us honest people’ against ‘them, dishonest Roma’ (as in the eternal fight between good and evil). The xenophobic climax is reached with the use of the term ‘scum’ to define the entire Roma group. This is a case where a stereotype is taken to an extreme, becoming prejudice, and where categorisation becomes a hate instrument. Roma people are pointed at as thieves or brigands from whom Italians have to defend themselves. In this sense, a strong nationalism underlies this particular example of categorisation, which is deeply rooted in the reader’s belief that the country needs to be ‘deromanised’. All in all, what is perceived as a huge social problem is given an extreme solution: as in the worst examples of racism, the user pictures society as a place where people should be divided on the basis of their ethnicity.

As the examples discussed above show, hateful discourse, prejudice-based remarks or even incitement to violence against certain individuals/social groups often arise because their identity and social roles have been respectively reduced to their ethnicity and anti-social actions, so much so that they are perceived as a threat to one’s (or to the whole nation’s) identity. In this respect, defence mechanisms, which emerge from the generalisation of particular characteristics allocated to determined social groups, aim at responding to the unpleasant emotions triggered by some perceived stereotype and at preventing the anxiety generated by the fear of a possible identity crisis and an attack on one’s own life context.

3.2 Stereotyping Vulnerable Groups

• Uladzislau Ivanou

Negative stereotypes and their influence on social inequality may often be underestimated, but the connection between stereotypes and the explosion of hate speech is nowadays becoming increasingly obvious. In our study of hate speech in Lithuanian newspaper comments, stereotypes were found mainly in comments made in response to articles encompassing either a neutral or a positive attitude towards the populations usually affected by xenophobia and homophobia; however, due to space restrictions, this section will focus solely on homophobia and its expression through the use of stereotypes. That said, and before moving on, it is important to note that, in the Lithuanian context, stereotyping of the LGBTIQ community affects male individuals engaging in homosexuality more than it does female ones.⁴ That is why the absolute majority of stereotypes concerning the LGBTIQ population in the present comments analysis applies to gay men and includes stereotypes identified through the use of keywords such as ‘gay pride’, ‘LGBT’, ‘homosexuality’, ‘homosexuals’, ‘gays’ and ‘sexual minorities’.

Stereotyping is not just a phenomenon but also a process, since stereotypes evolve and are constantly enriched. For example, as we will see, in the Lithuanian context, gay men are not only viewed as ‘chicken hawks’, but can also be stereotypically perceived as ‘zoophiles’, ‘fetishists’ or even ‘democratic scum’. This stereotyping process poses what has been labelled a “stereotype threat” (Inzlicht and Schmader 2011), where hate speech transforms into action and can lead to hate crime. History has numerous examples of initially harmless stereotypes gradually transformed into isolated and later collective displays of hate speech and, finally, actions and crimes, involving mass extermination and harassment of vulnerable groups: witches, Roma, Jews, Armenians, homosexuals etc.

To understand stereotyping as a phenomenon and process, the context in which a stereotype is used is important. For instance, the stereotype of a ‘feminist’ will differ among conservatives, Christians and leftists, as would the stereotype of a ‘redneck’ among feminists and blue collars. A stereotype’s (positive or negative) connotations should also be taken into consideration. For example, according to van Ypersele and Klein (2006), gay stereotyping in Lithuania, which has always been negative in nature, is characteristic of hasty and extremely reductionist collective evaluations that are reproduced across generations. As Cuddy et al. (2009) put it, homosexuality belongs to the category of contemptuous stereotypes.

According to the Stereotype Content Model (henceforth SCM) (Fiske et al. 2002), any stereotype includes two levels of content: a descriptive one, which encompasses those qualities of a certain group that trigger emotions (and are therefore mocked in our setting), and an explanatory content, which deals with the underlying idea that motivates the expression of a stereotype in a certain context. Considering their descriptive and explanatory components, the stereotypes concerning homosexuality that were identified on the basis of the online comments collected as part of the C.O.N.T.A.C.T. project in the Lithuanian context are provided in Table 3.1.

Table 3.1 Stereotypes related to homosexuality in the Lithuanian C.O.N.T.A.C.T. corpus

Stereotypes are produced under the influence of a certain socio-political culture and are affected by significant external influences. Thus, in Lithuania, the influence of the Russian culture during the Russian-empire as well as the Soviet era should be taken into consideration. When it comes to stereotypes of gay men, it is possible to find both the Soviet trace or mediation of the Soviet and later Russian culture in the expressions ‘liberal gays’, ‘democratic trash’, and the local, national trace, when there is talk about homosexuality as a threat to national prosperity. A socio-linguistic analysis of image stereotypes underscores a mixed nature of stereotyping where an interplay of global and local influences is evident, with stereotypical images of gay men in Lithuanian including ‘piderastas’ (faggot), ‘pedikas’ (fag), ‘homikas’ (woofer), and ‘žydras’ (banana crammer), all of which are borrowed from Russian (cf. Jasiūnaitė 2005, 2006, 2009; Zaikauskas 2007: 114–115).

In terms of prevalence, stereotypes concerning gay men can be classified as typical, that is, universal and known in the neighbouring countries and in Europe as a whole (e.g. ‘Homosexuality is a disease’, ‘Gay men are effeminate’) and rare (‘Gay people are selfish’). Still, some universal stereotypes acquire additional local shades of meaning: thus, for example, the stereotype of homosexual ‘promiscuity’, and ‘decay of virtue’ gets extended in the Lithuanian setting to encompass an extreme form of liberalism, as is seen in the description of gay individuals as ‘democratic scum’ (“demokratijos šlamštas”), since decay of virtue is often associated in the region (Belarus, Latvia, Lithuania) with excessive democratism. In this setting, due to the freedom of gay people in the West, the value of democracy is discredited under the influence of the traditional, and sometimes quite authoritarian, political stance that was inherited from the USSR and is still upheld in domestic nationalism beliefs. This is also evident in the widespread, in our data, stereotype of ‘тaлepacты’ (people tolerating gays) in Europe, who are deemed to be too lenient with homosexuality.

Turning to what can be described as a rare stereotype of homosexuality stereotype present in the local media, we find the belief that homosexual individuals are selfish

• (4) O LGBT visuomenė, kurios tikslai egoistiniai ir visą visuomenę vedantys į niekur, meilės nenusipenė ir niekada nesusipelnys. ⁵

  LGBT people are selfish, they don’t deserve to be protected by the state.

Yet, the belief that gay people ‘think and love themselves only’ and ‘are not ready to be responsible’ and create a family is quite paradoxical, since homosexual individuals have no right to assume such a responsibility in Lithuania, where neither same sex union nor adoption are allowed by the state.

All in all, the investigation of stereotypes related to homophobia (and xenophobia, by association) is not just a research curiosity, but rather an inquiry into the weaknesses of our society, and its findings can inform both politicians and the public about some issues that should be addressed not only by politicians, but by education and media specialists too. As Barthes, who defined the stereotype as something solid, unshakable, unchanging and—at the same time—monstrous, notes, it is possible to presume that politics has no unshakable and unchanging territory (1975: 63). That is why policy-making in relation to the detection and prevention of hate speech and crimes should also have the objective of minimising negative stereotypes on top of preventing incitement to violence.

3.3 From ‘Patriotism’ to Hate: Axiological Urgency in Online Comments Related to Refugees

• Monika Kopytowska, Julita Woźniak and Łukasz Grabowski

In his Fear of Small Numbers: An Essay on the Geography of Anger‬, Appadurai points to “a lack of tolerance of any sort of collective stranger” tied to uncertainty resulting from blurring “the boundaries of national peoplehood” (2006: 45). The collective Self, contingent on membership in social groups and the shared identification with these groups, along with the in-group versus out-group construction, gains particular prominence in times of conflict and crisis of political, ethnic, cultural, religious, or economic nature. Defining the Other allows for the (re)definition of the Self and “functions to promote straightforward feelings of identification, empathy or disapproval” (Fowler 1991: 15). The dynamics of this process is captured by van Dijk’s “ideological square” (1998: 33), set to present ‘us’ in a favourable light and ‘them’ unfavourably, and consisting in emphasising ‘our’ good properties/actions, while highlighting ‘their’ bad properties/actions. In this sense, it is related to what Chilton calls delegitimisation, which involves

acts of negative other presentation, acts of blaming, scape-goating, marginalising, excluding, attacking the moral character of some individual or group, attacking the communicative cooperation of the other, attacking the rationality and sanity of the other (2004: 47).

Within the Media Proximisation Approach (Kopytowska 2015a, b), this process of polarisation is discussed in terms of cognitive-discursive operations within the domain of axiology characterised by three functions:

• 1. establishing axiological status: that is, ‘our’ values/norms;

• 2. delineating axiological conflict: that is, the incompatibility of ‘our’ values/norms with ‘their’ values/norms; and,

• 3. conveying axiological urgency: that is, responding to a threat posed (often by ‘their’ actions) to ‘our’ values/norms and accepting moral responsibility to act.

This axiological conflict is, for example, reflected by the most frequent migrant-related topoi/themes in the UK press, as listed by Hart (cf. Table 3.2), which, connected with the concept of physical or mental threat, are likely to generate fear and evoke strongly negative emotional responses towards migrants (Hart 2010, see also Richardson and Colombo 2013).

Table 3.2 Migrant-related topoi in the UK press (after Hart 2010: 67)

Another way to promote anxiety and panic is the use of metaphors conceptualising immigration as an invasion and as flooding the country (cf. Mahtani and Mountz 2002). Perceived in this way, migrants and refugees inevitably constitute a threat to the collective Self and the survival of a community as a cohesive unit (cf. Buzan et al. 1998): in the particular setting of the current migration crisis, coming from a predominantly Muslim background, they are likely to bring in beliefs and traditions incompatible with the European Christian worldview.

This stance gains even more relevance in the case of ethnically and religiously homogenous societies, such as the Polish one. Here, the sense of threat and axiological urgency is not only justified quantitatively (religious/ethnic majority), but also substantiated with historical experience and collective memory. More precisely, over the centuries, the sovereignty of Poland and its people’s status quo have been threatened by various Others during the time of partitions, World War I and II, and the Soviet Union’s domination. The cult of struggle for national integrity and militant opposition against the enemy have become a hallmark of Polish patriotism, with its slogan ‘Bóg, Honor, Ojczyzna’ (‘God, Honour, Fatherland’).

Against this background, we will explore, in this section, the discursive dynamics of Polish online ‘patriotism’ and its interface with fear-mongering and incitement to hatred against the Other(s). More specifically, we will demonstrate how, by appealing to collective memory, existing stereotypes and cultural/national values, media texts producers and commenters create a sense of axiological urgency and arouse strong negative emotions, thus possibly bonding the in-group and legitimising verbal and physical aggression directed at the out-group perceived as the threat and the enemy.

Refugee-related hate speech with “patriotic” undertones has been chosen for several reasons. Firstly, Poland is one of the EU countries ‘experiencing’ the crisis and ‘moral panic’ without being directly affected by the physical presence of migrants and refugees. Constituting a pillar of the dominant narrative of the country’s ruling conservative government and of right-wing media, this anti-migrant rhetoric has played an important role in the (mediated) construction of the crisis in Poland (cf. Kopytowska et al. 2017; Kopytowska and Grabowski 2017). Secondly, the analysed data powerfully reflect a peculiar form of the Polish ‘patriotism’ thriving on national pride, collective memory, belligerent courage, and the need to unite against an external enemy. Thirdly, such instances of anti-migrant discourse provide evidence of how ethnic and religious homogeneity can be effectively exploited in collective identity formation and Us versus Them construction in times of (perceived) crisis.

To examine how Polish “patriotism” is conceptualised and used as motivation for and justification of hate speech directed against refugees and migrants, three salient concepts associated with the national collective identity, namely ‘Polska’ (‘Poland’), ‘Polak’ (‘Pole’), and ‘nasz’ (‘our’), were identified and analysed in the Polish C.O.N.T.A.C.T. corpus. More specifically, after identifying comments for the keyword ‘uchodźcy’ (‘refugees’) following the common C.O.N.T.A.C.T. methodology,⁶ we tagged and parsed them using Sketch Grammar for Polish, which was developed on the basis of the tagset of the IPI PAN Corpus of Polish and implemented into the SketchEngine software (Kilgarriff et al. 2014).

In our corpus, ‘nasz’ (‘our’) collocates with the following nouns: ‘kraj’ (‘country’), ‘ojczyzna’ (‘fatherland’), ‘dom’ (‘home’), ‘zasada’ (‘rule’), ‘rodak’ (‘compatriot’), ‘kultura’ (‘culture’), ‘demokracja’ (‘democracy’), ‘historia’ (‘history’), ‘terytorium’ (‘territory’), ‘ulica’ (‘street’), ‘kobieta’ (‘woman’), while isolating comments made in response to YouTube videos alone, it correspondingly collocates with: ‘kraj’ (‘country’), ‘teren’ (‘territory’), ‘cywilizacja’ (‘civilisation’), ‘ojczyzna’ (‘homeland’), ‘rodzina’ (‘family’), ‘przodek’ (‘ancestor’), ‘granica’ (‘border’), ‘dziecko’ (‘child’). Subsequent word sketches and concordance analyses revealed several interesting patterns in the semantic prosody of these words and their “axiological potential”. For example, ‘Poland’ frequently appears in the phrase ‘Polska dla Polaków’ (‘Poland for Poles’), but also in such statements as that it is not a place for ‘szumowin imigracyjnych’ (‘immigration scum’), and it will not accept ‘tych bydlaków, pasożytów, gwałcicieli, terrorystów’ (‘these beasts, parasites, rapists, terrorists’). Spatial appropriation is also often visible:

• (5) nasza Ojczyzna należy do Nas Polaków ⁷

  our Homeland belongs to Us Poles

• (6) Nie pozwólmy żeby to ścierwo wkradało się na nasze tereny ⁸

  Let’s not allow this carcass to sneak into our territory

• (7) PiS nie wpuści brudasów do naszego kraju ⁹

  Law and Justice will not let allow these slobs to enter our country

• (8) nie ulegac Niemcom z ich bandyckim planem podrzucenia najezdzcow-dziczy do naszej Ojczyzny ! ¹⁰

  let’s not give into Germans with their thuggish plan to bring invaders-savages into our Homeland!

Refugees and migrants are not only axiologically downgraded by being dehumanised through animal metaphors, but are also presented as a threat of both moral and physical nature:

• (9) Nie szanują absolutnie niczego w naszej cywilizacji ¹¹

  They have no respect for anything in our civilisation

• (10) będą nasze dzieci, nasze wnuki zabijać za allaha ¹²

  they will kill our children, our grandchildren for Allah

• (11) będą nas rabować, bić gwałcić nasze kobiety ¹³

  they will rob us, beat up and rape our women

Finally, particular instances of the Other’s savagery are provided, as in the following comment:

• (12) pierwsze ataki islamistow w Niemczech na Polakow:

  • zamordowanie meczeta Polki,

  • zamordowanie Polaka kierowcy TIRa,

  • podpalenie Polaka bezdomnego,

  to sa poczatki, musimy byc przygotowani Polacy rodacy ze takie ataki beda coraz czestsze, nalezy zachowac czujnosc w Niemczech, Polsce i innych krajach UE, poniewaz Polacy stawiamy tame nachodzcom ta fala bedzie napierala na Polske i Polakow mocno,jak potop szwedzki i radziecki. ¹⁴

  first attacks of islamists in Germany affecting Poles:

  • killing a Polish woman with a machete

  • killing a Polish lorry driver

  • setting fire to a homeless Pole

  This is just the beginning, Polish fellows we have to be prepared for the fact that such attacks will be more and more frequent, we need to be on alert in Germany, Poland and other EU countries because we Poles prevent the invaders this wave will hit Poland as hard as the Swedish or Soviet Deluge.

In view of such a threat, counteraction seems to be a moral obligation and patriotic duty. Commenters thus pledge to protect the Polish people:

• (13) MY będziemy BRONIĆ naszych kobiet i dzieci przed zagrożeniem. ¹⁵

  WE will DEFEND our women and children against this threat.

• (14) Będziemy po swojemu z nimi załatwiać, Polacy znowu wezmą w swoje ręce rozprawę z islamem! ¹⁶

  We will deal with them in our own way; the Poles will again crack down on Islam!

At the same time, using both imperative forms and modals with deontic meaning as collocates of ‘we’ and ‘Poles’, they call on their compatriots to be vigilant and take action:

• (15) musimy się wszyscy przygotować na odpór tej zarazy. ¹⁷

  we have to prepare to fight off this plague.

• (16) musimy się połączyć wziąć się w garść i im pokazać co potrafią Polacy zanim będzie za późno. ¹⁸

  we have to unite, pull ourselves together to show them what Poles are able to do before it is too late.

Religion also emerges as a salient issue, and Islam is presented as incompatible with the Polish culture:

• (17) Polacy mowia NIE islamizacji Polski i basta! ¹⁹

  Poles say NO to the islamisation of Poland and that’s enough!

• (18) Polacy nie zgadzają się na islam w Polsce ! ²⁰

  Poles do not agree to Islam in Poland!’

Interestingly, references are also made to Jan III Sobieski, a Polish king credited with turning back the last great wave of Muslim expansion in Europe through his victory against the Turks in the battle of Vienna in 1683. In some comments, his actions are given as an example of patriotic spirit and something to be cherished and continued. From this perspective, letting in refugees, who are Muslims, will be a crime and lack of respect for national history and values:

• (19) i usilnie namawiam rodaków do bezwzgledności! Po co Jan III Sobieski ich pogonił, tak sobie a my mamy to w d…e? ²¹

  and I strongly urge my compatriots to be ruthless! Why John III Sobieski chased them away, for nothing and we don’t give a f…k?

As far as cultural and moral values are concerned, the Polish nation emerges as both self-sufficient and superior:

• (20) My Polacy NIC nikomu nie jesteśmy winni i nie potrzebujemyich “ubogacania” bo nie potrzebna nam jest ta ich “kultura”na poziomoe VII czy VIII wieku podniesiona do rangi “bogactwa kulturowego” za pomocą noży, kałachów i ładunków wybuchowych. ²²

  We Poles do not owe ANYTHING to anybody and we do not need their “enrichment” because we do not need their “culture” from the level of 7th or 8th c. elevated to the level of “cultural richness” with the help of knives, Kalashnikovs, explosives.

• (21) Naród Polski nie wyraża zgody na mieszanie wrogich obcych kultur i religii ze zdobyczami wartości narodowych w Polskiej przestrzeni terytorialnej. ²³

  The Polish people does not agree to mixing other hostile cultures and religions with the heritage of national values on the Polish territory.

Also, its virtue seems to lie in its readiness to take up arms if the need arises:

• (22) Jesteśmy jednym z najbardziej walecznych nacji w Europie. Sam wezmę udział w obronie moich rodaków, jeśli trzeba będzie. ²⁴

  We are one of the most gallant nations in Europe. I myself will take part in defending my compatriots if necessary.

• (23) Jesteśmy Polakami i Patriotami i chcemy dobrze dla Kraju naszych przodków którzy też o to samo niejednokrotnie walczyli z bronią w ręku. ²⁵

  We are Poles and Patriots and we want all the best for the Land of our fathers who often took to arms to fight for this.

As revealed by word-occurrence patterns in the analysed corpus, frequent appeals to the national pride, identity and history on the one hand, and to the need to protect Poles from both physical and moral threat, on the other, seem to represent a way of motivating people to act against refugees and migrants. As the threat is presented as imminent, there appears to be an axiological urgency to act, manifested in exclamations and imperatives expressing commands. To evoke fear of and anger at the refugees, their potentially harmful actions are given both in the form of factual occurrences (past and present tense, perfective and progressive aspect) and imminent acts of violence (future tense). At the same time, references are made to heroic deeds and sacrifice in the Polish history, to national heroes and to events which are salient in the Polish collective memory.

All in all, an opposition is constructed between a Christian, European, civilised world, with Poland at the forefront, and the world of the primitive Other.²⁶ Having no intelligence and morality, this Other has no respect for ‘our’ values and since reasoning with ‘them’ is out of the question, the only way to protect ‘our’ values is to use force. Hence we find in our corpus frequent calls to actions (involving physical violence) which should be (or are intended to be) taken. Importantly, since cyberspace, with its interactive and intertextual potential, allows groups and individuals with similar (often radical) ideas to connect, this hostile form of Polish ‘patriotism’ becomes salient in online discourse, thus generating a spiral of hate (cf. Kopytowska et al. 2017) in subsequent comments and conveying a sense of axiological urgency: We have to act before the Other(s) invade and destroy us.

3.4 Metaphors Related to Othering the Non-natives

• Fabienne H. Baider, Anna Constantinou and Anastasia Petrou

Recurrent linguistic strategies and specific discursive choices are often employed with a view to constructing the exclusion of the out-group and the cohesion of the in-group (Baker et al. 2008). Such discursive choices include referential strategies, like epistemic modalities attributing negative qualities to the out-group, exploitation of existing stereotypes, aggregation (i.e. referring to a homogeneous group that also shares the same intentions), as well as intensification (i.e. the excessive use of quantifying adverbs or adjectives). Such figures of speech reinforce conscious or subliminal fears related to the LGBTIQ community or to immigrants and encourage socio-cultural practices as well as interpersonal relations on the basis of negative tropes.

Metaphors are particularly important to study since “understanding the systematic nature of metaphor choices” allows us to understand in turn how “entire belief systems are conceived and communicated” (Charteris-Black 2005: 3, our italics). In particular, metaphors of threat, leeches or parasites are typically used to ostracise the non-natives (Musolff 2015; Baker et al. 2008), and have also been described as being coherent with the beliefs, actions, or imaginings of the person using them. In a way, metaphors can reveal the underlying conceptual frame of their producer and give access to a set of assumptions made by competent members of a discourse community about the ‘typical’ aspects of a member of a minority or any person belonging (or appearing as belonging) to that group. This then leads to the conceptualisation of metaphors as creating or confirming stereotypes (Zinken 2003).

In light of the above, the aim of the present section is to understand how in the context of the small Orthodox island of Cyprus, where almost no refugees had landed during the summer of 2015 and where religion plays an active part in politics and everyday life (cf. Baider 2017), xenophobic metaphors are used to construct the social Other in social media. Due to space limitations, this section focuses only on our analysis of comments retrieved on the basis of the keywords ‘refugee(s)’, ‘migrant(s)’ and ‘foreigner(s)’ in line with the common C.O.N.T.A.C.T. methodology. ²⁷

For these particular keywords, we collected 2446 comments. Our analysis of these comments in terms of polarity revealed that more comments were negative than positive but not overwhelmingly so, as Table 3.3 shows.

Table 3.3 Polarity analysis of the Cypriot C.O.N.T.A.C.T. corpus

Having collected and classified our data in this way, we then proceeded to identify the most common linguistic means that are used to negatively categorise the social groups at hand. Here, the most frequent means include metaphors, insults, proverbs and irony/sarcasm, but for the purposes of this section we will focus on metaphors as a means for Othering migrants, foreigners and refugees. Most work dealing with the use of metaphors in discourse related to migration (see, for instance, Santa Ana 1999) has shown that it is often being conceptualised as a natural or manmade disaster (typically referred to using lexemes related to floods, tsunamis or pollution) and migrants, refugees and foreigners as menacing animals, bacteria and beasts of burden.

Table 3.4 summarises those conceptual metaphors that we encountered in our collected comments the most, alongside some examples of phrases that seem to trigger the relevant classification within them.

Table 3.4 Recurrent metaphors used for Othering immigrants in the Cypriot C.O.N.T.A.C.T. corpus

Even though Table 3.4 shows the most typical anti–immigration discourse tropes, it is not the same metaphors that are found for each category of Other in our dataset. For example, the category ‘refugee(s)’, which is the most diversely constructed one, comprises mainly metaphors characterising the relevant individuals as disgusting animals (‘worms’ in 24) or pests (‘mice’ in 25):

• (24) τόσο θράσος τα σκουλίκια, αχάριστα αδέσποτα παρουσιάζονται με απαιτήσεις. ²⁸

  They have audacity these worms; these ungrateful stray animals come over with demands.

• (25) Οταν μια χωρα σε φιλοξενη σε ταιζει κσι εισαι ασφαλεις απο τον πολεμο που υπαρχει στην χωρα σου και εσυ αντι να μεινης στην πατριδα σου να πολεμησεις φευγεις σαν ποντικος διαμαρτιρεσαι αχαριστε? ²⁹

  When a country is hosting, feeding and protecting you from the war that is going on in your country, and you leave from your homeland like a mouse, instead of staying and fighting, how can you be complaining, you ungrateful?

Conversely, the category ‘migrant(s)’ seems to be more commonly attributed metaphors of outlaw, violence and disease:

• (26) otan efigan oi pappoudes mou epian me to plio kanonika kai me ton nomo stin agklia! eperasan apo iatrikes epitropes kai 1000 dio alla, toutoi kamnoun etsi? i adinatis na katalavis tin diafora metanasti me ton lathreo ³⁰

  When my grandparents left, they went with the ship normally and legally to England! They went through medical tests and a thousand other things; are these ones [the migrants] doing the same? Or are you unable to understand the difference between a legal and an illegal migrant?

As for the category ‘foreigner(s)’, it is more typically approached using the metaphor of amorality, which includes prostitution or sexual promiscuity and lack of decent/moral behaviour as the following examples of comments in response to an article entitled ‘A foreign woman abandoned her 12 year old child to go on a trip’ show:

• (27) Ayti einai nootropia poutanou. epiase o poutanos ton moulla tziae epie stin xwra tou. Alla en epire ton mikro mazi tis (…). etsi o poutanos eprotimise ton moulla para to mwro tis ³¹

  This is the mindset of a whore. The whore got the mullah and went to her country, but she didn’t take the boy with her (…). So the whore chose the mullah instead of her baby.

• (28) Prepei na bolla anomali, mana-teras, psyxoanomali… Thee mu…na analavei to moro to kratos tse sta tsakkidia e opia k an einai e akatanomasti. ³²

  She must be such a pervert, a monster of a mother, a schizo… Oh my God! The baby needs to be taken care of by the state and she should go to hell whoever she is…

Metaphors are an intrinsic part of the Othering process, and central to identity construction. As such, they could easily lead to social exclusion and marginalisation processes as well. Indeed, in example (25) above, the fact that refugees left their country is interpreted as a lack of courage and therefore the metaphor of the mouse is used. What is more, this animal metaphor could reveal a conceptualisation on the part of the speaker which may in turn lead to an acceptance of treatment reserved for pests against the target of this trope. In this vein, metaphors could act as a transition from the argument ‘migrants should be deported’ to the conclusion ‘any means are justified to do so,’ as they are, at the same time, the results of unspoken premises and inferred conclusions.

All in all, this study confirms that previously identified metaphors used to Othering migrants and refugees in other languages are also found in Cypriot discourse. However, some tropes are typical of specific social categories, such as amorality being used only in relation to female foreigners in our data. From our brief exploration of the topic, it seems safe to conclude that metaphors function as a topos, i.e. as a place to look for arguments and a place where those arguments are ready for use (Zagar 2010). Indeed, much like topoi, metaphors are part of a categorisation scheme: they enable the speaker to construe an argument for a given conclusion, which, in this case, is the equation of the Other to a negative ontological (animal) or social (prostitute) category.

3.5 The Implicit Dimension of Discriminatory Discourse

• Rebecca Vella Muskat and Stavros Assimakopoulos

From a purely legal viewpoint, one needs to prove intent to stir up violence and hate toward a specific minority group in order to establish that some statement constitutes hate speech. However, in order to accomplish a thorough understanding of discrimination in language use as a social phenomenon, one would need to broaden the definition of the term, so that it also includes strategies used to implicitly impart a negative stance towards a given minority. As Reisigl and Wodak characteristically note, while discussing prejudice in racist discourse,

a categorisation according to the sentence structure of the most obvious prejudices is only partially able to grasp latent meanings, allusions, indirect strategies, vague formulations, implications, and forms of argumentation, all of which can extend beyond a single sentence and characterise written texts or oral discourse connected with prejudice and racism.

(Reisigl and Wodak 2005: 21).

In a similar vein, van Dijk also observes that “various types of implicitness play a prominent role in texts about minorities,” and attributes this tendency to “face-saving strategies [which] require that speakers avoid expressing explicitly negative propositions about minorities” (1992: 225). Indeed, the Maltese strand of the C.O.N.T.A.C.T. project research showed that, perhaps in fear of appearing intolerant towards migrants and/or the LGBTIQ community, most commenters who expressed a negative stance towards these groups did so implicitly, using a number of different indirect strategies.

Against this backdrop, this section will showcase how discrimination can be implied below the surface structure of the actual language being used, using examples from the online comments that were analysed following the common C.O.N.T.A.C.T. methodology in the Maltese context. Naturally, the starting point when one talks about implicit meaning is Grice, who famously coined the term implicature to describe meaning that is communicated over and above what is actually said by an utterance (Grice 1975). And even though we will not, in this section, be dealing directly with the notion of implicature—or the Gricean analysis of it for that matter, it should be acknowledged that it could, as a concept, encompass most of the indirect strategies that can be used to express a negative stance towards a minority. Perhaps the most pertinent aspect of implicature in this regard can be found in the post-Gricean, relevance-theoretic reanalysis of implicatures as propositions that can, among other things, “be communicated with varying degrees of strength, depending on the confidence with which a hearer can assume that they form part of the speaker’s informative intention” (Assimakopoulos 2017: 319). So, even though a statement may not constitute hate speech in the eyes of the law, it might still reveal the discriminatory attitude of its producer, to the extent that it could also be considered to be detrimental to the feeling of self-worth of members of a minority group. What is crucial in this respect is, of course, to take into account not only the explicit content of a given statement, but also the particular context in which it has been produced.

One of the most extensively researched categories of implicated meaning is that of irony, which can be easily identified in the example that follows:

• (29) We need to thank the geniuses who agreed with us signing the Dublin 2 convention. They want us to fingerprint immigrants to make it easier for them to identify and deport them back to Malta. ³³

At face value, the comment above has a particularly positive undertone, as it starts off with positively charged words, like ‘thank’ and ‘geniuses’. Yet, when looked at in its particular context, it is clear that it is meant as a negative comment against the Maltese politicians who signed the ‘the Dublin 2 convention’,³⁴ thus agreeing to Malta being solely responsible for the examination of asylum applications by refugees who enter the EU through its territory. The implicitly communicated negative meaning of this comment is derived from the combinatorial meaning of the two sentences it comprises, with the second sentence providing an explanation as to why the first one is intended as an ironical statement. In this second sentence, the user creates a distinction between ‘us’ (=the Maltese) and (the first, exophoric) ‘them’ (=other EU countries) to relay the information that ‘immigrants’³⁵ are unwanted and that the EU is using Malta as a dumping ground for this undesirable group of people. It is therefore evident that the user employs irony to communicate that the signing of the Dublin II convention was an unwise decision that has had negative effects on Malta. And even though the implicature at hand strongly communicates the user’s dismay at a particular political choice, it also carries a weaker negative stance toward migrants, since the irony of the first sentence in combination with the elaboration provided in the second one makes it clear that the user views migrants in Malta unfavourably.

The next example we will be discussing also belongs to the category of irony, but has a more pronounced mocking tone, and can thus be more succinctly described as an instance of sarcasm:

• (30) While the local taxpayer will foot the bill for the police time, court time and legal aid. Way to go! ³⁶

The news article with the headline ‘Libyan conditionally discharged after AWAS row,’ to which this comment was a direct reply reported on a trial in which a man from Libya, who pleaded guilty to the charges of causing material damages to the Agency for the Welfare of Asylum Seekers by breaking the glass surface of a desk when he was not attended to, was conditionally discharged and fined €80. Within this context, the second sentence of the comment in (30), which, when used in a neutral context, is a positive expression similar to ‘well done!’, is produced with a sarcastic tone, showing the user’s negative stance toward the fact that a migrant has cost the Maltese tax-payers money. This becomes clear by looking more closely at the language used in the first sentence of the comment. For one, the commenter uses the expression ‘foot the bill’, which is typically used in situations when the person paying a fee is somewhat unwillingly hoaxed into paying, in the same way a parent might have to ‘foot the bill’ for damages caused by their children. Then, this sentence also begins with the conjunction ‘while,’ which is often used to introduce information that contrasts with the main clause (the ‘main clause’ in this case being the narrative presented in the article, which merely reports the incident). Hence, the first sentence indicates a sort of forcible anchor on the taxpayer’s pocket, which is obviously not viewed favourably by the commenter. Moreover, the user modifies the noun ‘taxpayer’ with the adjective ‘local’, thus further emphasising their dismay that a foreign person has cost the Maltese money. Again, much like in the previous example in (29), even though the user strongly communicates their frustration with a particular state policy by means of this comment, they also weakly show a negative stance toward migrants and the presumed burden they put on the Maltese economy.

Turning to the issue of indirectly discriminating against the LGBTIQ community, the comment in (31) is a good example of a statement that may not constitute prosecutable hate speech in itself, but is strongly discriminatory in nature:

• (31) If adults have unwanted sexual urges or dreams, they should be able to get help. Especially they are married and have kids. Whether those urges are for men or young boys, it shouldn’t matter. Now if a parent sent his gay kid to NARTH, that’s a different story. ³⁷

To begin with, the commenter here does draw the line at parents forcibly sending their children to NARTH (National Association for Research & Therapy of Homosexuality), an organisation that offers gay conversion therapy, thus, recognising the danger that such therapy can bring on LGBTIQ youth. However, the language used in the rest of the comment indicates a strong negative stance toward homosexuality and a profound ignorance of LGBTIQ issues. Firstly, the user refers to non-straight sexual desires as ‘unwanted…urges’. Both these words have negative connotations and imply something unfavourable. Moreover, by way of the modal verb ‘should’, the user offers a suggestion for people with such ‘urges’ to ‘get help’. In doing so, the user implies that heteronormative values are hegemonic and any deviation from them creates an urgent need for the individual to seek help. Finally, the use of the inclusive ‘or’ in the sentence ‘Whether those urges are for men or young boys, it shouldn’t matter.’ strongly implicates that, in the mind of the commenter, homosexuality is on a par with paedophilia and thus warrants ‘treatment.’ In view of all this, then, the final, positive, statement that parents should not force their children to undergo gay conversion therapy is overshadowed by the overall negative stance the user has towards the LGBTIQ community.

Similarly, the comment in (32), which was made in response to a newspaper article about the civil union of a gay couple, might not seem at first to be overly discriminatory. Yet, if we break it down into its component parts and discern the meaning beneath the allusions being used therein, we might form a different opinion.

• (32) people marry because they fall in love, and although it’s a choice, it was meant to be like that even in the animal kingdom, for example swans mate for life, male and female, not male and male. ³⁸

The user that posted this comment may posit the idea that marriage is a choice, but frames the relevant clause with the conjunction ‘although’ (a conventional implicature à la Grice), which is generally used to present two contrastive arguments, thus indicating that even though marriage is a choice for people who fall in love, it is also a choice that comes with restrictions. By bringing in a comparison with mating in the animal kingdom, whereby all swan relationships are described as being heterosexual, the commenter subscribes to heteronormative ideals, implying in this way that any deviation from the heterosexual norm is unnatural. So, even though this comment concedes that marriage is a choice, somewhat echoing the main argument of most gay rights movements on the matter, it still exhibits a negative attitude towards the members of the LGBTIQ community.

Clearly, this short section cannot do justice to the far-reaching implications that that the study of indirectness can have for our understanding of discrimination in language use. What we hope to have achieved through this discussion of some online reactions to news items in the Maltese press is to have justified the need for going beyond the explicitly expressed and overtly communicated meaning when it comes to potentially discriminatory discourse. After all, as van Dijk rightly points out, discourse of any kind “may be seen as a semantic iceberg, of which only a few meanings are expressed on the surface of text and talk, whereas others meanings remain implicit knowledge stored in mental models” (1993: 109) that he elsewhere attributes to “(the users of) a text, [… and not to] the text itself” (2001: 104).³⁹ And given that the most effective weapon against xenophobia and homophobia (alongside many other forms of discrimination) is the promotion of a more inclusive mentality, the identification of such beliefs that can tacitly sway the public opinion towards discrimination is certainly pivotal.

3.6 Changing Participant Roles in the Expression of Hate Speech

• Sharon Millar, Klaus Geyer, Anna Vibeke Lindø and Rasmus Nielsen

Following Goffman (1981) and Levinson (1988), it has been generally acknowledged that there are various production and reception roles in interaction that go beyond the traditional, unnuanced categories of speaker and hearer. In Goffman’s terms, these roles align or position the individual in relation to an utterance, which he terms footing. Production roles may, for instance, be in the form of animator (the person who produces the talk or the text), author (the person who creates what is said or written), relayer (the person who relays the utterances of others) or principal (the person whose position or beliefs are established by the utterance), while reception roles include those of the addressee, bystander and eavesdropper. In this setting, the role of figure refers to the entity being talked or written about. These various interactional roles have been shown to be relevant to dialogically-oriented discursive strategies, such as constructed dialogue (Tannen 2007) and fictive interaction (Pascual and Sandler 2016), which have affinities with the Bakhtinian notions of polyphony and heteroglossia. Constructed dialogue is Tannen’s preferred term for reported speech since such speech is always recontextualised into new discursive contexts, while fictive interaction concerns “the use of the conversation frame to structure cognition, discourse, and grammar” (Pascual and Sandler 2016: 3) and covers phenomena such as talking to oneself, engaging in dialogues with virtual participants, or using rhetorical questions.

One could also argue that the technological affordances of online platforms impact on participant roles. For instance, hyperlinks allow the writer of the comment to relay content, or voices, from other sources, but, since such relayed content is also recontextualised into a new discussion, the resultant participant roles of the various voices can be quite complex. Against this background, we consider here changes in participant roles in relation to both fictive interaction, constructed dialogue and hyperlinks when it comes to the expression of hate in online reactions to news items in Denmark.

We begin with an example of fictive interaction that involves the manipulation of person deixis. The relevant comment (example 33) is in response to an article in the Danish tabloid Ekstra Bladet reporting how a man, who was selling his car online, was fined for writing ‘Fuck you, you Muslim’ to a bidder with an obviously foreign name, who offered what was viewed as an insultingly low price:

• (33) Så dig, som vi her kan kalde Hr F. Muslim, tag dig sammen og prøv at forstå, at selvom vi er yderst tolerante, er vi ikke dumme, dette giver dig ikke ret til at skambyde på en bil eller andet du måtte ønske af os tolerante danskere. Hold op med at lukrere på, at du er muslim, hvilken betydning har det egentligt for os… Betal manden de 5000 kr. tilbage som han har fået i bøde, blot fordi du pipper om din muslimske herkomst og opfører dig som en mand og ikke en kylling. ⁴⁰

  So you, who we here can call Mr. F. Muslim, pull yourself together and try to understand that although we are extremely tolerant, we are not stupid, this does not give you the right to make a disgracefully low offer on a car or anything else you might want from us tolerant Danes. Stop exploiting that you are a Muslim, what meaning does that have for us… Pay the man the 5000 kr back that he got as a fine just because you chirp about your Muslim origins and act like a man and not a chicken.

This comment is structured as interaction by using the 2nd person pronoun ‘you’ and a derogatory term of address, Mr. F Muslim, that plays on and perpetuates the original explicit language of the insult. This combination shifts the role of figure (the character talked about) to the reception role of the (most probably) non-present addressee. Moreover, the writer, who obviously is the author of the discourse, takes a collective perspective and in so doing distributes the role of principal (the person whose views or beliefs are established) across the in-group, ‘we’ Danes. The fictive dialogue then allows the writer to express a decidedly negative group stance towards the individual concerned on the grounds of his religious identity and towards his act of reporting the insult, while simultaneously portraying the in-group in a positive light. Minority identity is constructed as a means by which members of the minority group take unfair advantage of the put-upon in-group, and reporting what could be viewed as a hate incident is associated with cowardice. Hence, what is being legitimised is the discriminatory behaviour of the man selling the car.

The next example illustrates a different type of change in participant role, where the writer of the comment takes on the character of ‘all the immigrants’ in a fictive interaction. This comment is in response to a leader article in the tabloid Ekstra Bladet about a reception being held for people who became Danish citizens during the previous year.

• (34) Stakkels Pensionister .. de får ingen pleje eller støtte i hjemmet, hospitalerne har så få resourcer at de diskutere hvem der har et liv tilbage og som man vil ofre en behandling på. Fedt med alle invandrene som kan finde vej til Danmark og sige: ‘Maj ha penga .. manga penga’ ⁴¹

  Poor Pensioners… they get no care or support in the home, hospitals have so few resources that they discuss who will have a life back and who they will devote treatment to. Awesome with all the immigrants who can find their way to Denmark and say ‘Me have monay…much monay’

In evidence is a form of ventriloquism (cf. Cooren 2012) where the author of the text animates a generalised voice, manipulating pronouns and spelling to convey poor, foreign-accented Danish and, thus, perpetuating the stereotype of immigrants as having poor competences in Danish. At the same time, the content of what is said plays on the prevalent stereotype that immigrants are only interested in receiving welfare benefits. The fictive interaction functions as a means of negatively stereotyping an entire group. However, it also serves to hold this group responsible, since it shifts their role from the figure to the participant roles of both author, i.e. their words, and principal, i.e. their beliefs. Embedded in a sarcastic construction, the implied meaning is that it is far from ‘awesome’ that immigrants come to Denmark demanding money and this interpretation is strengthened by the contextual, and generalised, narrative about lack of resources for geriatric care and hospital treatments.

We will now turn to the role of hyperlinks in the expression of hate speech online. As noted by Klein, hate speech involves “the tactical employment of words, images, and symbols, as well as links, downloads, news threads, conspiracy theories, politics, and even pop culture” (2012: 428). The following comment relates to a newspaper article again from Ekstra Bladet reporting the intention of the Hungarian government to build a fence to keep refugees out.

• (35) “Instant justice” til fup-flygtninge http://m.youtube.com/watch?v=1SCdXHrykfI ⁴²

  “Instant justice” to bogus refugees

This comment is clearly negative, given the categorisation ‘bogus refugees’, but the link is integral to the comment if the exact nature of ‘instant justice’ is to be identified. While the YouTube link was no longer accessible, having been removed for copyright reasons, its original source was still visible: ‘liveleak migrants beaten’. A search on LiveLeak suggests that the video was about ‘migrants beaten by truck drivers in Calais’, and was also labelled as ‘shocking footage’. While LiveLeak makes no mention of ‘instant justice’, compilations of ‘instant justice’ can be found on YouTube, although not necessarily related to any group, suggesting that the phrase has some form of memetic status. At one level, the writer of the comment is relaying the video to others, but through the process of recontextualisation, new meanings for this video have been created or authored by the commenter. S/he has embedded the video in a textual context that legitimises violence towards refugees as justice. In this respect, hyperlink use can arguably be seen as some form of multimodal constructed dialogue.

Our final example deals with issues of sexuality and originates from the national broadsheet, Kristeligt Dagblad, which focuses on issues of faith, ethics and existence. The comment responds to an article about the decision of the US Supreme Court to permit homosexual marriages.

• (36) Indtil to med samme køn reproducerer sig selv, så vil jeg fastholde at homoseksualitet er unaturligt i ordets egentlige betydning. http://www.mx.dk/nyheder/aarhus/story/16276482 ⁴³

  Until two people with the same sex reproduce, I will maintain that homosexuality is unnatural in the literal meaning of the word.

Here, the hyperlink is quasi-transparent, linking to an article in another Danish newspaper Metro Express (21 June 2015), which is about how a politician caused a considerable controversy by deleting from her Facebook profile a question from a well-known homosexual TV presenter, whom she subsequently blocked when he continued to ask this question. This politician is a member of the right-wing Danish Peoples’ Party and had been recently elected to Parliament, while the question at hand related to her previous assertion, in 2013, that all kinds of family life besides the traditional model were unnatural and against the laws of nature. Even though she had back-pedalled from her statements at the time, the TV presenter raised the issue again. The linked article is complex in terms of participant roles, with three voices present: the journalist, the TV presenter and the politician. The journalist has the roles of author and relayer, but since he inserts direct and indirect quotations into a new context as constructed dialogue, the authorial role is paramount. The TV presenter and the politician are both figures in that they are being talked about, but they are also given the role of principal, as through the constructed dialogue their positions and beliefs are established. The TV presenter, however, is the more prevalent figure and principal in this article. Hence, the rhetorical function of the link is ambiguous in a comment that posits the unnaturalness of homosexuality. The commenter’s negative evaluation aligns only with the previous statement of the politician so his focus seems to be on her role as principal regarding this former position. It therefore seems that the relationship between a hyperlink and the comment within which is it embedded may not always be transparent, as it was in example (35) above, especially if the material within the link is itself complex.

To conclude, our data points to the importance of indirect practices in the expression of hate speech as interpreted broadly. The use of fictive interaction, constructed dialogue and hyperlinks can serve to perpetuate stereotypes, and normalise and legitimise xenophobic and homophobic discourse as well as involve the presumed online audience through creative rhetorical strategies.