‘You sick, twisted messes’: The use of argument and reasoning in Islamophobic and anti-Semitic discussions on Facebook

This research used critical discursive psychology to analyse anti-Semitic and Islamophobic discourse on the English Defence League’s (EDL) Facebook page. The discussion by Facebook users began about ‘reopening’ concentration camps, in which to incarcerate Muslims. Facebook users also expressed anti-Semitic discourse such as Holocaust denial, and the idea that Jews ‘could have done more’. The analysis focuses on the reasoning used when expressing this extreme idea, and how this was contested by other Facebook users, through the use of three strategies: (1) the construction of ‘sickness’, (2) Muslims as ‘the new Nazis’, (3) devictimising Jews as victims. This research shows how the EDL used positive aligning with Jews as means to present Muslims as problematic, and how such alignment resulted in the marginalisation of both Jews and Muslims. Findings are considered in terms of how critical discursive psychology can uncover the function of extreme discourse on social media, and the potential implications of hate speech online.


Introduction
On 7 January 2015, two gunmen killed 10 people in the headquarters of the French satirical newspaper Charlie Hebdo in Paris, and a further two people outside of the headquarters. The attack was allegedly motivated by Charlie Hebdo's controversial caricature drawings of Muhammed in 2011 and 2012, which appeared to be mocking Islam (The Telegraph, 2015). Two days later, a kosher supermarket in Paris was subjected to an attack by a lone gunman. Four people were killed in the siege, all of them Jewish (The Independent, 2015). The gunman had pledged allegiance to ISIS and was a close friend of the gunmen in the Charlie Hebdo attack, who were being held by the police at a print shop. More recent attacks by ISIS such as those in Paris in November 2015 1 and the Manchester arena bombing 2 continue to fuel the construction of Muslims as a threat, and Islamophobia has become a significant issue in contemporary British society and elsewhere in Europe, particularly by far-right groups and movements (Fekete, 2011). Islamophobic discourse online has also increased (Allen, 2014), particularly after terrorist attacks (Sadique et al., 2018). This article will use critical discursive psychology (CDP) to address how Islamophobia is articulated using strategies of reasoning, on the social media website Facebook. Research (e.g. Klein and Muis, 2018) shows that the far-right have increased their circulation of Islamophobic discourse in recent years, so the current article aims to critically analyse how this is responded to by users of Facebook. The research will discuss the importance of analysing Islamophobic discourse online, as well as how Jews were used as means to achieve opposition to Muslims.

The Far-right
Historically, the far-right have a dilemma of maintaining nativist nationalist values while simultaneously reaching out to a mainstream audience for support. Billig (1978) argued that one way of managing this dilemma is to disguise extremist ideologies in order to appear more moderate. For example, the National Front adopted an 'anti-Zionist' position rather than 'anti-Jewish', revealing how prejudice discourse is framed as 'cultural' rather than 'biological'. A strategy used by far-right politicians to appear reasonable in anti-Islamic positions is to justify opposition to Muslims based on their apparent intolerances (Verkuyten, 2013).
The English Defence League (EDL) is a far-right street protest movement formed by "Tommy Robinson" (real name Stephen Yaxley-Lennon) in 2009 (Goodwin et al., 2014). The EDL's (2016) mission statement claims that its aims include to educate the British public about Islam, which includes incorporating the promotion of a "balanced depiction of Islam as a religion and ideology". As there is no formal membership or 'joining' process for the EDL, it is difficult to establish exactly how many 'members' or supporters there are (Kassimeris and Jackson, 2015), though Bartlett and Littler (2011) established an approximate membership figure of around 25,000-35,000 members.
The EDL attracts attention and causes controversy by its demonstrations across cities in the United Kingdom (Baker et al., 2008). These demonstrations draw in large numbers and are renowned for being aggressive and violent in nature (Busher, 2013;Oaten, 2014). The EDL argues that the purpose of demonstrations is to campaign against Islamic extremism and not all Muslims (Treadwell, 2012;Treadwell and Garland, 2011); however, this distinction can become blurred in cases where anti-Islamic chants take place (Garland and Treadwell, 2010). Facebook is the EDL's most popular mode of communication (Bartlett and Littler, 2011), and the EDL are renowned for circulating Islamophobic discourse online (Copsey et al., 2013). The social media presence of groups like EDL and Britain First means that like-minded people can now join an online community that allows them to express their Islamophobic views, such as Muslims as a threat to Western culture (Oboler, 2013), while still remaining physically distanced from such groups (Barlow and Awan, 2016;Jacks and Adler, 2015). Bleich (2011Bleich ( : 1581 defines Islamophobia as "indiscriminate negative attitudes or emotions directed at Islam or Muslims". Maussen (2006) argues that Islamophobia is multifaceted and encompasses various forms of discourse, speech and acts, that all emanate from an identical ideological core. Islamophobia has also become more prominent on digital media (Horsti, 2017), along with Islamophobia 'myths' that Islam is misogynistic, violent and a threat to Western liberal values (Kumar, 2012;Törnberg and Törnberg, 2016). Islamophobia has been argued to have become a form of "accepted racism" (Hafez, 2014: 484), and the far-right have increased in popularity across Europe due to the mainstreaming of Islamophobic ideologies (Wodak, 2015a).
The EDL claims that Islamophobia is a rational reaction to Islamic extremism in the United Kingdom (Kassimeris and Jackson, 2015). Similar distinctions between opposing Islamic extremism and not opposing Muslims have been identified in the discourse of politicians such as Geert Wilders in an attempt to appear as reasonable (e.g. Verkuyten, 2005). Researchers argue that the rise in Islamophobia is similar to anti-Semitism in Britain during the First and Second World Wars (Linehan, 2012). Wodak (2015a; see also Hafez, 2014) argues that Islamophobia has somewhat 'replaced' anti-Semitism as a significant contemporary issue in political discourse and that anti-Semitism is no longer seen as a prominent issue, although research suggests that offline events (such as allegations of anti-Semitism in the UK Labour Party) can cause a surge of anti-Semitic discourse on social media (Williams and Burnap, 2018).

Discursive psychology and prejudice
Discursive Psychology is an approach that argues that prejudice is a form of discourse (Edwards and Potter, 1992) rather than the result of faulty cognitive processes as the traditional social psychology theories propose (Ashmore and DelBoca, 1981). Traditional social psychological approaches treat what people say as true representation of what they are 'really' thinking. This notion is criticised by discursive psychologists because such approaches fail to account for the complex interactional work that is going on when people are talking (Wetherell, 1998). Discursive research focuses on how prejudice is articulated and legitimised through discourse, for example, the well-established "norm against prejudice" by Billig (1978: 95; see also Augoustinos and Every, 2007). This is the denial of being prejudiced, while at the same time, making potentially prejudicial arguments.
Discursive research has shown that far-right politicians in Finland use online blogs as a platform to share their nationalist views (Pettersson and Sakki, 2017;Sakki and Pettersson, 2016), as well as how political far-right groups such as Britain First communicate on Facebook regarding Islamophobia (Burke, 2018). Discursive research analysing how Facebook users respond to far-right posts is lacking, with the majority focusing on the discourse of politicians themselves (e.g. Burke, 2018;Goodman and Johnson, 2013;Pettersson and Sakki, 2017;Sakki and Pettersson, 2016). Members of the public have less of a dilemma that Billig (1978) identified of needing to appear reasonable when expressing prejudicial views. Thus, this research aimed to analyse the responses by Facebook users to far-right Facebook posts, and the discursive strategies used by Facebook users when both rebutting and supporting far-right ideologies.

Method
Data were collected from the official Facebook page of the EDL 3 between March 2014 and 2015, as part of a wider project analysing the communication strategies of the farright on Facebook. This part of the project focuses on how the far-right and supporters talk about Muslims during the aftermath of the Charlie Hebdo attack. It was found during this period of time that the far-right used supporting Jews as means to oppose Muslims. The analysis presents comments from one of the Facebook posts in order to examine in detail the strategies of reasoning used when expressing Islamophobic and anti-Semitic views (for other publications on how the far-right convey Islamophobia on Facebook, see Burke, 2017Burke, , 2018; The post of focus is from 28 January 2015, the day after Holocaust Memorial Day and on the 70th anniversary of the liberation of Auschwitz. The EDL shared a video on its Facebook page entitled "70 years After Its liberation, A Drone Captures the Auschwitz we Should Never Forget". The video was produced by the BBC news, and was originally posted on the website 'Israeli video network'. The video is 2:29 minutes long and begins with a drone capturing the view of the railway tracks leading into Auschwitz-Birkenau, and then shows the entrance to Auschwitz. The drone films over the empty concentration camp, with subtitles explaining each feature. The EDL did not post a caption with the video so has not left its own commentary about Auschwitz; nonetheless, its stance is made explicit through sharing the video from Israeli video network. This is similar to the strategies used by other far-right parties to distance themselves from the label that they are fascist (Wood and Finlay, 2008). The Facebook post of the video generated 59 comments in response over a 24-hour period. The extracts presented are from the 59 comments to the video.
CDP was used to analyse data. Discursive psychological approaches argue that language itself should be studied both in use and as part of interaction, focusing particularly on how language is used by speakers to both empower and justify actions (Augoustinos and Every, 2007). This is termed the "action orientation" of talk by Edwards and Potter (1992: 2). Looking at 'action' in talk means what is achieved through talk rather than what this tells us about individuals' attitudes or psychological phenomena. CDP embraces contexts where identities and positions are carried over from other conversations, meaning that analysts consider not only the detail of the discourse, but its broader context (Wetherell andEdley, 1999, see also KhosraviNik, 2017). CDP argues that discourse reflects not only the immediate context in which it takes place, but broader patterns in discourse that reflect global contexts. For example, local expressions of racist discourse may reflect broader patterns of racism in society. This is what Wetherell (1998: 405) referred to as "the social and political consequences of discursive patterning".
What makes CDP 'critical' is that discourse is analysed in wider contexts such as historical or political (see Unger et al., 2016), so is more open to considering the wider cultural understandings of discursive phenomena. CDP focuses on the formation of identities and interaction, and the interactional work that is performed through discursive accounts (Wetherell, 1998). CDP argues that individuals are active in constructing their own identities against larger social contexts. The analysis in this research examined how actions such as Islamophobia were mobilised through discourse, as well as how Muslims and Jews were constructed as 'the other'.

Analysis
The analysis identified three main rhetorical strategies used in reasoning when expressing Islamophobia and anti-Semitism online, each of which will be discussed in turn. First, the construction of opposing the idea as 'sick', as well as how discussion between Facebook users can escalate. Second, the subject position of Muslims as 'the new Nazis', as well as expressing support for Hitler. Finally, 'devictimizing the victim', which encompasses comments displaying Holocaust denial and the idea that Jews were responsible for their own fate.
The structure of the comment thread is as follows: in the first replies to the video, the idea of reopening concentration camps is expressed by several Facebook users. There is immediate conflict between Facebook users over whether the reopening of concentration camps is 'good' or 'bad'. The first reply was a comment in direct response to the video. Following this, fourteen comments were left in response to the first comment, creating a 'micro thread' underneath this first reply. Comments were subsequently left in response to the video, separate from the discussion about reopening concentration camps. Extracts are presented as they appeared on Facebook (with line numbers added), so all spelling and grammatical errors are left intact.
First, we examine the construction of 'sickness' to pathologise 'negative' opinions regarding Muslims. The accusation of another Facebook user being sick is used to say that the view is so extreme. This works to rhetorically frame the accused as not reasonable.

The construction of 'sickness' and 'filth'
Extract 1 The first extract shows an argument escalating following the most popular and replied to comment. P1 starts with the construction of 'reopening' concentration camps to place Muslims in them: that, you clearly support these terrorists! 7. P4: I can be a commander of that facility 8. P5: Maybe you should be put in there too [P1]? 9. P1: Why? I am not a terrorist. . ..unlike many muslims! P2 offers P1 explicit support (2) and escalates the notion being an "excellent idea" as well as three 'thumbs up' emoticons to indicate approval. P1 later downgrades putting Muslims in concentration camps, to putting the "many muslims" (4) who are terrorists in concentration camps. This is an attempt to appear as reasonable, as it shifts the focus onto only the "problematic" Muslims who are terrorists, rather than constructing a generalisation that all Muslims are terrorists. This also constructs P1's position as 'real' and rational.
Conflict over the construction of sickness starts at P3's comment. P3 begins with a rhetorical question (3) to criticise this idea and follows with a two-part construction of a moral category which starts with "sick, twisted", which is then repeated into a two-part subject position of "sick and twisted" by P1 (4). This is an example of how 'sickness' can be constructed as either supporting Muslims or being opposed to Muslims. P5's suggestion of putting P1 into a gas chamber (8) upgrades the disapproval by switching who should be placed into concentration camps. P1 reverses the sickness onto "many muslims" and draws upon the construction that supporting Muslims implies that you are not 'us' but are 'one of them', and that is problematic, emphasised as common knowledge through using "clearly".
The next extract is from later in the same discussion. An accusation of someone being "filthy" is made, seemingly based on their identity as a Muslim (P6, whom the author addresses by name): P7 replies to P6's representation of the Holocaust and presents the notion of Christian terrorist as absurd. P7 positions him/herself as offended by P6's comment based on his identity as a Christian, shown through the use of "my religion". The construction of this idea is presented as so outrageous that P7 is positioned as being on drugs (3) as if this account could not be presented reasonably when of sound mind. Further emphasis of the absurdity of P6's comment is shown through P7 using four question marks (3), although the accusation is softened through P7's use of "mate". P7 constructs P6 as a Muslim (a construction seemingly based on P6's name) and draws upon the 'us and them' distinction (Lynn and Lea, 2003) between Christianity and Islam. P7 uses derogatory language that orients to the Prophet Muhammad's marriage to a six-year-old girl in the form of a three part list (Jefferson, 1990): "filthy", "kiddy shagging", "murderer prophet". This use of the implicit pattern of moral transgressor categories works to establish Islam as 'intentionally' evil. The association of ethnic groups with dirt categorises them in a dehumanised way (Tileagă, 2007). P7 also shows entitlement by accusing P6 of 'daring' to make the comparison, yet inserts their own contrast (Atkinson, 1984) between Islam and Christianity. The use of "dare" also implies a moral transgression, that P6 has crossed a boundary, which P6 is not entitled to do as P7 as categorised his as 'lesser'. This section has examined the construction of 'sickness' in opposing arguments about Muslims, as well as insults that individuals perceived to be Muslims are "filthy". This construction 'others' Muslims and portrays them in a dehumanised category. In the same thread of comments, we see the co-categorisation of Muslims as 'the new Nazis' in addition to being terrorists. On the other side of this construction, we see the notion that Nazism is a positive idea, to enable the argument for reopening of concentration camps for Muslims.

Nazism
The following extract is a continuation of the debate about putting Muslims in concentration camps. The extract is several comments after extract 1, and continue on from the debate over who is and who is not 'twisted' for opening Auschwitz for Muslims: Extract 3 1. P8: Only one problem the Hitler is dead! but the great 2.
idea!! burn all the scum bags in one large oven 3. P9: Twisted would be to tolerate them anymore 4. P10: The islamists are the new Nazis. Simply repatriate 5.
them back to their homelands so they can practise 6.
their voodoo crap and leave the rest of us in peace!
Here, there is further support in the discussion about reopening concentration camps for Muslims, with an upgrade to it being a "great idea". P8 constructs Hitler's death as being the only drawback to this notion of reusing concentration camps. P8 refers to "one large oven" (2), a euphemism used to avoid the laden concept of gas chambers. This has connotations of genocide, and while Muslims are not mentioned, the notion of killing all Muslims is implied. P9 is reversing the previous use of the term 'twisted', (as seen in extract 1) whereby 'twisted' had been to support the idea of reopening concentration camps. P9 constructs the notion that to be twisted means to support Muslims. P9 refers to "them", in a similar way to P8, which implies that Muslims are being referred to and draws upon the common idea that Muslims are intolerable (Verkuyten, 2013).
Next, the subject position in the form of two categories is used, Islamists as "the new Nazis". However, P10 has several strategies to downgrade the hostility that has been identified so far in this discussion. First, P10 refers to "islamists" (4), so the opposition is not generalised towards all Muslims as was seen in extract 1, but now towards Islamists. Second, P10 softens the violence that has been used by previous authors by using the term "repatriate": to send away rather than to kill. However, this does justify prejudice in the form of cultural incompatibility. P10 also constructs Islamists as outsiders to the United Kingdom, which ignores the notion that Islamic extremists have come from Britain before (e.g. the London 2005 bombings, where three of the four suicide bombers were British born 4 ). Finally, P10 uses the idea of practising "Voodoo" which positions Islam as a dark, violent cult rather than a religion, in contrast with the 'peaceful us', and emphasises Islam as something absurd. The pairing of "Voodoo" with a swear word is used to mark the construction as explicitly racist, a strategy identified by Stokoe and Edwards (2007). This is a strategy to position Islamists as the violent ones, rather than P10. Despite Muslims being 'the new Nazis', and thus the incarcerators, they are still positioned as ideal to be prisoners in concentration camps and are simultaneously being categorised as two subject positions.
At the same time as the 'micro thread' of comments about reopening concentration camps was taking place, Facebook users were commenting on the video discussing the idea that Muslims are the New Nazis: Extract 4 1. P11: Muslims were with the Nazis, they have taken over 2.
from him, 2nd Holocaust happening right now all over 3.
being murdered in the name of Allah, Islam, how long 5.
will it take for the non-Muslims to wake up and 6.
realise what's happening, or are we waiting for 7.
another 6 million to be killed 8. P12: i agree with you [P11], it seems as though the world 9.
is sleeping, hope it wakes up before its to late P11 draws upon the metaphor of being asleep (5), to construct the idea that people need to face up to the ongoing global problem of Muslims, and that it does not take a lot to see the 'truth' of the matter to "wake up". P11 adds to the construction of Muslims as 'the new Nazis' by referring to a "2nd Holocaust" and as Muslims taking over from Hitler. This is placed as common knowledge in two ways: (1) through referring to Hitler as "him" rather than by name, and the reference to "another 6 million" -how many Jews were killed in the Holocaust (Cohen-Almagor, 2016). In addition to this, P11 constructs Muslims as historically aligned with the Nazis and expresses who is being killed "right now" by Muslims, constructing a current battle of good against evil. This in turn draws upon the idea that nothing is being done about it. This is used to present the identities and religions that P11 aligns with, as well as construct Muslims as powerful and harmful. Finally, P11 draws upon the statistic of how many Jews died during the Holocaust, as an implicit threat that the same will happen again. P11's parallel between the Holocaust and the present day makes moralising and reifying claims about the 'evils' of Islam. P12 directly aligns with P11's account through expressing explicit agreement and inserts an additional opinion using the same metaphor of the world being asleep, and again uses implicit threats, although somewhat softer than P11's threat in the form of 'hoping' that this will not happen.
This section focused on the construction of Muslims and Islamists as 'the new Nazis'. Resorting to talk of Nazism in arguments about Muslims has been recognised as a rhetorical device; labelling Muslims as Nazis and thus culturally recognisable as evil. In the final section, we examine a common feature of anti-Semitic discourse; Holocaust denial. An extreme form of anti-Semitism, Holocaust denial is the notion that the historical account of the Holocaust is either exaggerated or incorrect. For example, denial of the extent of the mass murders (Griffin, 2015). In some of these accounts, Holocaust denial is implied rather than explicitly denied, such as devictimising Jews as being responsible for their own fate and orienting to the idea that 'it wasn't just Jews' who suffered.

Devictimising the victim (Holocaust denial)
The first extract displays a common strategy (Griffin, 2015) in Holocaust denial; denial over the scale of murders that took place: Extract 5 1. P13: Oh its 1.6million now, when will the jews decide on 2.
which lies are their official story. 3. P14: Just in that camp. . ..

P15: Shame on all Jew Haters.
Here P13 orients to the caption in the video which states "An estimated 1.6 million people were killed at Auschwitz". P13 constructs the figure of 1.6 million as being a lie (possibly confusing this with the well-known figure of six million Jews who were murdered during the Holocaust). The argument that the figure is a lie is a commonly used trope of anti-Semitic conspiracy theories (Byford and Billig, 2001). P13 displays the common Holocaust denial of the scale of people murdered (Wodak, 2015b), and orients to the idea that this figure frequently changes. "Oh" is a change of state token (Heritage, 1984) used to indicate that P13 has received new information, and there is now a change in what is common knowledge. Treating it as a matter of surprise also treats it as a matter of contestation. This is further shown through using the word "now", implying that the figure of people killed has changed. This also constructs a portrayal of Jews as being inconsistent.
P14 constructs P13's account as that of confusion over the 1.6 million being how many people were killed in the Holocaust, rather than killed in Auschwitz alone. P14 corrects the first author's 'mistake'. P14 constructing the figure of 1.6 million as common knowledge and using dots at the end of his or her statement indicates that nothing more needs to be said. This could be a strategy to imply that this fact is obvious, or avoidance of directly saying that P13 has oriented to Holocaust denial. The term "Jew Hater" is used by P15 as a way of calling P13 anti-Semitic.
The next extract devictimises Jews by downplaying the impact of the Holocaust on Jews: Extract 6 1. P16: Its a lot less than what has been done an still 2.
going on on the Genocide of the Americas of 1 st 3.
Nations by the invaders and Ii am sick to death 4.
of jews crying about a few yrs of the 2 nd WW our 5.
Genocide started in 1490,s to the present with 6.
Only 4 million Fullbloods left and we have fought 7. For our lives over this time, jews didn't they just 8. did what they were told 50 guards could guard 9.
thousands easily walking or in trains to there 10. deaths was it stupidity or no guts roll over an die 11.
instead of fighting back maybe a few hundred would 12.
die over powering guards but thousands would have 13.
survived stop crying jews there are people's always 14.
when they're as stupid as his.
Strategies such as the construction of Jews as exaggerating the size of the effects of the Holocaust are manifestations of Holocaust denial (Cohen-Almagor, 2016). P16 draws upon the genocide of Native Americans and uses the figure of how many "Fullbloods" Native Americans are left, in order to contrast this with the Holocaust and challenge the Jewish suffering. This account downplays both the length of the Holocaust and the impact upon Jews through using normalising terms such as "few years" and "crying". Jews are also blamed for their fate, for "just" doing what they were told, the use of "just" constructs the lack of 'fighting back' as self-evidently wrong (Goodman and Burke, 2010). The size of the effects of the Holocaust are minimised by comparing them with the fate of other groups in order to undermine Jews. Jews are devictimised by being positioned as in control of their own destiny and of not taking control of something easily controllable. This places the blame of the Holocaust onto Jews and positions Jews as weak. P16 constructs his theory as reasonable by invoking a utilitarian argument that at least "a few hundred" would still have died during the Holocaust, in contrast with the "thousands" who would have survived. This means that P16 does not deny the Holocaust altogether in order to appear as rational, but by portraying its occurrence due to a lack of Jewish resistance constructs an implication of who are 'really' accountable.
There is an 'us and them' distinction used, the idea that there are people worse off than Jews. The author highlights the plight of 'us', a form of differentiating the self (Lynn and Lea, 2003). This strategy allows P16 to show concern for people belonging to his own 'us' group, native "Fullbloods", while maintaining an anti-Semitic argument. This constructs the 'us' group as having worse hardship. This also allows P16 to refute the idea that anti-Semitism is bigoted because he or she has a more 'reasonable' concern.
In response, the "shame on you Jew Hater" comment is used again, this time towards P16 directly. P5 (who has posted before, see extract 1) responds and accepts the first author's view and invokes the freedom of speech argument to present him/herself as reasonable and tolerant to differing views. The support for freedom of speech then allows P5 to make his subsequent insulting comment, utilising the common discursive strategy of linking opposing arguments with a lack of intelligence (Burke and Goodman, 2012).
The final section has examined the notion that victims are responsible for their own fate, how the victim has become devictimised. Strategies include Holocaust denial, undermining the effects of the Holocaust on Jews, and the idea that Jews 'could have done more'.

Discussion
This article has analysed the responses on Facebook to the video about the 70th anniversary of the liberation of Auschwitz, which occurred during the aftermath of the Charlie Hebdo attack, and the Hypercacher kosher supermarket siege which occurred 2 days later. Facebook users displayed anti-Semitic and Islamophobic views shown through three ways of reasoning strategies. First, conflict between users showing support for reopening concentration camps for Muslims and those opposing this idea, with a focus on who is 'sick', as well as Muslims being positioned as 'filthy'. Second, the co-construction of Muslims as the 'new Nazis'. Finally, comments displaying Holocaust denial and devictimising Jews. This denial involved denial of three types: (1) the use of gas chambers; (2) denial over the scale of people murdered; and (3) orienting to the notion that 'it wasn't just Jews', which, rather than drawing upon other victims of the Holocaust, uses Native Americans as an example. Within this strategy was placing blame onto Jews for their fate, as they 'could have done more'.
In each strategy, there is conflict and escalation. Logic and reasoning are used on both sides of a debate, even when Facebook users display violent, visceral, expressions of hatred. This has been shown to be a common strategy between two groups in conflict, whereby members try to persuade the opposing side that their position is a reasonable one (Finlay, 2014). The use of reasoning has also been found to be a strategy used by politicians to appear as rational when making potentially racist arguments over immigration (e.g. Goodman and Johnson, 2013). The application of CDP to social media discourse has shown that the EDL focused on the apparent tackling of anti-Semitism in order to win support from Jews. However, what the far-right initially presented on their Facebook pages as at least implicitly pro-Jewish (without initially mentioning Muslims), has taken a transition by Facebook users to not only anti-Islamic, but also anti-Semitic rhetoric. The absence of moderation and intervention from the EDL in these types of online discussions is an indication of the EDL's anti-Islam ideology.
Concentration camps as a negative phenomenon is displayed as common knowledge, but nonetheless constructed as suitable to be reopened for Muslims, with logic and reasoning applied to this extreme idea. The construction of bringing back gas chambers has previously been identified on Facebook discussions about asylum seekers (Burke and Goodman, 2012). Facebook users oriented to common Jewish conspiracy theories such as disputing the number of Jews murdered at Auschwitz (Cohn, 1967;Richardson, 2013). The effects of the Holocaust can be belittled; a manifestation of anti-Semitic discourse. Holocaust denial is a criminal offence in some European countries such as Austria, Germany and Poland. Thus, as Wodak (2015b) has highlighted, speakers commonly imply Holocaust denial rather than outright deny it, to avoid persecution. Yet the statements seen in this analysis are left on Facebook unreported, despite procedures in place for Facebook users to report hate speech on Facebook (Oboler, 2013). Future research could address how such discourse is monitored on Facebook and whether similar strategies are observed on other social media websites.
Online hate speech is easier to articulate in comparison to offline hate incidences, due to the lack of physical presence, being cheap and instant (Brown, 2018). While researchers argued that Islamophobia has somewhat 'replaced' anti-Semitism (Wodak, 2015a), this research supports the argument that anti-Semitism is still an important issue in contemporary discourse (Klikauer, 2018), in this case on social media, as well as how it intermingles with anti-Islamic ideology. Facebook users constructed the reopening concentration camps for Muslims, 'the new Nazis', as not only a rational idea, but an "excellent" idea. The notion of Islam being linked to Nazis has been identified as a strategy used by far-right parties to avoid being labelled as anti-Semitic (Hafez, 2014).
Previous research has shown that far-right parties need to be seen as being reasonable when promoting potentially prejudicial views (e.g. Billig, 1978;Burke, 2017;Goodman and Johnson, 2013). One can observe how some commenters are cautious to appear to oppose Islam as an ideology and not Muslims as individuals (see Wood and Finlay, 2008). However, in this research, Facebook users used the terms Islamic extremists and Muslims interchangeably and synonymously. Such discourse has led to some researchers arguing that there needs to be a distinction between the definitions 'Islamophobia' and 'Muslimophobia' (Cheng, 2015), or suggest a definition such as 'Anti-Muslim prejudice' that is set within concepts of cultural racism and racialisation (Meer and Modood, 2009).
Users on Facebook used reason in their arguments of getting rid of Muslims, while the far-right are maintaining that they are not opposed to Muslims. This suggests that farright supporters can express on social media what far-right organisations and politicians cannot on other platforms. That is, the underlying, more extreme, messages. The farright cannot express such extreme messages due to the aim of maintaining a portrayal of rationality and appeal to the mainstream. Social media discourse can reinforce the binary opposition of 'us and them', and the 'othering' of Muslims and Jews.

Conclusion
This research has been important for identifying how, in the aftermath of a terrorist attack, groups can be 'othered' as not 'one of us' and marginalised by far-right supporters on social media. The analysis has shown how one religious group (Jews) can be used as means to achieve opposition to another religious group (Muslims). Even in implicit defence of one group -in this case Jews -the extreme discourse can easily turn on them. What has emerged from the discourse of far-right Facebook pages is that it is not only Islamic extremists being targeted, as the far-right try to uphold, but ordinary, everyday, Muslims through the blurring of the two categories. Users employ reasoning in these discussions over reopening concentration camps and putting a twist on what it means to "never forget".

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.