“Who is really British anyway?”: A thematic analysis of responses to online hate materials

Introduction

On the 21^st of March 2016, 3,3 billion people worldwide were online (Internet Live Stats, 2016a). The relatively recent emergence of social networking sites (SNSs) has contributed noticeably to the growing numbers of Internet users. For example, Facebook had 1.6 billion users in March 2016 (Internet Live Stats, 2016b), and YouTube claimed, that in 2015, 1 billion users visited its page per month (YouTube, n.d.).

Despite their immense popularity, some SNSs have recently come under critical scrutiny. In Britain, for example, there is increasing government concern about the Internet’s role in the incitement of hatred, including racial hatred. There have also been claims by some researchers, advocacy groups, Internet service providers (ISPs), and charities that websites promoting hatred may contribute to the development and exacerbation of radical attitudes and behaviours and, as such, foster terrorist sympathies (e.g., Foxman & Wolf, 2013; Perry, 2000; Stern, 2002). However, to date, claims regarding the persuasive power of the Internet remain largely based on anecdotal observations, common sense assumptions, and on government concerns regarding the protection of members of the public.

The existence of hateful materials online seems indisputable (e.g., it can be evidenced by certain key word searches). In addition, online hate materials, including propaganda video clips on websites, such as YouTube, are regularly found on computers seized from terrorist suspects (e.g., the Boston Bombers, see: Robi, 2013; or the Lee Rigby murderers, see: Swinford, 2013). Consequently, lawmakers across the globe have begun to amend and develop legislation to account for the evolution of the Internet into a potential medium for stirring up and promoting hatred. In the United Kingdom alone, this legislation includes the Offences Against the Person Act, 1861, the Public Order Act 1986, the Malicious Communications Act, 1988, the Protection from Harassment Act 1997, the Crime and Disorder Act 1998, the Communications Act 2003, the Criminal Justice Act 2003 and the Racial and Religious Hatred Act 2006. Consequently, in the United Kingdom, several arrests and convictions have been made in relation to the publication and distribution of online materials with the intent to incite racial hatred (e.g., Colin White, see: BBC News, 2016; Ihjaz Ali, Kabir Ahmed, & Tazwan Javed, see: BBC News, 2010; Joshua Bonehill-Paine, see: Al-Othman, 2015; Matthew Doyle, see: Burrows, 2016; Teenager, see: Crown Prosecution Service, 2010). Arrests have also been made in Australia, China, Dubai, Russia and Kazakhstan where several people have been convicted of inciting hate via social media, including ethnic hatred (e.g., Ai Takagi, see: AFP, 2016; Alkhanashvili, see: Kazakhstan Newsline, 2015; Mukhtar, see: Radio Free Europe Radio Liberty, 2015; Pu Zhiquaing, see: Phillips, 2015; Tatyana Shevtsova-Valova, see: Latest.com, 2015), religious hatred (e.g., Yklas Kabduakasov, see: Spencer, 2015; UAE national, see: Agarib, 2016) and extremism (e.g., Vadim Tyumentsev, see: Reuters, 2015).

Academic research exploring online hatred has begun to emerge in several social sciences (mainly sociology and criminology). So far, this research has tended to have two main analytic foci, namely on providing (1) detailed descriptions of online hate materials or (2) characterisations of polarised individuals/groups. For example, research on online hate materials has focused on outlining the specific contents of hate-promoting websites and has provided insights into the use of propaganda to polarise and recruit new members to common ideological causes (e.g., Anti-Defamation League, 2005; Blazak, 2001; Bostdorff, 2004; Douglas, 2007; Douglas, McGarty, Bliuc, & Lala, 2005; Gerstenfeld, Grant, & Chiang, 2003; Levin, 2002; McNamee, Peterson, & Peña, 2010; Perry, 2000; Turpin-Petrosino, 2002; Zhou, Reid, Qin, Chen, & Lai, 2005). Research focusing on understanding the type of individual or group who produces hateful materials or hate speech online has tended to concentrate specifically on people who are already polarised, thus identifying not only some personality traits associated with hate speech but also some of the ways in which polarised individuals or groups share their ideology and deliberately use prejudice to create conflict between groups (e.g., Angie et al., 2011; Cammaerts, 2009; De Koster & Houtman, 2008; Doane, 2006; Erjavec & Kovačič, 2012; Ezekiel, 2002; Glaser, Dixit, & Green, 2002; Green, Abelson, & Garnett, 1999; Jacks & Adler, 2015; Wojcieszak, 2010).

However, thus far, research tapping into the broader processes underlying the development of online hatred remains limited – despite the reported increase in ‘hate blogging’, especially on SNSs, over time (e.g., Chau & Xu, 2007; Citron & Norton, 2011) and the recent surge in arrests and convictions for SNS-related hate offences. In particular, there is limited empirical evidence investigating how Internet users might respond to hateful online web content displayed on ordinary SNSs (i.e., YouTube, Facebook, Twitter). In particular, whilst one might expect to come across hateful/polarising content on certain websites (i.e., English Defence League, Stormfront), one might not expect to see such content during their visit to their familiar SNSs. As a result, we know little about how Internet users might respond to or be influenced by hateful online materials displayed on SNSs.

Many SNSs and online news web sites now allow, and even encourage, online users to express their views or responses to different web contents (including specific statements, pictures, video clips or TV programmes) via posting/commenting. These posts or comments are usually displayed publicly and, in turn, often seem to elicit further comments/posts from other users – who might agree or disagree with either the initial poster or with the web content in question. It seems therefore plausible that certain web content could lead SNS users to respond to such content in a polarised manner. These polarised responses could, in turn, lead to (virtual) conflicts amongst SNS users.

In fact, such virtual conflicts can be observed in so-called online ‘trolling’, where certain individuals (i.e., ‘trolls’) intentionally aim to stir responses amongst users. Trolling has been attributed to online anonymity (Hardaker, 2010; Herring, Job-Sluder, Scheckler, & Barab, 2002), which, in turn, links to the so called online disinhibition effect (ODE; Suler, 2004). In particular, the ODE is thought to be the result of a lack of social constraints mediated, predominantly, by online anonymity (Joinson, 1998; Suler, 2004). Whilst not all behaviour associated with online disinhibition has to be negative (i.e., increased self-disclosure), as noted above, online disinhibition can encompass negative communication, such as offensive language, harsh criticisms, anger, hatred and threats (Suler, 2004). Suler (2004) referred to this form of the ODE as ‘toxic’ online disinhibition.

Accordingly, Coffey and Woolworth (2004) found that whilst people’s responses to an alleged racially motivated murder reported in an anonymous online newspaper discussion board resulted in expressions of insults, racism, abuse and hatred (i.e., polarisation), no such responses were observed when the incident was discussed in a face-to-face meeting. They consequently attributed these polarised responses expressed online to online anonymity. In addition, Peddinti, Ross, and Cappos (2014) found that compared to identifiable Twitter users, anonymous users were less inhibited in relation to their Twitter activity. They also found that there are more anonymous Twitter users following sensitive topics, such as pornography than identifiable ones.   

Further, Doughty, Lawson, Linehan, Rowland, and Bennett (2014) found that controversial television broadcasts elicited many polarised (i.e., anti-social, abusive, negative, aggressive and hostile tweets - a phenomenon frequently referred to as ‘flaming’ (O’Sullivan & Flanagin, 2003; Suler, 2004; Turnage, 2008). In Doughty et al.’s study (2014) ‘flaming’, which is described as a negative aspect of the ODE, tended to be directed at the individuals or specific communities depicted in these broadcasts. In addition, Lange (2007) and Moor, Heuvelman, and Verleur (2010) investigated possible causes of ‘flaming’ behaviour on YouTube and found that people predominantly ‘flamed’ to express their disagreement with certain video or comment content. They also found that a small number of users ‘flamed’ for mere entertainment (Moor et al., 2010) and enjoyment (Lange, 2007). Lange (2007) and Moor et al.’s (2010) findings also indicate that, although ‘flaming’ seems prevalent on YouTube, it is not always perceived to be harmful, even by the recipients or ‘targets’ of ‘flames’.

Moreover, Faulkner and Bliuc (2016) explored discursive strategies of racist and anti-racist supporters expressed on online news web sites. They found that moral disengagement is a very common strategy used by those supporting racism in order to justify or ‘feel okay’ about their views. They also found that those who opposed racist views/acts rarely used this strategy when expressing their views online. Given that expressions of racial attitudes are no longer socially acceptable (Crowne & Marlowe, 1960; Furnham, 1986; Gawronski, LeBel, & Peters, 2007; Joinson, 1998; Steinfeldt et al., 2010), it does not seem surprising that moral disengagement is predominantly observed in those supporting racist attitudes.  

Current evidence therefore presents some strategies through which the expression of certain attitudes (i.e., racial) can be justified. In addition, it suggests that SNS content may not only facilitate discussion amongst online users but can also lead to attitude polarisation. However, to what extent does controversial web content succeed in influencing or changing people’s actual attitudes or views?

Steinfeldt et al. (2010) found that online newspaper forums often provide misinformation about specific groups (i.e., American Indians, in their case) which, in turn, can reinforce existing negative stereotypes and thus facilitate the expression of racist attitudes. More specifically, Lee and Leets (2002) argue that White supremacist web content can influence attitudes and views of polarised as well as ‘neutral’ adolescent online users. Furthermore, Rauch and Schanz (2013) found that, in the context of White supremacist online content, frequency of Facebook use influenced users in one of two ways. In particular, compared to infrequent Facebook users, frequent users were both more likely to agree with or reject racist messages (e.g., about White superiority), thus suggesting that those who frequently use Facebook are at higher risk of responding in a polarised manner to its content compared to those who infrequently use it.

In sum, the evidence outlined above suggests that the content of materials posted on online forums or SNSs can influence and polarise the expression of people’s views as well as lead to online abuse– even in cases where the initial web content is not specifically problematic or hateful. Yet, the extent to which explicit online hate materials might elicit hateful responses in online users more generally remains unexplored. For example, given recent media coverage (e.g., Castle, 2013; McNab, 2015; Nirvana News, 2016; Poch, 2015; Risen, 2014) focusing on the use of YouTube and other social media sites (e.g., for promoting domestic violence or for terrorist propaganda and recruitment), it seems increasingly likely that people come across hateful content without necessarily seeking to do so. This then, raises the question of how SNS users in general might respond to explicit hate materials.

This study therefore aims to explore the following research question: How do YouTube users respond to online materials that can be described as explicitly inciting hatred?

Our research question here is deliberately broad in order to allow for a qualitative, bottom-up and inductively-driven analytic approach – which has been deemed the most appropriate research strategy for under-researched areas (Braun & Clarke, 2006). It is therefore important to stress that this study aims to offer an in-depth understanding of people’s responses to a very specific type of online hate material (i.e., a video clip with racist content) within a particular SNS context (i.e., YouTube). In other words, this study seeks to provide a detailed description of the qualitative nature of people’s online responses in a specific context – that is, it does not set out to make any claims about online responses to hateful materials in general. However, the authors hope that the qualitative findings outlined below will provide a springboard for further quantitative research in this field to evaluate the extent to which the key findings can be generalised.

Method

Materials

Data source. This analysis was based on comments made by users of the social networking site YouTube in response to a particular video clip uploaded on its website. The video clip depicted a White woman who racially abused other passengers. The woman, later identified as Emma West, was travelling with her toddler son on the London underground in October 2011. The scene was filmed by a passenger on his/her mobile phone and was subsequently uploaded as video clip onto YouTube by one YouTube user on 28th November 2011, and lasted approximately five minutes (for a detailed description of the clip’s content, please see below). The clip led prompted an unprecedented number of viewers to respond to its content via YouTube’s commenting feature – and these responses provided the raw data for this study.

The nature of our choice of analytic materials was therefore somewhat opportunistic. In particular, when the video clip in question went viral online, the authors became aware of the analytic potential the accompanying comments presented for addressing the current research question. To the authors’ knowledge, the video clip was the first of its kind (in the UK) to have elicited such large numbers of views and comments. Using this data sours, therefore, had several methodological advantages. First, the voluntary exposure of YouTube users to the clip’s content circumvented any ethical concerns an experimental design (i.e., the experimental use of hateful stimuli) would have posed. Second, users’ voluntary choice to comment on the clip’s content was likely to render responses more ecological valid than responses elicited in experimental settings. Third, unlike in many laboratory experiments, there was no researcher engagement with ‘respondents’, thus circumventing experimenter bias or social desirability effects.

The time-frame for data collection, in turn, was determined by the sheer volume of online comments made in response to the YouTube clip. Forty-eight hours after the clip’s upload, the authors decided that the volume of comments was more than sufficient whilst still presenting a manageable analytic task. Also, the original video clip was removed at approximately the same time because the police started a criminal investigation of the woman depicted in the clip.

Consequently, the original raw data consisted of 71.295 comments. The actual content of the video clip is described below, followed by an outline of the different types of YouTube users’ responses the clip elicited after its upload, as well as the legal consequences Emma West (i.e., the woman depicted in the video clip) faced for her behaviour.

Ethical considerations. The online comments (see above) that constituted the data source for this study were available within the public domain. As such, the responses were visible and freely accessible to anyone using YouTube. YouTube’s terms of service explicitly state that, by uploading/posting content (including comments) onto the site, users automatically consent to this content being used (i.e., reproduced, distributed, prepared in derivative works) by third parties. However, in order to ensure that the YouTube users whose comments formed the basis of the analysis could not be identified further, all user names were anonymised.  

The YouTube video clip. At the start of the video clip Emma West can be heard posing the rhetorical question “...what has the UK come to...” - which she then proceeds to answer herself by stating “...a load of black people and a load of f***ing Polish”. Emma then verbally attacks several fellow passengers for “not being English”, as well as telling them to “go back to their own country”. Next, a Black woman asks Emma to refrain from swearing for the sake of the children present on the train. In response, Emma demands the Black woman state which country she is from, to which the Black women responds that she, too, is English. Emma appears not to believe her. The clip continues by depicting Emma telling the same woman to “go back” to her “own country”. Emma then attempts to specify which country the Black woman should “go back” to; she is, however, unable to articulate this correctly. In particular, she refers to an awkward combination of the two countries Nicaragua and Siberia, which in this specific combination sounds to other passengers like the racially offensive term N****R. This, in turn, can be seen to upset a young Black man behind Emma so much that he stands up, visibly agitated, in response to which two fellow passengers immediately get up and quickly calm him down; he then returns to his seat. Shortly after, a young White woman can be seen to rush towards the upset young Black man and embrace him. During this, Emma can be seen to argue with other female passengers calling out “You ain’t f***ing British! You are Black!” Then a baby’s cries can be heard, to which a seemingly angry White woman steps forward and shouts at Emma to “shut up” for waking up her baby. The woman can then be heard stating that she is English and asking Emma what she has to say to her. Whilst the argument between the two women continues, the video clip finishes.

Online YouTube responses. The video clip went viral after its upload and was viewed on YouTube more than 3 million times within the first 48 hours of its upload. Viewers’ responses included comments making death threats towards Emma, her child, and other YouTube users, general threats of violence, as well as offensive and racist comments, which resulted in YouTube’s removal of the clip from their website after approximately 48 hours. Since then, Emma West was arrested and charged with a racially aggravated public order offence. Emma pleaded guilty to racially aggravated harassment, alarm or distress. Additionally, several copies of the video clip have re-emerged on YouTube, resulting in an estimated 11 million views to date, including repeat viewings.

Analytic Strategy

Data preparation. YouTube comments to the video clip were first imported into Microsoft Word and then formatted uniformly (Arial, size 12 with single line and one paragraph spacing between posts) to provide visual ease during the analytic process. This also aided initial familiarisation with the content. Next, comments that solely consisted of Computer-Mediated-Communication (CMC) abbreviation acronyms (Baron, 2004) such as LOL, and response tokens such as yeah, were deleted if the connection or relevance to other comments could not be easily established. Additionally, repeat comments made by the same users (i.e., displaying the same verbatim content) and comments made in a foreign language (30 in total) were deleted, leaving approximately 71.000 comments for analysis.

Data analysis. Based on the relative lack of existing research in this area, as well as the absence of an overall theoretical framework with which to account for how online users respond to online hate materials on SNSs, a bottom up, inductive approach to the data was taken. Overall, the thematic analytic process followed Allen, Bromely, Kuyken, and Sonnenberg’s (2009) procedure. Generally, Thematic Analysis (TA) permits for a degree of epistemological and theoretical flexibility and is therefore not bound or limited to any specific pre-existing theoretical framework - yet, it does allow for the consideration of relevant theories during the analytic process (Braun & Clarke, 2006). This flexibility thus allows for the identification of themes at both the semantic (i.e., ‘micro’) and the latent (i.e., ‘macro’) level of analysis (Allen et al., 2009; Braun & Clarke, 2006). Given the broad inductive scope of our research question, both levels of analysis are drawn on here.

The thematic analytic process involved reading and re-reading the entire data set several times to become more familiar with the data and to identify themes related to the overall research question. A theme was deemed as such if, first, it tapped into relevant aspects concerning the overall research question and, second, if it represented a response-trend across the entire data set (Braun & Clarke, 2006). Reading and re-reading the data set continued throughout the formulation of themes, ensuring that the final themes remained data driven and that the chosen YouTube extracts (i.e., comments) accurately illustrated the corresponding theme/sub-theme. Themes were compared and were either separated further to show different nuances in meaning, or grouped together according to their overlap with one another (Allen et al., 2009). This process allowed for the development of an analytic hierarchy, by ordering the data into overarching themes and corresponding sub-themes (Allen et al., 2009). An overarching theme was defined as such if it appeared to be more inclusive (i.e., higher in the thematic hierarchy) and possessed greater analytic scope and explanatory strength than a sub-theme (Gleibs, Sonnenberg, & Haslam, 2014). Overall, the aim of the thematic coding process was to identify and organise emerging themes that were internally homogenous, externally heterogeneous and had explanatory power (Allen et al., 2009). Attention was also paid to prevalence which was determined by the frequency of responses that illustrated each theme (Braun & Clarke, 2006).

Results

Overall, the majority of responses condemned the content of the video clip. Analysis of YouTube users’ responses to the video identified four overarching themes: (1) Making Sense of Emma, (2) Meeting Hatred with Hatred, (3) Us versus Them, and (4) Contesting Britishness, with the latter two being most prevalent. These main themes are examined in detail below under separate section headings. Each section includes a broad definition and a narrative description of the respective theme content as well as evidence from the actual online comments. Corresponding sub-themes will be presented as sub-headings. It should be noted that here, as in other qualitative analyses (see Gleibs et al., 2014) there is a degree of overlap between themes. For example, all themes reflect how posters, to some extent, attempted to position themselves (and other posters) in relation to racism. It should also be noted that many of the extracts below contain grammatical and spelling errors – which were only corrected in some cases to help clarify the meaning of these extracts.

Making Sense of Emma

Posters were obviously stirred by the content of the video clip and their responses often took a stance on racism. The video content seemed to evoke a range of negative emotions (i.e., expressions of distress) in several respondents. These emotions included shame (“Such a rude, ignorant women makes me ashamed to be British”), disgust and embarrassment (“You disgust me. You are the reason I am sometimes embarrassed to admit that I'm British.”), upset (“This video sickens me. She needs a reality check. I'm deeply upset by this footage.”), and shock (“I'm actually shocked! I can't believe what I have just watched?!"; “I am so shocked and appalled by this woman...”).

Notably, respondents who expressed negative emotions condemned what they saw as Emma’s overt, racist behaviour. Many posters seemed to deal with what they had seen in the video clip by attempting to make sense of Emma and her behaviour. This sense-making focused on Emma’s personal characteristics, and particularly on attempting to attribute Emma's behaviour to internal, intra-psychic processes (i.e., alcohol and/or drug consumption):

OMG that woman is vile. And obviously either drunk or on drugs (judging from facial expressions and slurring). Luckily we don't all feel like her and she is one of a minority of people.. (Poster A)

This extract demonstrates how many respondents who disagreed with Emma’s racist behaviour, tended to attribute her behaviour to negative personal characteristics, which, in turn, enabled them to distance themselves from Emma. This distancing is evidenced by “we don’t all feel like her and she is one of a minority” which, not only rejects racism but also positions the poster in opposition to Emma and other perceived racists and thus removes any basis for socially identifying with her. As will become clear later, the process of distancing also relates to another theme, namely Us versus Them.

The link between disapproval of Emma’s behaviour and attempts to make sense of it by focusing on Emma’s perceived negative personal and socio-economic characteristics becomes even clearer in the extract below. Specifically, here the poster attempts to make sense of Emma’s behaviour by explicitly referring to her perceived low intelligence, low education, and low socio-economic status. Moreover, here these personal characteristics (i.e., negative stereotypes) are seen to be associated directly with racist attitudes:

The fact she can't string her opinions without saying F*** says a lot more about her intelligence. would she happen to be on benefits and live in a council home by any chance? Its usually those people that moan about others but are happy to scrounge benefits and freebies of others who work in Britain and pay taxes like the people she insults (Poster B)

Here, there is an allusion to Emma’s racism as a direct consequence of her perceived low intelligence and low socio-economic status. In particular, these attributions seem to serve primarily to label Emma as a racist – in other words, as one of “those people” (i.e., an out-group member). Here, this strategy to position Emma as a racist (i.e., an out-group member) further seems to serve as a means of making sense of Emma’s behaviour.

 In addition to explaining Emma’s behaviour on the basis of a perceived lack of education, the following extract invokes a comparison between Emma and immigrants (i.e., immigrants "have more skills" and better work ethics than Emma). This comparison further serves as a way of discrediting Emma. The poster also tries to makes sense of Emma’s behaviour by linking it to a particular lifestyle (i.e., sexual transgression and work avoidance), which has been linked with poverty (Parisi, 1998):

This is what happens when u don't get an education that informs you of the realities of immigration. immigrants of all races have to work twice as hard to get jobs here and often have many more skills, in speaking several languages. I wonder what this ignorant fuck's skills are? probably shagging around and talking about how 'blacks' have ruined the country when she probably has never held down a job in her life! (Poster C)

Making sense of Emma’s behaviour was also often associated with reference to Emma’s role as a mother (i.e., "a bad mother" and "a bad role model"). In particular, responses frequently expressed concern for Emma's child, such as “I feel sorry for her little boy”, “she (i.e., Emma) used the child as a shield”, “I hope social services will take her child away", or “poor child”. This, then, corresponds to broader strategies that have been observed in the media (Thomas, 1998) which aim to discredit single mothers through negative stereotyping and their portrayal as immoral and neglectful.

By referring to Emma and/or other perceived racist posters in hateful terms, comments frequently went beyond simply trying to make sense of Emma’s behaviour, attempting to take a stance on racism:

Most of,if not all of the people backing this horrible cunt are unemployed dole spongers who use foreigners as an excuse for not getting a job.Waists of space racist pricks. most of the English football team are black. Show some respect (Poster D)

This, then, goes beyond trying to make sense of Emma: here, insulting and negatively stereotyping Emma and other racists serves to demonstrate the poster’s opposition to their views. As will be seen, this use of hostile and aggressive language also links to the next theme (i.e., Meeting Hatred with Hatred). Also, the poster’s reference here to the ethnic diversity of the English football team – a symbol of national pride – foreshadows the final theme Contesting Britishness.

Overall, Making Sense of Emma encapsulated respondents’ attempts to understand and explain Emma. Specifically, respondents drew on a range of negative stereotypes both on a personal level (i.e., invoking alcohol/drug abuse, low intelligence, low education, ‘bad mother/role model’) and a social level (i.e., invoking a low socio-economic status) in order to account for Emma’s behaviour and to distance themselves from (Emma’s) racist views by positioning her as ‘other’ (i.e., an out-group member).  This process frequently involved the use of strong, hateful language as the following extract illustrates:

Its funny how this silly lady thinks its 'the blacks' that are ruining her country when she is sitting on a tram with her son clearly on drugs and carrying on and the only person decent enough to tell her not to swear in front of kids is black..... I know who I would rather live next door two, the black lady and not the coked up whore. (Poster E)

Meeting Hatred with Hatred

This thick bitch would have got kicked the fuk out of by me the dik head in the back that takes his bag off why didnt he do anything.. Faggot bitch would get killed lil tart if i was there i would kill the bitch !!!! FUK YU DIE BITCH (Poster F)

This extract epitomises the essence of Meeting Hatred with Hatred as another overarching theme in this analysis. In particular, comments encapsulated by this theme depicted posters’ tendency to strongly object to Emma's behaviour by means of demeaning and dehumanising her. Specifically, and somewhat paradoxically, many posters responded to Emma’s behaviour by using hateful (including racist/abusive) and aggressive language, such as “she has a mentality like Hitler”, “she shouldn’t be allowed to breed”, “ugly ass throwback gobshite of a tredless bucketpussy usless eater oxygen thieving minus 80 I.Q”, “lazy”, "whore", "cunt", "chav", “slag”, “anti-social, scum-sucking retard”, "pikey", “albino monkey”, or “honkey”. This use of hateful language in response to Emma's behaviour seems paradoxical in so far as it reflects an attempt to counter Emma’s apparent racism and hatred with further expressions of hatred (hence Meeting Hatred with Hatred).

Along similar lines, posters frequently dehumanised Emma by referring to her as “a piece of filth”, "gutter trash", “White trash”, “mosh mault”, “monstrosity”, “plankton”, “a thing” or a "scum (bag)". These terms invoke a perceived lack of humanity on Emma’s part, thus placing her outside ‘normal’ human society and calling for her social exclusion (e.g., into "the sewer/gutter/trash", a place where society discards its unwanted waste). The dehumanisation of Emma thus represents another way in which posters attempted to distance themselves from her and the video’s content.

In addition, posters who condemned Emma’s behaviour called for her punishment, albeit on a symbolical level rather than a literal one. For example, posters referred to “this lazy sponging cunt should be put down by a vet. Although animals are better looking that that scum.I sugest she goes back to where she came from the sewer”. As these examples illustrate, Emma’s dehumanisation and calls for her punishment frequently went hand in hand – indeed, the latter seemed predicated on the former. These calls for punishment seemed to underscore further posters’ apparent opposition to racism. Further calls for Emma’s punishment included “hitting her”, “knocking her out", “beating her”, “stabbing her”, “slashing her throat”, "deporting her", "locking her up in the sewer with the rats", or to "rot in hell" – which, again, highlight posters’ opposition to Emma's views and behaviour.

Overall, Meeting Hatred with Hatred encapsulated one particular method posters used to express their disagreement with the video clip’s content. In other words, in their use of aggressive, abusive, threatening, and hateful language posters attempted to distance themselves from Emma’s expressions of racism. These attempts at distancing are also evident in the following theme.

Us versus Them

This prevalent theme describes how posters’ attempted to either distance themselves from or affiliate with Emma and her views by invoking a sense of “us” versus “them”. This, in turn, echoes previous work focussing social identity processes (e.g., Blackwood, Hopkins, Reicher, 2012, 2013, 2015; Reicher, Haslam, & Rath, 2008; Tajfel & Turner, 1979; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). For example, posters condemning Emma tended to position themselves alongside other posters who shared their own seemingly anti-racist position – whilst, at the same time, distancing themselves from those who advocated or supported Emma's views (and vice versa). In particular, those posters who did not share Emma's views tended to refer to her and those who shared her views, as ‘them’ - in other words, as out-group members. Posters opposing racism also tended to refer to other like-minded YouTube users as us, that is, as in-group members. Us versus Them consisted of two sub-themes that were labelled Sense of Injustice and Sense of Deprivation.

As already indicated, many posters condemned racist views and consequently distanced themselves from Emma:

@Participant X ¹: […] All racism is wrong. This woman is an embarassment to where I come from. Hate engenders hate. Even the biggest, most intolerant racists must recognise exceptions to their misplaced and ignorant rules. Why not try giving everyone the benefit of the doubt? I don't want to be mistaken for someone like you, or this stupid woman on the video, so why should I assume anything about anyone else? (Poster G)

This poster distances him/herself from racism in two ways: first, on a social level, by calling “all racism wrong” and by labelling racists, as a group, as “intolerant” and “ignorant” and second, on a personal level, by describing Emma in negative terms, calling her an “embarrassment”, “stupid”, and “I don’t want to be mistaken for (i.e., associated with) someone like you” (i.e., someone with racist views). Conversely, posters supporting Emma’s views attempted to distance themselves from immigrants and immigration supporters:

@Participant X ²: True, that was adorable how she stood up for her race and the bravery to say what everyone was thinking in front of adversity and to all the immigrant invaders who came to enforce their foreign oppressive laws and genocide [on] the land of the indigenous Whites. (Poster H)

This comment displays categorical thinking through its use of distancing pronouns, such as “their” foreign oppressive laws, implying that laws were not voted for or made by us (i.e., here White people). Consequently, the poster here distinguishes between the in-group self (i.e., the ‘us’ supporting Emma’s views) and ‘the other’ or out-group (i.e., those opposing Emma). Specifically, the above terminology depicts ‘us’ as good, while ‘they’ are bad. Here, the contrast between negative terms, such as “invaders”, “oppressive”, “genocide”, “indigenous White", and positive terms, such as “adorable”, “stood up” and “bravery” emphasises the distinction between the in-group and out-group (i.e., perceived racial differences). The term ‘genocide’ here also links to a perceived threat, ostensibly posed by immigrants, to posters’ social identity (i.e., social identity threat; Branscombe, Ellemers, Spears, & Doosje, 1999; Reicher et al., 2008; Verkuyten, 2013) which will be discussed in more detail under Contesting Britishness.

Some comments in support of Emma not only created a sense of Us versus Them through use of in-/out-group defining pronouns, but also by displaying high levels of hostility towards out-group members, as illustrated in the following quote:

GOOD FOR HER! THEY CAN ALL FUCK OFF BACK TO WHERE THEY BELONG, AND THE TRAITORS WHO BROUGHT THEM TO THE U.K. MUST BE ARRESTED AND FUCKNG EXECUTED FOR TREASON!!! FUCK OFF! FUCK OFF! FUCK OFF!!! (Poster I)

This comment clearly demonstrates the poster’s positive self-representation and negative other-representation, through out-group defining pronouns (i.e., they/them) and, in doing so, distances him/herself not only from ‘foreigners’ (i.e., ‘they’ do not ‘belong’ in the UK), but also from those people in the UK who support immigration (i.e., "traitors", who should be "executed for treason").

This hostility towards immigration supporters indicates a certain degree of conflict as the recipients of this hostility also belong to a potential us (i.e., the British). Here, British nationals supporting immigration are categorised as out-group members by those opposing immigration which suggests that hostility expressed towards ostensible out-group members serves as a strategy to create inter-group conflict and justify action against them.  

Overall then, Us versus Them tapped into broad social identity processes. That is, the theme and its two subordinate themes (see below) encapsulated posters' attempts to position themselves and other posters in line with the construction of in- and out-groups. As such, respondents' positioning frequently drew on the issue of immigration – which is exemplified by the following sub-themes, Sense of Injustice and Sense of Deprivation.

Sense of Injustice. This sub-theme illustrates the way in which posters constructed a sense of Us versus Them by invoking immigration. In particular, those who voiced racist views tended to express their perceived sense of injustice by claiming that immigrants receive preferential treatment. Respondents who opposed racist views tended to argue that everyone “born in Britain” should be considered "just as British as we (i.e., White British) are" and therefore "have the same rights to live in Britain" (i.e., here, a sense of injustice consisted of denying immigrants their right to claim to be born in Britain as Britishness). These types of responses also related to notions of ‘Britishness’ and link to the final theme Contesting Britishness. There were also many posters who argued that immigration was an inevitable consequence of Britain's colonial history:

Hey lady - you don't like people from other countries or ethnic backgrounds in your country or calling them British or English - you guys started it! invading and looting all those other countries, staying in India, Pakistan, parts of Africa etc - now it's our turn!! PS. - rest of Britain who aren't racist like this stupid cow - we love you, we love your pubs, your beer & stouts, your football and your culture. Peace! (Poster J)

Again, here a sense of Us versus Them is created through use of specific out-group-defining pronouns (i.e., you and your to address Emma and other British people), and in-group defining pronouns (i.e., our and we in relation to members of other countries and ethnic minorities). The poster also proclaims immigration an inevitable and direct consequence of British colonialism. In particular, his/her references to invading and looting express a sense of unfairness and injustice towards people from Britain's former colonies. Whilst this also suggests that s/he identifies as non-British, the poster also distinguishes between two different versions of British identity, namely a non-racist and racist British one, thereby differentiating between different types of British national group membership.

There were, however, also posters who endorsed Emma's stance on immigration:

[…] Islamic extremests do run round screaming "kill the infidel" and the police stand by and do absilutely nothing!. One law for black and another for white isn't working, any more than a housing or education policy that discriminates against established nationals, and merely fosters disharmony. (Poster K)

Although this poster does not directly categorise him/herself as a member of a specific in-group, this comment creates a sense of Us versus Them (i.e., Islamists extremists and Black people versus White people) through ethnic stereotyping. In particular, this post implies that all Muslims are Islamic extremists and not White – a misleading fusion which is likely to reinforce existing negative stereotypes. More specifically, the comment expresses discontent concerning the British establishment (i.e., laws and government policy), whereby the British criminal justice system is accused of unjustly differentiating between groups by "discriminating against established nationals”.

Whilst the poster in the above extract implicitly suggests that current laws/policies are unfair and contribute towards social division, other posters accused the British government and criminal justice system explicitly of committing injustices against those who are ‘really British’. Specifically, their sense of injustice referred to accusations of perceived persecution of ‘real Brits’ for ‘minor crimes’, such as "expressing politically incorrect views", whilst maintaining that ‘severe’ crimes, such as "murder", "rape", or "burning poppies in the street", committed by those ‘non-British’ (i.e., immigrants) "remain unpunished".

Thus, some responses to the video clip accused the British government of applying different rules to different people (i.e., favouring immigrants and discriminating against ‘the British’) - which directly related to posters’ own attempts to differentiate between groups (i.e., Us versus Them). This sense of injustice (i.e., the claim that the establishment unjustly differentiates between groups) here also relates to the next sub-theme.

Sense of Deprivation. The comments falling under this sub-theme conveyed posters’ sense of feeling deprived. In particular, those holding racist and anti-immigration views tended to express a sense of deprivation via creating a sense of Us versus Them (i.e., here ‘us’ versus ‘foreigners’). Some posters blamed ‘foreigners’ for their own financial insecurity and unemployment, including familiar phrases, such as “they take our jobs”, “they are prioritised for housing” and “they receive special benefits and free health care”, whilst "we lose out to them”. Consequently, and as illustrated in the following extract, posters also expressed their sense of deprivation by drawing explicit group distinctions (i.e., racial out-grouping):

I wish more people would stand up for what they believe in we have a freedom of speech but somehow this has been withdrawn with this and why??? she has an opinion and is expressing it..... if people are offended then sorry they shouldn't watch it or be here to be offended against it..... we have got too many immigrants due to legislation..... and what a load of **** it is..... white population lose out on housing jobs and benefits because of foreigners..... how is that fair??? (Poster L)

The above reference to "the white population losing out on housing, jobs, and benefits because of foreigners..." not only suggests a sense of deprivation in general but, particularly, a sense of ‘us’ being deprived compared to ‘them’ (i.e., see fraternal relative deprivation; Runciman, 1966). For instance, ‘our’ (i.e., White British) socio-economic insecurity is here blamed on ‘them’ (i.e., immigrants).

Overall, the theme Us versus Them reflected social identity processes as posters attempted to distinguish between different in- and out-groups (e.g., Blackwood et al., 2012, 2013, 2015; Reicher et al., 2008; Tajfel & Turner, 1979; Turner et al., 1987). In particular, posters attempted to align themselves with other posters whose views they shared (Us) and to differentiate themselves from those whose views they opposed (versus Them). In doing so, comments frequently touched on notions of ‘Britishness’ – and thus specifically on the issue of who should be included in this social category and who constitutes Us. This then, relates to Contesting Britishness, the final theme, to which we will turn next.

Contesting Britishness                                                

In the majority of posts, attempts were made to define what it means to be British by constructing and contesting different versions of British national identity). Here, posters did not simply try to position themselves in relation to Emma - instead, this theme indicates the extent to which respondents attempt to go beyond taking a specific stance on racism. In particular, this prevalent theme describes posters’ ambivalence about the notion of ‘Britishness’ or national identity content, by capturing the ways in which comments contested the criteria by which ‘Britishness’ should appropriately be defined. This is perhaps unsurprising especially as the video clip showed Emma challenging the British identity of a Black fellow passenger (i.e., referring to her as non-British). Also, given the diversity of YouTube users, one might expect variations in posters’ references to national identity and sense of Britishness.

Whilst endorsement of racist views and an anti-immigration stance need not necessarily go hand in hand, there frequently appeared to be a degree of overlap between the two. For example, here, respondents who positively embraced immigration did not always seem to differentiate between racism on the one hand and anti-immigration on the other. They therefore frequently referred to those who rejected immigration as racists.

Overall, respondents’ accounts of the criteria defining ‘Britishness’ ranged from being “multi-cultural", "tolerant", "accepting of others", "patriotic", "born in Britain", "White", to "holding a British passport", or rejecting the existence of ‘true Britishness’ altogether:

She doesn't have respect. No one is PURE english no more, not even the queen and the royal family is and I bet she isn't either. I'm mixed my father is english and my mum's thai, I was born here in the UK and I am a BRITISH. I follow every thing she does. .. I do the same as she, what just because My mum's from a different country? Just because these people are different colour means there not British.. HOW SHE WAS SAYING THIS IS MY BRITAIN.. NO BITCH THIS IS OUR BRITAIN. LIVE WITH IT! (Poster M)

Here, the poster clearly rejects Emma's version of ‘Britishness’ and, particularly, the view that anyone can be considered ‘purely’ British – including the ultimate symbol of ‘Britishness’, namely the Royal Family (ETHNOS, 2005). In particular, s/he rejects the idea that ‘Britishness’ can be determined by one's skin colour. In doing so, this poster not only distances him/herself from Emma's views, s/he also implicitly affirms immigrants’ and non-White people’s right to claim a British identity.

The following two sub-themes relate to the degree to which the boundaries of ‘Britishness’ are understood as inclusive of ‘others’. These were Britishness Under Threat, in which multiculturalism was claimed to be either the ‘undoing’ of ‘Britishness’ or, conversely, as its very essence, and “We are all Immigrants”, in which notions of 'Britishness’ were explicitly linked to Britain’s history.

Britishness Under Threat. The content of a number of posts seemed to express fears that Britain might be ‘losing’ its ‘Britishness’ due to immigration. In particular, these fears tended to relate to posters’ perceived inability/restriction to express their ‘Britishness’ (i.e., not being allowed to “fly the George Cross on St. George's Day”, or to “refer to Christmas”). In other words, here ‘Britishness’ is seen to be under threat. Along the same lines, some posters expressed a sense of identity threat by references to “cultural genocide” or “cultural cleansing” in relation to immigration. Here, such references create a sense of social identity threat (Branscombe, et al., 1999; Reicher et al., 2008; Verkuyten, 2013) amongst those supporting Emma whilst simultaneously invoking conflict with the perceived out-group (i.e., foreigners).

Interestingly, most posters who self-identified as English appeared to endorse a British national identity (rather than an English one). In particular, respondents who explicitly referred to their English heritage rarely voiced concerns about losing their Englishness –  instead, when national identity was seen to be under threat this tended to be expressed in terms of the British losing their ‘Britishness’. In addition, posters frequently differentiated between a British we/this country from a non-British them/other countries (Condor, 2000):

She has a valid point, albeit her language is really strong. The white, working classes in this country are becoming increasingly marginalised by the flood of immigrants from Africa, Asia and Eastern Europe.She's frustrated and angry. The present and preceding governments have failed her for years, and all she sees on public transport are ugly, threatening black and brown faces.Welcome to Tony 'War Criminal' Blair's vision of Britain for the 21st century. (Poster N)

Here the British government is reproached for allowing “the flood of immigrants” into Britain – the poster thus distances him/herself from the government’s vision of Britain. This distancing is further emphasised by stating that the government “have failed” the “White working classes” which, together with references to “ugly, threatening black and brown faces”, underscore the poster’s sense of threat to his/her national identity.  Also, the poster’s use of the term "marginalisation" is interesting here. First, the term conveys a sense of (relative) deprivation for which s/he holds immigrants responsible (see also above). Second, the use of this particular term here expresses a sense of identity threat by implying that White people have become a minority in Britain.

However, many other posters rejected this notion of Britishness Under Threat altogether. These posters argued in favour of immigration, suggesting that it enhances and enriches ‘Britishness’. This, in turn, was linked to calls for greater equal rights in terms of who should be allowed to call themselves British:

I'm British (white) and know I have no greater right to call myself British than those who became citizens today. I hope the ignorant minority of white nationalists will flow into our cultural past, but I doubt it. No society ever lives without ignorance, prejudice or racism completely. I would say I'm ashamed of this woman but I see no connection between myself an her just because of our race. Don't get frustrated by her ignorance, Britain's a multicultural society, an I'm proud of that. (Poster O)

Here, then, multiculturalism becomes the very essence of what it means to be British – which is in direct opposition to the views and behaviour depicted in the YouTube video. The poster clearly dis-identifies with Emma (i.e., "see no connections with her") and distances him/herself from White nationalists in general (i.e., by referring to them as "ignorant"). Rather than threatening British national identity, here multiculturalism is construed as defining ‘Britishness’.

Other posters argued that throughout British history different cultural groups came to Britain and therefore shaped ‘Britishness’. This not only gives people the right to call themselves British, it also undermines ‘Britishness’ as a fixed category. Some posters even suggested that “We (i.e., the British) are all Immigrants” – a claim which we will discuss next.

“We are all Immigrants”. Many YouTube comments explicitly and positively embraced immigration, especially given Britain’s colonial history. For example, some posters responded to those rejecting immigration by suggesting that, under British Empire rule, people from British colonies had not only been encouraged but, at times forced, to live in Britain:

Did anybody force the British to buy African slaves? By buying Africans, they made them British. The same history told me that the British used extreme force to capture several African settlements. They even sent the leaders of some of these settlements into exile. An example is King JaJa of Opobo. The British traveled for Months to look for Africa. They found Africa and raped her. They have to bear the consequences of their actions. No complaints is entertained from any British about Africans (Poster P)

This then raises the issue of who has the right to call him/herself British and bestow this identity on others. The specific argument here is that African descendants have the same right to call themselves British based on their ancestral history. Here, people who reject immigration are being reminded, that, although Britain’s colonial history might appear glorious to some, British history is rooted in the (bleak) past of those living in the colonies – which, in turn, here justifies immigration.

The YouTube comments also challenged Emma’s concept of ‘Britishness’ by referring to Britain’s pre-colonial history. For example, some posters undermined Emma’s notion of ‘Britishness’ by suggesting that:

if you look at the history of britain [we] are pretty much all immigrants… a mix of roman and saxons and norman and viking… further back you hav the celts and the picts… who came from somewhere else in europe… (Poster Q)

Again, the argument here is that ‘Britishness’ is, in fact, a function of the nation’s cultural diversity – and as this diversity is evident in ancient British history, no one should thus be considered as more or less British. The implication here is that there is no inherent or fixed essence to ‘Britishness’; instead, as Britain’s history and present continue to be shaped by cultural diversity and multiculturalism, what it means to be British evolves and expands. This position, then, lead to an advocacy of tolerance and diversity, which is reflected in the following extract:

There are good and bad in every race. Every one has a right to there own opinion but no one has the right to disrespect and abuse people based on there race and religion this is not freedom of speech. White people are not the only ones who contribute to this country every where you go you will see black and brown doctors, teachers, politicians, police officers and soldier’s many have died for this country. Do you know why? Because this is our country! We are different but equal! Peace (Poster R)

Overall, as a prevalent theme, Contesting Britishness encapsulated attempts to define ‘Britishness’– in other words, it captured efforts to define ‘us’. Specifically, respondents contested the criteria that determine who should and should not qualify as British, which often related to immigration. This contestation sometimes included perceived threats to posters’ sense of Britishness (i.e., social identity threat; Branscombe, et al., 1999; Reicher et al., 2008). However, for many others, Britishness constituted a broad, flexible and evolving social category with immigration/multiculturalism at its very heart.

In sum, the current analysis indicates that YouTube users’ responded to explicit hate materials by attempting to define or contest the meaning of posters’ national (i.e., here predominantly British) identity. Responses also frequently served to position users according to their stance on racism. As such, whilst some YouTubers responded in a racist, aggressive and abusive manner, the majority of posters did not automatically endorse, or raise their responses beyond, the level of hatred displayed in the video clip.

Discussion

The objective of this study was to explore how general users of a particular SNS (i.e., YouTube) responded to online materials which could be construed as explicitly inciting hatred. To address this objective, comments in response to a YouTube video clip which depicted ‘Emma’, a White female train passenger racially abusing other passengers, were analysed.

Thematic Analysis (TA; Braun & Clarke, 2006) identified four superordinate themes: (1) Making Sense of Emma, (2) Meeting Hatred with Hatred, (3) Us versus Them, and (4) Contesting Britishness. Making Sense of Emma depicted attempts to explain Emma and her behaviour by invoking both intra-personal and social level explanations (i.e., alcohol/drug abuse, lack of intelligence, lack of educational and/or low socio-economic status) which also served to distance posters from Emma. Meeting Hatred with Hatred captured ‘inflamed’ responses to the video clip’s content where hateful language was used, including calls for symbolic punishment, to express opposition to Emma and the views embodied by her. Us versus Them captured posters’ attempts to construct in- and out-groups in order to position themselves according to their own stance on racism. This positioning frequently occurred with reference to immigration and related to expressions of a sense of injustice and perceived relative deprivation. The final theme, Contesting Britishness, depicted attempts to define (or contest) the British national identity, especially in relation to Britain’s colonial history and immigration. Whilst some YouTube posters expressed a sense of their national identity being under threat, others suggested that there is no such thing as ‘Britishness’ – and, for some, immigration/ multiculturalism constituted the very essence of ‘Britishness’. These themes link to existing theories as well as previous empirical findings which will be discussed below.

Posters frequently attempted to distance themselves from Emma’s behaviour by positioning her as an out-group member (i.e., a racist). This, in turn, was achieved through invoking negative stereotypes (i.e., low intelligence, low socio-economic and single mother status) and the use of hateful, dehumanising language. Similar distancing strategies have been observed in previous work (e.g., Angie et al., 2011; Coffey & Woolworth, 2004; Douglas et al., 2005; Erjavec & Kovačič, 2012; Faulkner & Bliuc, 2016; Moghaddam, 2005; Mols & Jetten, 2014; Steinfeldt et al., 2010) where the use of (hateful) language has been described as a means of creating negative social comparisons and reinforcing existing stereotypes – or ‘othering’. Here, this ‘othering’ strategy served not only as a means of Making sense of Emma but, paradoxically, also led to Meeting hatred with hatred, as expressed in Emma’s dehumanisation.

The dehumanision of perceived out-group members has been observed previously as a common linguistic ‘othering’ strategy (e.g., Angie et al., 2011; Coffey & Woolworth, 2004; Faulkner & Bliuc, 2016; Moghaddam, 2005). Dehumanisation of perceived of out-group members has also been associated with attempts to foster a sense of moral righteousness or moral disengagement (Cammaerts, 2009; Faulkner & Bliuc, 2016) and the creation of a sense of in-group superiority (Bliuc, McGarty, Hartley, & Muntele Hendres, 2012; Mumford et al., 2008). For instance, Faulkner and Bliuc (2016) found that racists used moral disengagement strategies such as dehumanisation to justify expression of their views. Interestingly, in this study, dehumanisation was used by both those who supported or opposed racist attitudes to justify expressing their views. In line with this, some sub-themes (i.e., Sense of deprivation, Sense of injustice and Britishness under threat) echo some of Faulkner and Bliuc’s (2016) moral disengagement themes.

Furthermore, it has been suggested (e.g., Moghaddam, 2005) that one of the functions of dehumanisation is to enable in-group members to justify punitive treatments of individual out-group members and out-groups – which, in this study, was reflected in posters’ demands for Emma’s symbolic punishment. Similar calls for symbolic punishment have also been observed in other online hate contexts (e.g., Angie et al., 2011; Coffey & Woolworth, 2004; Doughty et al., 2014; Erjavec & Kovačič, 2012). Some posters’ attempts to counter Emma’s apparent racism with further expressions of hatred (i.e., Meeting hatred with hatred) seem to be in line with existing ‘flaming’ research (Coffey & Woolworth, 2004; Doughty et al., 2014, Lange, 2007; Moor et al., 2010). In particular, there was some evidence of ‘flaming’ (i.e., use of aggressive/abusive/racist language to oppose certain web content) in the online comments examined here. Specifically, posters appeared to use hateful language, or ‘flames’, not only to express their disagreement with the video clip, but also in order to mirror the norms conveyed in the video clip (i.e., as evidenced by Emma’s own aggressive, racist, insulting and hostile language and behaviour). This corresponds to previous research on online aggression which suggests that aggressive expressions on the Internet are often a reaction to perceived aggression from others (Law, Shapka, Domene, & Gagne, 2012).

The most prominent themes identified in the current analysis – namely, Us versus Them and Contesting Britishness – explicitly tapped into social identity processes and, as such, not only correspond to previous findings but also fit within a broader social identity approach (Blackwood et al., 2013, 2015; Reicher et al., 2008; Tajfel & Turner, 1979; Turner et al., 1987; Verkuyten, 2013). For example, Angie et al. (2011) relate the use of negative stereotyping and categorical thinking – as evidenced by ‘us’ versus ‘them’ – to social identity formation. In particular, Angie et al. (2011) found the use of categorical thinking more prominent in non-violent ideological groups than in violent ones. Categorical ‘us-versus-them’ thinking was also observed by Blackwood et al. (2012), Bliuc et al. (2012) and Reicher et al. (2008). In line with previous work, in order to confirm their identification with like-minded respondents, or their group identity (i.e., here, non-racists), many posters in this study adopted similar categorical thinking strategies to those identified in previous work (e.g., Angie et al., 2011; Charteris-Black; 2006).

Posters who supported Emma’s behaviour also distinguished between ‘us’ and ‘them’ – specifically by using terminology that depicted ‘us’ as good and ‘them’ as bad. This, too, corresponds to previous observations that positive self-representation is often pitched against negative other-presentation (Douglas et al., 2005; Verkuyten, 2013) in order to create negative social comparisons which, in turn, aim to establish a sense of legitimacy and to persuade like-minded others (Charteris-Black, 2006). Here, negative other-presentations were frequently linked to expressions of a sense of injustice or relative deprivation. For example, posters who endorsed Emma’s views tended to claim that immigrants and ethnic minorities were in receipt of preferential treatment by the State or by the police - which, in turn, mirrors previous findings (e.g., Coffey & Woolworth, 2004; Douglas et al., 2005; Mols & Jetten, 2014). These posters also invoked the notion of socio-economic insecurity as a consequence of immigration; attributing blame to immigrants has previously been linked to scapegoating, disgruntled sentiments, prejudice and hate crimes towards minority groups (Bowling & Phillips, 2003; Sibbitt, 1997; Young, 1999; for a detailed review see Walters, 2011).³

Taken together, these findings closely relate to the concept of social identity threat (Branscombe, et al., 1999; Reicher et al., 2008, Verkuyten, 2013) which derives from the social identity perspective (e.g., Blackwood et al., 2012, 2013, 2015; Reicher et al., 2008; Tajfel & Turner, 1979; Turner et al., 1987). Whilst there are different types of social identity threat, the current findings relate most closely to what Branscombe et al. (1999) have termed ‘distinctiveness threat’ (i.e., the fear that ones’ group distinctiveness is prevented or undermined). This suggests that perceived threats to ones’ group’s identity may lead to discrimination including expressions of overt racism and hostility (Pereira, Vala, & Costa-Lopes, 2010; Reicher, et al., 2008; Verkuyten, 2013).

In line with previous findings (e.g., Faulkner & Bliuc, 2016), posters in this study, who supported racist views, attempted to justify their opinions by expressing a perceived sense of threat to the distinctiveness of their social (i.e., national) identity (Branscombe et al., 1999) on the basis of immigration. It has been shown previously that one of the responses to multiculturalism can be the expression of a perceived (national) identity threat (Mols & Jetten, 2014; Sindic, 2008). Issues relating to perceived social identity threat were also evident in our final theme – namely, Contesting Britishness. In particular, posters who supported racist views invoked as sense socio-economic threat to the White British working class.

Another way in which the themes Us versus Them and Contesting Britishness tapped into social identity processes relates to YouTubers’ contestation of their national identity as British. In particular, Faulkner and Bliuc (2016) previously observed how Australians disputed in online newspaper comments the perceived values and characteristics considered representative of their national identity. In addition, differences in peoples’ versions of ‘Britishness’ have been identified previously (e.g., Condor, 1997). In particular, differing accounts of ‘Britishness’ are frequently found in political discourse where they serve as rhetorical tools to influence peoples’ interests and behaviours according to political agendas (Sindic, 2008). Therefore, as political agendas change, so might the boundaries/definitions of ‘Britishness’.

Along similar lines, national identities are not fixed and, as Cohen (1994) has argued, their boundaries are generally ‘fuzzy’. Thus, national identities are continuously re-defined through interactions with different 'others' (Condor, 2000; Hopkins & Murdoch, 1999).

Exposure to online hate materials and escalation

Coffey and Woolworth (2004) observed that some online discussions can be characterised by escalation (i.e., in terms of racism, aggression, abuse). However, in the context of the present study, the authors did not find any straightforward evidence of escalation; instead, posters’ responses seemed more complex and, as discussed above, often contested relevant social categories (i.e., ‘us’ versus ‘them’, ‘Britishness’). One reason for the lack of evidence of escalation here is likely to be found in the relative short time-span (i.e., 48 hours) posters had available to respond to the video clip (i.e., before the clips’ removal from YouTube).

Another likely reason is the asynchronous nature of the responses under investigation here and the corresponding non-sequential way in which posts were displayed on YouTube: unlike in Coffey and Woolworth’s study (2004), given the volume of simultaneous responses to the YouTube video, posters were rarely able to respond sequentially (i.e., directly) to specific others. Some respondents attempted to overcome this issue by creating a response sequence addressing other posters directly (i.e., starting their responses with an '@' and indicating the recipient’s username). Yet, by the time their posts were displayed, given the sheer number and speed of postings, these attempts at creating response sequences would have been hard to identify visually (i.e., they would have appeared ‘lost’ amongst hundreds of other comments).

The potential differences between synchronous versus asynchronous communication and user density in online spaces have been highlighted previously (Wojcieszak, 2010). In particular, Wojcieszak (2010) questioned whether online response synchrony and user density would facilitate or prevent people from responding in a more polarised (i.e., escalated) manner over time. Thus, the relative short time, asynchrony and density of the comments analysed here might, in part, have prevented further polarisation and escalation of hateful responses. The specific relationship between response time, asynchrony, density and escalation of hateful responses therefore warrants further investigation in order to clarify the possible factors contributing to polarised responses.

Finally, whilst collating the online comments, it was observed that several respondents were removing their comments from the website. This removal seemed prompted by other respondents’ appeals to the service providers (i.e., YouTube) or the police to delete or investigate certain comments for their perceived offensive and illegal nature (i.e., hate speech and/or incitement to racial hatred). Therefore, the removal of posts seemed related to posters’ concerns about potential legal consequences and a decreased sense of online anonymity. More importantly, the observed removal of posts suggests that a certain degree of self-policing is practiced online.

It was also observed that some posters blocked or spammed others who expressed racist views, as well as advised others to do the same (via commenting). Consequently, responses by blocked or spammed posters became no longer visible to anyone on YouTube, suggesting that online hate materials can, in fact, elicit self-censorship. This self-censorship, in turn, seems to further support the finding that social networking users do not always automatically endorse online hate materials.

Limitations

First, as with qualitative work in general, one of the limitations of this study lies in the fact that the current findings cannot be generalised (e.g., across all online hate materials or different SNS). Here, responses to a certain, racially motivated hate incident on one specific SNS (i.e., YouTube) were explored. The authors therefore cannot claim to explain hatred across all SNSs or all hate-motivated incidents, such as those based on disability, faith, gender, transgender or subculture. Instead, the broad focus of the current research question aimed to provide a starting point for research by exploring the nature of online replies to explicit hate materials. In particular, the present study aimed to shift the currently dominant ‘end-product approach’ (i.e., already polarised individuals/groups and their content/interactions) to a focus on the potential processes underlying the development of polarisation in the context of online hatred. We found that whilst online racism can lead to some degree of polarisation, the majority of responses did not suggest polarisation per se, but instead indicated social identity processes at play (i.e., reflecting the construction of in- and out-groups, illustrating respondents’ group affiliations).

Second, whilst one of the study’s strengths lies in the realistic nature of the data, this observational approach prevents the authors from making inferences about the representativeness of the responses (e.g., the sample might not represent users across other SNSs). The authors neither had access to respondents’ demographics, nor could they determine whether respondents posted with multiple or different user names (i.e., identities). Yet, given the substantial number of comments in response to the video clip (i.e., over 71,000 posts), it is unlikely that these were exclusive to a specific (demographic) group – instead, there was evidence of some diversity in backgrounds (e.g., gender, ethnicity, geographic location).

Finally, as in previous studies (e.g., Coffey & Woolworth, 2004), the findings do not allow the authors to draw inferences about the effects that the relatively short period of time in which people were able to respond to the video clip might have had on the content of their posts. In particular, it remains unclear whether longer-term online interactions result in repeated and/or increased expressions of hatred or whether long-term interactions/discussions change direction and automatically lose focus, therefore reducing or diverting expressions of hatred.

Conclusion

Expressions of hatred (e.g., racism) are increasingly becoming a feature of life online, especially on social networking sites (e.g., YouTube). It is therefore becoming increasingly important to understand how ‘everyday’ online users respond to and engage with such hateful materials. Many YouTubers who responded to the hateful online material analysed here condemned its content. As such, the findings obtained in the current context indicate that online hatred may not always be automatically endorsed by online users; instead we observed a degree of contestations in online users’ responses. Whilst the current findings cannot be generalised, we nevertheless hope they provide a springboard for future research which should take into account not only a wider variety of hate materials but also different social networking sites.

Notes

1. This comment was made in response to another comment, expressing opposition to the poster’s views.

2. Please note that this comment was also made in response to another comment. In this instance, the comment was aimed at a poster who shared the views of this particular poster.

3. At the time of the video clip's upload in October 2011, Britain was in an economic recession and characterised by socio-economic insecurity. This may therefore have exacerbated posters’ sense of relative material and financial deprivation.