‘It’s Not a Race, It’s a Religion’: Denial of Anti-Muslim Racism in Online Discourses

: This article investigates the denial of anti-Muslim racism in online discourses. It does so by examining Facebook posts responding to a bystander anti-racism video about a Muslim woman. Particularly salient on social media is the lack of rules, etiquette or social taboos around racism controlling what people post and how they express themselves. This allows comments that are blatantly racist and antagonistic rather than concealed and subtle as is more socially acceptable in ofﬂine spaces. Using critical discourse analysis, the article will delve into the rhetorical and linguistic strategies the posters use to deny that racism toward Muslims can exist. It will expose how the denial of anti-Muslim racism is used in attempts to silence Muslims and anti-racists as well as to convince the general public that Muslims deserve the ill treatment they receive. However, exposing the strategies of racism deniers gives us a better understanding of how to resist such discourses.


Introduction
Denial of racism is a main feature of modern racism, both at the individual and political level (Nelson 2015). Denying the existence of racism is strongly tied with the use of disclaimers, which are obvious signs of underlying prejudices or antagonistic attitudes, if not indications of racism itself ( van Dijk 1993). Since racism has been outcasted as something in the realm of the irrational (Billig 1988), it makes sense that those with racist agendas will attempt to conceal their racist attitudes by denying they, or someone else, is racist.
Studying denial of racism becomes both more simple and more complex when it comes to prejudice towards Muslims. Denial of anti-Muslim racism can involve the simplistic trope 'Muslims are not a race', leaving space for Muslims to be verbally or physically attacked with impunity. However, deconstructing the speaker's argumentation and logic behind these denials is more complicated due to the different ways the speakers discursively construct Muslims, the concept of racism and the relationship between the two. This article investigates how anti-Muslim racism and its denial play out in online discourses. Recent studies show that people show less inhibition when it comes to expressing racist and discriminatory thoughts online in comparison to offline discussions (see, e.g., Coffey and Woolworth 2004;Steinfeldt et al. 2010). This study thus furthers our understanding of expressions of anti-Muslim racism and its denial in anonymous, online environments free of the constraints of social taboos and etiquette that would normally control how speakers moderate their language in offline settings.

Background
In 2015 the 'Challenging Racism Project' research team at Western Sydney University (WSU) in Australia commissioned four bystander anti-racism videos. The videos each show a scenario in which one person is a victim of racist discrimination, abuse or harassment. Each video shows two versions of the same event: in one version, the bystanders say and do nothing. In the second version, bystanders confront the perpetrator, report the perpetrator and/or defend the victim.
This article concerns one video which depicts several passengers harassing and mocking a Muslim woman on a Sydney train due to her wearing a headscarf 1 . In the first version of the video, the other passengers, the bystanders, do nothing and the Muslim woman hurries fearfully out of the carriage to escape the verbal abuse and mocking of the perpetrators. In the second version, one man speaks up to tell the perpetrators to 'leave her alone' after which the other bystanders also speak up to defend the Muslim woman. The perpetrators decide to leave the carriage and the Muslim woman thanks the bystanders.
After the videos were posted on the university's Facebook page, viewers left thousands of comments in response. The comments provide insight into the viewers' thoughts on one premise of the video: that is, that the Muslim woman was a victim of racism.
Unlike the other three videos, where the ethnicity or race of the victim was the reason for the victimisation, in this video, it appears it is the woman's religion that is the basis for the perpetrators' verbal attacks. In many cases, the commentators argued that it is not racism because 'Islam/Muslim is not a race'. However, as will be shown, Muslims are racialised as 'others' with the headscarf simply identifying the wearer as Muslim. This article will thus examine the online comments on the video and how the commentators justify their denial that what happened to the woman was anti-Muslim racism.
Since the video was posted, Islamophobia has been an ongoing problem. The Christchurch mosque shootings, in which 51 people were killed and 40 injured, shows the extreme consequences of people's vitriolic attitudes towards Muslims. Iner (2022) has also found that Islamophobic incidences continue to happen to Muslims in everyday settings, especially hijab-wearing women. In addition, online Islamophobia is also 'of concern, given its easy and speedy reproduction and widespread and long-lasting distribution in the absence of strict monitoring and prevention of online hate' (Iner 2022, p. 8) Reports of online Islamophobia incidents to the Islamophobia Register increased 18 fold in the two weeks after the Christchurch attacks (Iner 2022). It does not appear that attacks on Muslims will decrease or cease any time soon, either offline or online. This article thus investigates these online attacks against Muslims, not only to better understand hatred toward Muslims in general, but to deconstruct the online posters' discursive strategies of denial and their justification of the ill-treatment and prejudice toward Muslims

Offline and Online Discourses
From the 1980s to the early 2000s, scholarly works on race and racism began discussing the social taboo against making openly racist remarks against outgroups. This taboo stems from the advent of 'new' or 'modern' racism where culture instead of biology was used to explain a negative hierarchisation of outgroups (see, e.g., Barker 1982) and involves a denial of being racist . Being racist is inextricably tied with being prejudiced which has been tied to irrationality (Billig 1988). Thus, those who wish to express negative news about outgroups have to be careful to construct these views as justified, warranted or preferably, not even negative at all (Rapley 1998). Denying and mitigating prejudiced views towards ethnic minorities serves to present oneself as reasonable and 'not racist' (Condor et al. 2006).
During those decades, capturing racist talk for purposes of research and analysis was difficult (Stokoe and Edwards 2007, p. 341). Nowadays, the ability for people to make comments on social media anonymously allows researchers to gather types of data that would not have been publicly available or easily accessible to researchers before social media came into existence. Technological advances have allowed people to communicate through multiple electronic platforms (Cleland 2014). The existence of digitally enabled communication such as online forums and social media sites such as Facebook, Twitter and Instagram have enabled research to be conducted on these platforms. In more recent times, the discursive requirement of 'political correctness' is commonly suggested not to be applicable online (Malmqvist 2015). The Internet has allowed racist messages to be communicated, not only by text, static images and symbols, but also by downloadable videos, music, and interactive online games (Bliuc et al. 2018). Etiquette that would be expected in face-to-face interaction may be quite different in interactions via the Internet (Vasquez 2021). This is due to 'factors such as anonymity, invisibility and minimisation of authority online' (Matley 2018, p. 37). Coffey and Woolworth (2004, p. 10) argue that while communication in person tends to see more restraint and often, a 'tempering of words to make points', 'the anonymity the Internet affords gives prejudiced people license to publicly express racial attitudes'. Online forum comments have allowed people to perpetuate stereotypes and express overtly racist attitudes (Steinfeldt et al. 2010). Individuals can 'use online channels to validate their beliefs and achieve a sense of belonging to (virtual) communities of like-minded supporters of racism' (Bliuc et al. 2018, p. 84). The more extreme type of behaviour in online environments stems arguably from the 'online disinhibition effect', whereby 'psychological restraints that serve to block or conceal emotions and undisclosed needs are found to be lowered in cyberspace in various online personal behaviours' (Lapidot-Lefler and Barak 2012, p. 434). Users of social media frequently employ the phrase 'freedom of speech' to defend the right to voice their own opinions, while they simultaneously silence others on social media platforms (Lim 2017, p. 420).
There is growing evidence that online platforms allow racist comments and abuse to flourish. For example, in Iner's (2022) study encompassing online Islamophobia after the Christchurch attacks, she found people posted extreme content online that most likely would not be acceptable in person. This content applauded the Christchurch terrorist, expressed joy for the attacks via insensitive jokes, sought more killings of Muslims and shifted the blame for the attacks on Muslims (Iner 2022). Ozduzen et al. (2021, p. 3350) argue that 'the digital space serves as a ventriloquist for the obscene that cannot be said in person'. One major finding is that biological racism, based on biological heredity and superiority/inferiority, is making a comeback through online platforms (Ozduzen et al. 2021). For example, promotions of 'ethnic cleansing' and fearmongering about 'White Europe' being threatened have been identified on French website discourse (Froio 2018). Some websites even endorse the Islamo-substitution conspiracy theory whereby Muslims are purported to be replacing the autochthonous French population (Froio 2018). Lentin (2018, p. 401) argues that the 'denial and redefinition of racism . . . has become a central formulation for the expression and legitimation of racism'. In everyday offline settings, Augoustinos and Every (2007) argue that in denials of racism, racial views are carefully justified and constructed as balanced and rational. Presenting oneself as 'evenhanded' and 'balanced' are important discursive tools to mitigate against accusations of racism, while simultaneously downplaying the extent of racism (Augoustinos et al. 1999). The dominant conception of racism stems from a Eurocentric formulation that is unable to capture the individual lived experienced of those at the receiving end (Lentin 2020). This formulation relies on historical understandings of racism based on pure biological racism from the past, such as in the case of the Holocaust, Apartheid and Jim Crow segregation (Hesse 2011). Thus the denial that something is 'not racism' often originates from this misconception (Lentin 2020). Understanding racism as only the more extreme biological racism of the past conveniently allows people to declare their views as 'not racist' but rather, rational and justified.

Denial of Racism
Many White people follow a double strategy of positive self-presentation, on the one hand, and a strategy of expressing subtle, indirect or sometimes more blatant forms of negative other-presentation, on the other hand ( van Dijk 1992). Indeed, especially in public discourse, outgroup derogation seldom takes place without expressions of ingroup favouritism or social face-keeping ( van Dijk 1992). The denial of prejudice is strategic and is much more designed 'to make a good impression' rather than being 'sincere and honest'. The positive self-presentation often takes the form of disclaimers to mitigate the racist statements that will follow. van Dijk (1997) argues that for politicians, . . . it is necessary to make sure that such negative talk and cognitions are not perceived as biased or prejudiced, let alone racist. Closely related to the moves of positive 'I' presentation are the usual disclaimers in which speakers deny that they are racist or otherwise biased: "We have nothing against immigrants [or minorities], but . . . " This appears to hold true even in the more contemporary climate with the rise of far-right politicians, political parties and organisations. Even the far-right fascist British National Party sought to justify their claims about Islam and Muslims by quoting from the Qu'ran in a leaflet they published entitled 'The Truth about I.S.L.A.M. 2 ' (Allen 2010). The party even recruited a small number of fringe Sikh and Hindu groups to argue they were presenting an 'insider' view (Allen 2010). Similarly, an analysis of Pauline Hanson's 2016 maiden speech to the Australian Senate shows that she constructed her views as factual, objective and reasonable (Sengul 2019). She presented herself as tolerant of immigrant groups other than Muslims to legitimise her attacks on Islam (Sengul 2019). Thus no matter how extreme the view is, people will continue to justify themselves and deny that they are racist.
While White supremacist discourse is re-entering mainstream political discourse and popular culture, those from the alt-right still attempt to make their extremist racist views palatable by, for example, using racial stereotypes under the guise of humour to deny literal claims (Paul 2021). The alt-right also adopt memes to justify their toxic discourse as parody, which can mask problematic content as a joke (Lamerichs et al. 2018). Even online hate groups attempt to legitimise their racist ideologies to win more supporters-authors of Stormfront, credited as the first hate website, cite credible publications such as the Wall Street Journal, New York Times and the Encyclopedia Britannica to frame White supremacist doctrine as well-researched, valid sources of information (Meddaugh and Kay 2009). Even those spouting extreme, racist discourses will attempt to conceal anything perceived to be irrational and/or problematic, if only to appeal to a wider audience and thereby have stronger social influence.
A different way of denying the existence of racism is to disparage anti-racism efforts or to depict anti-racist education as something contravening the norm or a form of racism in itself. This adheres to the 'colour-blind racism' concept that Bonilla-Silva (2003) developed. This concept is prevalent in societies that are seemingly free of differential treatment based on race or colour but actually ignores the fact that race and colour still play a role in determining the life chances of people of colour (Bonilla-Silva 2003). Colour-blind racism is clearly evident in the Black Lives Matter movement in which accusations of racism (against Whites) were abundant and the All Lives Matter movement arose as a response. On Twitter, those refusing to look at issues of racism replace the rhetoric of #BlackLivesMatter with #AllLivesMatter, claiming impartiality and even superiority (Carney 2016). However, while the claim that all human life is valuable is not wrong per se, it intentionally erases the complexities of race, class, gender, and sexuality and supports a colour-blind ideology that favours White supremacy (Carney 2016). Disparaging anti-racism efforts appears to be a proxy for normalising racism toward non-Whites and denying such racism exists since it is the 'norm', where any challenges to this 'norm' are themselves seen as discriminatory and damaging.

Anti-Muslim Racism and Its Denial
Muslims do not constitute a 'race' by nineteenth-century 'scientific' definitions (see, e.g., Miles and Brown 2003). However, Hochman (2017) argues that 'racialised groups' are groups that have been misunderstood to be biological races and are very much real. 'Muslim' has now acquired meaning beyond religion and now also describe a racial category: those whose ancestry goes back to Islamic countries (Gotanda 2011). Meer and Modood (2009, p. 344) argue that 'we should guard against the characterization of racism as a form of 'inherentism' or 'biological determinism', which leaves little space to conceive the ways in which cultural racism draws on physical appearance as one marker, among others, but is not solely premised on conceptions of biology in a way that ignores religion, culture and so forth'. The essentialisation associated with biological racism does not necessarily apply here, as culture can be both essentialised and de-essentialised for illiberal and oppressive purposes (Verkuyten 2003). For example, one can demand that Muslim women stop wearing headscarves due to a belief that headscarves belong to an inferior and misogynist culture. Carr and Haynes (2015) argue that the term 'racism' is necessary to recognise how Muslims' outgroup status and oppression is constructed, legitimated and lived.
Conversely, Roy (2009, pp. 8-9) asserts that faith communities should be recognised on the basis of an individual and free choice because while racial and ethnic identity are ascribed from the outside, 'faith is a choice'. However, the 'choice' argument is a thinly veiled disguise to package racist attacks on minorities in a more palatable way that does not refer to race (Loke 2012). This leads to a dangerous slippery slope of expecting religious adherents to stop acting or appearing ostensibly religious to prevent discrimination and harrassment against themselves. In the video, which is the subject of the article, one could claim that the Muslim woman is responsible for the harrassment because she could choose to take off her headscarf if she does not want to be harrassed for wearing it. Arguments that one can hide or change one's religion puts the onus on Muslims to change aspects of themselves or forego certain beliefs rather than the onus being on the perpetrator to forego his or her prejudiced views (Cheng 2015).
Furthermore, faith can be ascribed from the outside: anti-Muslim hate crimes have not been limited to Muslims, as people can be targeted for simply 'looking Muslim' (Considine 2017). There have been several tragic high profile incidences where immigrants from South Asia have been attacked or killed in the US because they 'looked Muslim' regardless of their actual religious background (Mishra 2017). Sikhs have been attacked due to prejudice against the 'Muslim-looking other' (Jhutti-Joha and Singh 2020). Thus claiming that prejudice against Muslims cannot be racism is a form of denial that any wrongdoing against Muslims-or other visible minorities-has occurred.
The 'culture-as-natural-difference' discourse is utilised to deny what one has said is racist because one has not mentioned race or racial difference (Every and Augoustinos 2007, p. 414). This is clearly the case for the denial of anti-Muslim racism, which often includes references to cultural incompatibility between Islamic and Western cultures. As Considine (2017, p. 8) argues, 'racism, in the context of cultural orientation or imagined cultures, involves prejudice or discrimination against individuals because of their perceived cultural preferences'. In Finland, for example, politicians exploit cultural differences to warrant their anti-Muslim views, such as declaring that 'Islamists do not accept the festivities of us infidels' and defending this as simply expressing one's opinion (Pettersson 2020, p. 44). Another strategy of denial politicians use is to frame their anti-Muslim hate speech comments as factual and common sense, for instance describing problems allegedly caused by Muslims as 'true problems', which serve to deny that they have any racist intentions (Pettersson 2020).

'Reverse Racism'and White Privilege
The concept of White victimhood arose during the 1960s in the US and UK as an opposition to race-related social policy (Nelson et al. 2018). In Australia, it manifested itself in the 1980s and 1990s with a growing opposition to social policy intended to benefit Indigenous Australians (Hatchell 2004), new migrants and asylum seekers (Sharples and Blair 2021), with White Australians believing those groups receive more benefits and privileges than themselves. Claims of anti-White 'racism' (or 'reverse racism') became prominent in public discourse during this time.
Claims that White Australians are being discriminated against stem from 'White paranoia' (Hage 2003) that migrants are disrupting the dominance of White Australians and the belief that White Australians are the true owners of national space and identity (Sharples and Blair 2021). A sense of 'White victimhood' arises from the beliefs that Whites are downtrodden and neglected (Amin 2002). In addition, Whites believe the 'special privileges' given to Indigenous Australians are 'unfair' because they contravene liberal-democratic principles of merit equality (Dufty 2009). In the US, one reason why Whites are keen to highlight to existence of reverse racism is because some see racism as a zerosum game where lower levels of anti-Black bias are considered linked to higher levels of anti-White bias (Norton and Sommers 2011). While the US context in which Norton and Sommers' study took place is different from the Australian context, it explains why some Whites perceive measures to increase equality to be at their expense (Norton and Sommers 2011). Far-right Australian politician Pauline Hanson's parliamentary motion 'It's ok to be white' in 2018 represents one of the clearest examples of this kind of White supremacist/alt-right discourse (Sengul 2021). Hanson spoke of 'the deplorable rise of anti-white racism and attacks on Western civilisation' (Karp 2018). Through her 'reverse racism' discourse, Hanson has not only denied that she is racist but also denied negatively racialised people the right to define what constitutes racism (Sengul 2021).
Those who reported witnessing and experiencing anti-White racism fail to acknowledge structural, White privilege (Nelson et al. 2018). The saying 'for those accustomed to privilege, equality feels like oppression' is apt here. Those claiming to be victims of anti-White racism are likely to be ignorant of how oppression and marginalisation of non-White Australians occur on a daily basis, rooted in the colonial dispossession of Indigenous people and exclusion of non-White migrants from Australian national identity (Nelson et al. 2018). Tollefson and Yamagami (2008) argue that 'reverse racism' is a non-sequitur, designed to maintain existing social hierarchies and discourage active measures to combat social inequality. If we consider Wodak and Reisigl's (2015) conceptualisation of racism, racism also contains the process of economic exploitation and social and political exclusion. It is only possible to execute this exploitation and exclusion if one is in a position of power over others. As Due (2011) argues: . . . racism requires both prejudice towards a group of people based on the social construction of race and the power to oppress those groups of people. Thus it does not make sense-at least within Australia as a country with a history of colonisation-to discuss racism from marginalised groups towards dominant ones. This definition is important in that, whilst acknowledging that individual acts of racism can and do occur, a broader definition of racism involves discourse which functions to further marginalise groups who are already disadvantaged.

Methodology
In December 2015, four bystander anti-racism videos were posted on the Challenging Racism Project's profiles across the social media platforms Facebook, Twitter and YouTube. This study seeks to analyse the comments made in response to one of these videos on Facebook. The video was produced as part of a broader project that sought to examine bystander anti-racism and develop anti-racism campaign materials to encourage bystander intervention. The video is based on a real-life scenario set in a suburban Sydney train, in which the 'target', a Muslim-Australian woman wearing a hijab, is approached by three perpetrators who proceed to harass her and pull at her hijab, asking her sarcastically where she got her scarf or her 'tea towel' from.
The video was posted onto Facebook twice. The first time was as a generic post on the WSU Facebook page on the 15 December 2015 accompanied by the text: Did you know that almost 40% of racist incidents occur in public spaces, including on public transport? Today our Challenging Racism Project launch a series of Bystander Anti-Racism videos the help the public #speakupspeakout. Find out more at westernsydney.edu.au/challengingracismproject The second was as an advertisement on the WSU Facebook page on 18 December 2015. As the WSU Facebook page is public, as opposed to a private Facebook group that is only accessible to members (British Psychological Society 2021), or a personal profile page (Zimmer 2010), the comments on the video are thus considered public, as the posters can reasonably expect to be observed by strangers (British Psychological Society 2021).
Comments made in response to the video were collected in three different ways. Firstly, responses were manually retrieved from the video comment section on Facebook as they were posted over a three week period between 16 December 2015 and 6 January 2016. A second set of responses were archived due to their offensive nature and recovered by WSU from the same time period. A third set of data was retrieved using R in December 2016, allowing the collection of new responses made since the videos were originally posted.
The three sets of data were collated, with duplicate responses, due to the multiple data sets, removed. This resulted in a data set of 4843 unique responses, that is, Facebook comments. Validity checks were conducted on the data, differentiating the two video sources, the original post and the advertisement, as well as comments made as replies within each data set. A second validity check was also conducted to remove comments written in languages other than English, comments made by automated bots, comments that consisted only of 'tags' in which merely the name of another Facebook user would appear, comments comprised solely of emoticons, links and memes and any invalid text (i.e., gibberish or text that was otherwise incomprehensible). This resulted in 3497 comments being potentially suitable for analysis. Comments that were posted in response to other comments made and could not be deciphered on their own were not included in the coding for the purposes of this study, for example, comments such as 'Hahahaha', 'Oh', or 'Agreed 100%'.
The majority of comments had a negative opinion of the video with many anti-Muslim and anti-Islam sentiments being shared. Altogether, there were only approximately 300 comments that were considered to directly support the video or the messages being imparted. For the most part, the posters used the video as a platform to air prejudiced and hostile views about Muslims and Islam, or about immigrants or other minority groups.
Since this article is premised on the belief that the denial of racism is a key element of the expression and legitimation of racism today (Lentin 2018), the excerpts chosen for an in-depth analysis were those whereby the posters follow the 'Muslims/Islam are not a race but a religion' argument to deny that the perpetrators' actions on the train were racist. In addition, during this process, a disturbing denial of a different kind that had not been anticipated was uncovered, that is, not denying that racism has occurred per se, but a denial that there is anything wrong with what happens in the video because Muslims allegedly do not deserve better treatment. Altogether, 111 comments were analysed in detail for this article and the excerpts presented here were chosen to represent the different strategies of denial that were revealed.
Quotes are provided verbatim, thus errors in spelling, grammar and punctuation are left as in the original post. In order to preserve the anonymity of the posters whose comments are analysed, pseudonym initials are used. Initials instead of names are used to not reveal the gender of the posters. In addition, while it is possible to decipher the gender of the name used in the posts, the neutral pronouns 'they' and 'them' are used for the posters, since we cannot know whether the name chosen is real or depicts the poster's real gender.
The comments were coded using critical discourse analysis, in particular the Discourse-Historical Approach (DHA) developed by Wodak (2001, 2009). As a strand of critical discourse analysis (CDA), DHA also focuses on critique, ideology and power. While DHA sees ideology as a vehicle for establishing and maintaining unequal power relations through discourse, one of DHA's specific aims is to "demystify' the hegemony of the specific discourses by deciphering the ideologies that establish, perpetuate or fight dominance' (Reisigl and Wodak 2009, p. 88). In addition, language is not powerful on its own, but rather is made powerful by powerful people. It therefore critically analyses the language use of those who have the power to change and improve conditions (Reisigl and Wodak 2009). While the subjects of this study may not be 'symbolic elites' who occupy official positions of power such as politicians, journalists or teachers ( van Dijk 2000), Australia has been founded on the premise of White superiority and hegemony. As van Dijk (2000, p. 15) argues: In principle all members of dominant groups are directly or indirectly involved in systems of inequality such as racism. Even if not engaging in acts of discrimination or 'everyday racism', dominant group members often at least profit from discrimination: on the whole they tend to have better jobs, salaries, housing, education, health and status than minority group members.
This study is thus premised on the notion that those denying anti-Muslim racism benefit from a system of inequality and privilege-a system whereby dominant group members receive unearned advantages (Sue 2003). Structures, policies and institutional practices are set up to benefit members of the dominant group and hinder other groups 3 .
The posts were coded following DHA principles, namely to determine how the posters attempt to assert their dominance and their version of events in the public online realm.
To this end, the following research questions were asked of the data: How do the online posters deny that anti-Muslim racism has occurred?
What discursive strategies do the online posters use to legitimise their arguments?

Reversal and Displacement
Many posters 'reversed' the problem by accusing Muslims of worse crimes than what happened to the Muslim woman on the train. This has three effects: firstly, it displaces the original act of racism and attempts to invalidate it. It draws attention to crimes that Muslims supposedly commit so in contrast, the original act seems trivial and irrelevant. Secondly, it blames the victim by implicating her in these much more serious and violent crimes by association. Thirdly, it shifts the role of victim to Australian/Western/White society or oneself in attempts to gain sympathy. The reversal/displacement posts contained some of the least politically correct statements and certainly contravened social taboos by being irrational and overly emotional (see, e.g., Billig 1988; van Dijk 1992).
The following excerpt is a typical example of how posters reverse/displace the problem, naming Muslims as perpetrators instead: Well done for addressing a problem that isn't a problem. Badly acted, badly scripted, completely unrepresentative of true events. Meanwhile, sharia zones, muslim rape gangs, hate preaching, ISIS sympathisers, all exist as subjects it's just too edgy to address. But then again, I guess to be racist, you have to be white to begin with.-AB In this excerpt, the denial is direct and AB does not even attempt to conceal it. While 'addressing a problem that isn't a problem' is ambiguous in that it is not clear if they believes such events never happen, or that they do happen but they are not problematic (i.e., normalising harrassment of Muslims), the next sentence confirms AB believes such events do not even happen. They then contrast this with 'true' events that they believe exist. In comparing this with offline discourse, a denial that racism is true is also a known strategy, whereby the speaker argues that those who say something or someone is racist, does not know what racism is or is exaggerating ( van Dijk 1997). Here, AB deflects the denial into an attack on Muslims, accusing them of violent crimes and racism, succumbing to stereotypes of the 'dangerous, deviant Muslim' (Poynting and Briskman 2018). Cacho (2000) identified a similar swap of 'victim' and 'villain' in American newspaper discourses, which take on a White victimhood approach and portray White middle-class citizens as victims and undocumented immigrants as villains. However, traditional media does follow certain journalistic standards and editorial practices in attempts to provide objective coverage in general, which is not the case for non-traditional internet sites (Melican and Dixon 2008).
AB adopts a White victimhood mentality and sarcastically declare that only Whites can be racist. Here, adopting a discourse of White victimhood is a deflection strategy allowing AB to imply that racists who are Muslim, and supposedly non-White, will always get away with it while Whites do not due to a system of discrimination against Whites. The effect appears to be an attempt to incite pity and empathy for Whites who AB believes are the true victims of Muslim perpetrators, and not vice versa as depicted in the video. Similarly, a study by Bloch et al. (2020) on a US anti-immigration online forum found participants argue that White people, men in particular, are victims of systematic oppression. The data present in that study reflects the Trump-era, where a shift from a colour-blind approach, whereby race supposedly no longer matters (see Bonilla-Silva 2003), to a White supremacist ideology occurred (Bloch et al. 2020). However, as Carney (2016) has argued, the new colour-blind approach incorporates White supremacy, whereby existing systems privilege White supremacy, even if interpersonal racism were suddenly erased. This second excerpt is somewhat different in that it does not deny racism per se, but rather justifies it. This is arguably more disturbing than denial because the poster is insinuating that racism toward Muslims is acceptable.
Oh brother, Here it goes. The propaganda of the left of guilt and how everyone is a raciest. As Muslims invade and change the culture of nations and kill innocent people just living their lives. If you feel any anger or frustration. Here is the campaign to make you feel guilty and liable you as a racist. How Asinine!!! Don't fall for it.-CD This too is a displacement of racism using the 'reversal' strategy. As in the above excerpt, RW diverts attention to Muslims' supposed crimes and chooses an extreme, violent crime that is designed the shock readers. However, CD does not deny that the video depicts racism; rather they attempt to justify the anti-Muslim racism seen in the video. Similar to AB's discourse, CD adopts a White victimhood mentality whereby Whites (not stated, but implied) are the true victims of 'propaganda of the left of guilt'. CD believes that 'we' are double victims: firstly, Muslims are invading 'our' countries and trying to kill us, and secondly, 'we' are being called racist for feeling justifiably angry about this. As Stewart (2019) argues, anti-Muslim nativists believe that Muslims and Islamic institutions contaminate Australia's pristine cultural environment. According to CD, this then justifies racist attacks. CD deliberately falsely interprets the point of the video by declaring that the video is propagating that ' . . . everyone is a raciest [sic]'. CD's comment also seems to imply that people's anger and frustration can play out as attacks on Muslim, which perhaps also explains why they do not deny that the video depicts racism. As Atton (2006, p. 580) argues, in far-right racist discourses, racism is considered to be 'born out of suffering and repression, not hatred'. That is, racists are the oppressed victims whose racism is presented as a 'reasonable reaction to the imputed racism of the Other' (Atton 2006, p. 580).

'Tit-for-Tat'
The 'tit for tat' strategy usually involves a sentence containing the phrase 'If I go to their country . . . ', or 'If I go to a Muslim country . . . ' This strategy follows a hypothetical scenario of the speaker/writer going to a member of the outgroup's supposed country of origin and either behaving respectfully and obediently, or being treated badly by the intolerant and prejudiced locals, or both. This is then used to claim that Muslims behave improperly in 'our' country (unlike the writer/speaker) and justify why 'we' should treat them with the same intolerance as they would us. This strategy is very problematic for three reasons: firstly, it assumes that Australia is only the home country of White, Anglo or European people, and implies that non-Whites' national identity must lie elsewhere (Haw 2021). The White/Anglo/European speaker or writer thus automatically ascribes themselves the identity of being 'Australian' while excluding others (Haw 2021). Secondly, the speaker/writer uses these hypothetical claims to justify the harrassment and vilification of Muslims or other outgroups. Finally, the writer/speaker portrays Muslim countries in harsh, negative terms without providing any evidence that this portrayal has any truth to it. The 'tit-for-tat' strategy is in some ways a positive self-presentation and negative otherpresentation strategy, which is often seen in political discourses discussing immigration and multiculturalism (see van Dijk 1993).
The following excerpt exemplifies the 'tit for tat' strategy. Unlike in political speeches, the writer, EF, does not emphasise positive self-presentation, but rather Australians' right to treat Muslims however they wish-which evidently means badly. It is likely that, unlike politicians for whom face-keeping is important, EF does not feel a need to present themself positively, since they remain anonymous and receive no repercussions from their post. Instead, they emphasise their distaste for the video's message that Muslims should not be harrassed.
Why is it when we have more Muslims coming into Australia than Europeans we all of a sudden have to be so tolerant and respectful because it might "upset" them. Whatever happened to free speech? I'm not saying that every person who sees a Muslim should verbally abuse them but we don't need to make these promotional videos just to show how weak and numb minded we are towards them because if we were a minority in their country they wouldn't do the same for us.-EF EF clearly racialises Muslims here as the antithesis of 'European', indicating that they believe Muslims are not just adherents of a religion, but rather belonging to a specific ethnicity or race. They categorise Muslims separately from Europeans, indicating there can be no such thing as European Muslims. This confirms Ozduzen et al.'s (2021) argument that biological racism is becoming present again through online platforms. Their perception of an 'Australian' is clearly White and not Muslim. This is further confirmed when EF declares 'if we were a minority in their country . . . ' According to EF, Muslims are not Australians and come from another country where the locals would treat (European) Australians badly. Due to this, EF believes we should have the right to treat Muslims however we like and there is no need to treat them with tolerance and respect. This is an attempt to invalidate the point of the video because in EF's eyes, there is nothing wrong with the scene in the video where the antagonists verbally mock the protagonist and pull at her headscarf. In fact, they see this as a right to 'free speech'. This aligns with Lim's (2017) argument that social media facilitates the freedom to hate along with freedom of expression, where individuals exercise their right to voice their opinion while silencing others. In this case, the attempts at silencing are not directed at other users, but at the producers of the video.
EF's assertion that there are more Muslims than Europeans coming to Australia subscribes both to the ignorance about actual numbers of Muslims living in Australia and to an alarmist strategy designed to scare readers. In 2016, an Ipsos poll found that Australians estimated the proportion of Muslims in Australia to be 12 percent when the real percentage was 2.4. Australians also believed that the Muslim population is growing much faster than it is, thinking Muslims would make up 21 percent of the population in 2020 (Ipsos 2016) whereas the 2021 national census found that 3.2 percent of Australia's population reported Islam as their religion (Australian Bureau of Statistics 2022). While EF does not mention any numbers specifically, the discourse still subscribes to the 'numbers game', a well-known move in political discourse in the negative presentation of immigration ( van Dijk 1993). EF sets up a scene where it appears Muslims are invading Australia and suggests that ill treatment of Muslims was acceptable when fewer Muslims were migrating to Australia. EF argues that we need to keep this fast growing number of Muslims under control with harsh measures, justifying the ill treatment as depicted in the video. This strategy is thus not a direct denial that racism towards Muslim occurs, but rather a denial that this racism is unacceptable. This type of denial does not typically present in offline discourses as it would reveal the speaker/writer to be irrational and blatantly racist.

Reverse Racism
The reverse racism strategy involves using complaints and cries of unfairness towards Whites in order to show that Whites are the true victims of racism, not non-Whites.
Pretty sure white Australians get bullied everyday and just because we do not appear physically ethnic or fit the bill of a stereotypical religion of any sort-just plain old white-means that we are the culprits. Let's face it, if we want to get down to the nitty gritty pretty sure we are all discriminated against and we should all stand up for one another regardless. The media forgets that white Australians with no specific religious background get treated the same its just not highlighted because it's easier to blame. I am aware that there are many different ethnicities and religious stances in this video and what is happening is wrong but what if the role was reversed and it was a white Australian being bullied by a Muslim woman and a Chinese man? Would that be impactful? Food for thought.-GH GH subscribes to similar thinking as the participants in Hatchell's (2004) study who believe because White people can be racist toward Indigenous Australians, that the reverse is also true. According to this type of thinking, racism occurs only as individual and personal encounters without any context that takes into account political, social or economic factors. The 'White victimhood' mentality is evident here, where GH describes Whites as not only victims of racist non-White perpetrators but of a society which does pay sufficient attention to the injustices against White Australians. In this way, according to GH, Whites are disadvantaged because while 'we are all discriminated against', Whites are always considered perpetrators only and cannot expose the racist acts perpetrated against them.
As with the tit-for-tat strategy, GH presents hypothetical 'White victimhood' scenarios to support their argument. They do not provide any real-life examples of the type of racism that they are talking about. They hypothetical examples are designed to convince readers that Whites are being treated unfairly. GH believes that racism against Whites is a more important issue than racism against a non-White person, the latter of which is depicted in the video. This aligns with Sharples and Blair's (2021) findings that some White Australians believe they are treated the worst out of any group in Australian society. One of the perpetrators in the video is actually a man of Asian appearance by deliberate design to offset any accusations about the perpetrators being all White people. However, GH still presents the 'role reversal' idea as unique and innovative. The effect of this is to diminish what the Muslim woman experienced in the video. They end their comment with a rhetorical question and 'food for thought' as if their comment had been insightful and clever, and not something anyone else had considered before.
Finally, GH uses several disclaimers in order to maintain positive self-presentation, which is a similar strategy to offline discursive strategies. Firstly, they state that 'we should all stand up for one another regardless'. The 'regardless' here is contradictory when they clearly emphasis Whites status as victims and not as perpetrators. Secondly, they admit that what is happening in the video is wrong, but attempt to turn the tables and discount the fact that the vast majority of people reporting racial discrimination are from non-White backgrounds (Blair et al. 2017). Experiences of racism are much higher among Aboriginal and Torres Strait Islanders and people from a Language other than English (LOTE) background. In institutional settings, everyday settings and everyday racism, Indigenous or LOTE background people experienced racism at approximately twice the rate of their White counterparts (Blair et al. 2017) 4 . In fact, eighty percent of Muslims have experienced unfavourable treatment due to their religion, race or ethnicity (Australian Human Rights Commission 2021). To claim that everyone, regardless of their religious, racial or ethnic background experiences racism at the same rate is untrue and only serves as an attempt to invalidate the fact that non-Whites are often treated less favourably in nearly all aspects of life compared to Whites.

'Muslims Are Not a Race'
The 'Islam is not a race' excuse is commonly used in racial harrassment and vilification of Muslims (Savoulian 2018). This type of argument usually takes one of two forms: one form disagrees that there is such a thing as anti-Muslim racism, but concedes that it is perhaps religious discrimination. This form usually acknowledges the negative effects on Muslims. The second form only uses the 'Islam is/Muslims are not a race' argument in order to invalidate any negative treatment Muslims receive resulting from hatred of or hostility toward Muslims. This is clearly the form IJ uses in the extract below: So they are not racist hate our way of life hate us and a fucking news flash is its not a race it's a religion they are not like a aboriginal when u call a aboriginal a black cunt that's racist do u think the Muslims give a fuck about us y use a women with a rag on her head for a racist add use a native from Australia a race that deserves respect more then these religious bashing terrorists.-IJ This excerpt starts off sarcastically in a deliberate misinterpretation of the video and uses crude language and swearing, something that is very unlikely to occur in political discourses or other face-to-face discourses. There is nothing in the video to suggest that the Muslim woman or Muslims in general 'hate our way of life'. IJ's response is childish and petulant, deliberately trying to displace the original racism by pointing out that Muslims are the ones who are racist. This is very similar to the displacement strategy discussed in the previous section where Whites are portrayed as victims of Muslims' racism. However, IJ suddenly switches strategy by following the 'it's not a race, it's a religion' argument.
IJ's discourse then enacts biological racism straight after claiming Muslims are not a race. They argue that the video should have used a Muslim woman ('woman with a rag on her head') for a racist instead of a 'native from Australia'. We can assume that IJ does not mean Indigenous Australian when they say 'native' as they had previously used the term 'Aboriginal' to denote an Indigenous Australian. Instead, it is highly likely that IJ sees White Australians as 'natives' and argues that they deserve more respect than 'religious bashing terrorists', clearly referring to Muslims. There is a clear racial categorisation and hierarchisation here: White Australians ('natives') are superior and deserve the most respect, and Muslims deserve the least, with Indigenous Australians somewhere in-between, though closer to the bottom near Muslims. The reference to 'black cunt' indicates this, even though IJ could always deny that they were actually calling Indigenous peoples that. Muslims are not just adherents to a religion in the eyes of IJ here: if that were the case, even 'native Australians' could also be Muslim, but IJ clearly differentiates the two. In a similar vein, Carr and Haynes (2015) found that the microaggressions Muslims in Ireland experienced were often related to the idea that being Muslim and being Irish are mutually exclusive. The positive attribute of their Muslim identity racialises Muslims as non-Irish.
The perception of White Australians as 'native' subscribes to the now defunct doctrine of terra nullius, or empty continent, which had been embodied in Australian law for almost 200 years (Augoustinos et al. 1999). The myth of terra nullius has seen discursive constructions of Australia's past as European settlement of an empty continent (Augoustinos et al. 1999). Whites are of course then seen as the original inhabitants of Australia. Overall IJ's discourse serves to present a picture where Muslims are the racists and White 'natives' are the victims, and never vice versa, thus denying that racism towards Muslims is ever possible.

Discussion and Conclusions
Posters of online comments on a bystander anti-racism video used a variety of strategies of denial of anti-Muslims racism. Two types of denial were identified in the data: first, the typical denial that what had occurred in the video was racism at all, but second, a more disturbing denial that there is nothing wrong with what we witness in the video. The latter type of denial accepts that this might be racism, but argues that Muslims deserve their ill treatment, or at least that we do not need to treat them decently or respectfully. This kind of denial would normally not be present in offline discourses as it does not show the speaker to be rational, decent and 'not racist'. However, under the shield of anonymity in online spaces, speakers are bolder and more daring in putting forward such an argument.
More general strategies of denial included displacement of the original act of racism by accusing Muslims of worse offences; offering hypothetical scenarios of Muslims treating 'us' badly to explain away 'our' treatment of 'them'; accusing the video producers of reverse racism against Whites; and of course the oft cited quote that 'Muslims are not a race'. While the strategies are different in their own ways, they all have the same effect of diminishing what the Muslim woman experienced in the video as well as silencing those who wish to highlight that racism toward Muslims is a problem in Australian society. Indeed the online comment posters seemed to take offence at the fact that there could be anything wrong with (White) Australians' treatment of minorities. Portraying White Australians in negative light for vilifying Muslims is immediately considered to be discrimination against Whites.
The existence of social media has now enabled researchers to access such data, which was hitherto inaccessible: that is, people speaking freely about cultural, racial and religious diversity without feeling restrained by social conventions, social taboos or etiquette around politeness. The drawback to social media is the explosion of highly racist and offensive material that is published online for anyone with an internet connection to see. Social taboos around making blatantly racist remarks have not erased racist thoughts and attitudes, but simply masked them. In an unrestricted and anonymous environment such as on social media, users do not need to abide by any social conventions around what is or is not acceptable. Therefore, it is likely that statements leaning on biological racism, which has been generally concealed by social taboos in the past few decades, will be used more and more frequently in arguments and comments in online discourses.
This presents some complexities when investigating discourses on Muslims. With most other ethnic or racial groups such as Black people or Asians, there would generally not be any denials that they are a race, even though there may be denials that it is about race (as opposed to culture or lifestyle). However, in the case of Muslims, there can be a denial that racism against Muslims is possible while simultaneously racialising Muslims and insinuating that they are inferior to other racial and ethnic groups. This type of contradictory discourse was uncovered in this dataset.
Some of the online discourses here have indeed displayed a disturbing shift back to biological racism as Ozduzen et al. (2021) have found. This also indicates a perception of Muslims as not just adherents to a certain religion, but as a group of people with genetic commonality that is inheritable. Another kind of discourse that appears to be particularly salient in relation to Muslims is the 'If I go to their country . . . 'argument, the 'tit-for-tat' strategy. This seems to stem from the belief that all Muslim countries are oppressive, misogynist and hostile toward foreigners. It also adheres to the 'positive selfpresentation/negative other-presentation' discourse so prevalent in political discourses (van Dijk 1993), but presents only hypothetical scenarios rather than real ones. Creating imaginary scenarios to prove a point indicates a disturbing trend toward vilifying Muslims. The online environment would only encourage such discourses because the writers are not held accountable in any way for what they write or for any misinformation they spread.
Finally, there were many instances of self-victimisation and accusations of 'reverse racism' where Whites were seen as victims and Muslims the perpetrators. In seems that some are intent on viewing the world from this perspective no matter how solid the evidence is that Muslims are being harrassed and vilified in everyday life. Due (2011) is certainly correct in arguing that the purpose of this kind of discourse is to further marginalise groups that are already disadvantaged. The difference between online and offline discourses appears to be that in online discourses, the writers make little attempt to be politically correct or portray their arguments as rational or logical.
Overall, while denial of anti-Muslim racism existed before the rise and normalisation of social media, the online environment has given writers free reign to say whatever they wish with impunity. This leads to discourses that are not only irrational and illogical, but also vilifying and harmful. The disturbing intention of the denials is to stop people from taking action against racism. In this particular case of a bystander anti-racism video, the vitriolic response is heightened because the video is encouraging bystanders to take action against racist abuse. This may have hit a raw nerve among deniers of racism-not only do they feel accused of racism themselves, but they are also aggrieved that they are supposed to take a stance against racism toward non-Whites. If we can expose the strategies of racism deniers, this is one step toward deconstructing their arguments and fighting back. Institutional Review Board Statement: Ethical review and approval were waived for this study due to the public and anonymous nature of the data.
Informed Consent Statement: Informed consent was not possible while protecting the privacy of social media data due to the anonymous nature of the data.

Conflicts of Interest:
The author declares no conflict of interest.

1
The video can be found on Youtube at: https://www.youtube.com/watch?v=7JUzzBmbchs (accessed on 6 October 2022). 2 The acronym was deliberately used to stand for 'Intolerance, Slaughter, Looting, Arson and Molestation of women'. 3 It should be noted that the dominant group does not always consist of Whites. It depends on the country and context. For example, Lan and Navera (2021)'s study analyses the discriminatory ways in which the dominant Han Chinese write about Chinese Muslims in China.