Racism is not just hate speech: Ethnonationalist victimhood in YouTube comments about the Roma during Covid-19

Abstract

Research shows that racism and xenophobia soared during the Covid-19 pandemic and this was certainly the case with the Roma in Romania. In this article, using critical discourse analysis, we analyse comments left below a television news clip posted on YouTube early in the crisis. This gives us valuable access to the way racism and xenophobia are linguistically expressed in social media, particularly in this Romanian context. It yields insights into how more overt forms of racism can sit alongside others which are less so, all united by a sense of shared embittered victimhood on behalf of Romanian citizens. We show how this takes place as the affordances of social media allow for a collective expression of frustration and mobilisation, reflecting on how social media may increase exposure to more extreme forms of racism. (Critical discourse analysis, Covid-19, online racism, Roma, social media, white victimhood)

Introduction

At the time of writing, the Covid-19 pandemic has impacted the lives of people worldwide, especially those from poor and marginalised communities (Hamiduzzaman & Islam 2020; Millán-Guerrero, Caballero-Hoyos, & Monárrez-Espino 2021). This was certainly the case among the Roma in Romania where the pandemic starkly revealed existing social inequalities between the Roma and non-Roma, making racism and xenophobia more visible and pushing already vulnerable communities into deeper levels of marginalisation and social exclusion (Creţan & Light 2020; Matache, Leaning, & Bhabha 2020). Scholars and activists drew attention to increased hate speech directed at the Roma during the pandemic (Plainer 2020; Matache & Bhabha 2020). The Roma were blamed for the spread of the virus because of lack of self-discipline or for not complying with prescribed health and hygiene instructions (Berta 2020). Such views were found across national news and in social media, which carried accounts of returning Roma migrants spreading the virus and posing a health risk to the entire country (Costache 2020; Matache & Bhabha 2020). While there was no evidence that the virus was spreading faster among Romani communities, such fears resonated with deeply rooted stereotypes about the Roma's poor hygiene habits or their image as irresponsible, disruptive citizens (Tileagă 2015; Berta 2020). In this context the Romanian government issued strict lockdown and quarantine orders for several towns inhabited by Roma communities (Creţan & Light 2020).

In this article we analyse the comments posted in response to a news report uploaded on YouTube about the lockdown of one Romanian town with a large Roma population. The town of Ţăndărei or ‘the time bomb’ (Gândul 2020), as it was labelled at the beginning of the pandemic, is a Roma community in Romania which, over time, has gained a negative image in the media due to ongoing allegations of criminality, including child trafficking, instigated against rich members of the Ţăndărei community (Pitu 2020; Stoica 2020).

Because of the already negative image of Ţăndărei across Romanian media, forceful policing and use of the military to enforce the lockdown were not questioned but regarded as natural and legitimate (Plainer 2020). The town was at the same time subjected to abuse and hate speech on social media. One such example, which drew international criticism, was a meme posted on Facebook by a well-known Romanian scholar (Jipa 2020; Mutler 2020). The meme showed a group of crows resting on a fence, the top headline reading ‘Ţăndărei Airport’ and the bottom headline ‘All flights cancelled’. The association between Roma and crows is a well-established racist metaphor in Romania (Breazu & Machin 2019).

While the case study we use in this analysis may not characterise the treatment of all Roma during the pandemic, it provides an opportunity to examine the YouTube comments for one particular news report showing the police and military enforcement of the lockdown in Ţăndărei.

Using critical discourse analysis (CDA), we reveal the different realisations of racist ideology in the comments. We find more blatant racist hate speech. But many comments lack this. Rather they can be characterised as what scholars have described as new racism where overt racism is avoided in favour of referring to ethnic minorities in terms of social or cultural problems or as an economic burden (Ansell 2016). Yet running across all comments is a discourse of a sense of shared embittered victimhood on behalf of Romanian citizens. This phenomenon of ‘reverse victimhood’, or inflicted white injury, has been well documented in various contexts (Bonilla-Silva 2006; Bloch, Taylor, & Martinez 2020).

Research on racism and social media has pointed to the need to reveal and understand both blatant and more subtle forms (Ben-David & Fernández 2016). In Romania, in particular, scholars have discussed racism toward the Roma in relation to ethnic nationalism and right-wing populism (Turda 2007; Dobrincu 2019), yet this specific discourse of victimhood has received less attention (Vamanu & Vamanu 2013). Given Roma's long history of discrimination in Romania, the analysis here allows us to both look at how racism is realised in different forms on social media comments and also to shed light on the specific nature of ethnonationalism and victimhood in Romania.

This attention to less overt expressions of racism is also crucial given that there is a tendency for monitoring processes of major social media platforms to flag racism only when it takes explicit forms of hate speech. Siapera & Viejo-Otero (2021) argue that this is a kind of decontextualised approach to tackling online racism, since computational tools are designed to detect only words and expressions easily identifiable as racist, thus overlooking those subtle forms, such as ridicule, sarcasm, or reverse victimhood—all of which we consider in the present analysis.

While it may be the case that some who leave comments are simply more careful as to how they formulate their views, the point made by Matamoros-Fernández (2017) is relevant here in that such platforms tend to bring users into networks of increasing intolerance. There is a danger they may be likely to encounter more extreme versions of their existing views (Heikkilä 2017). Certainly, across these comments there are no interactions between those commenting as to what is a reasonable view to hold. Yet, as we show, the discourse of embittered victimhood provides coherences across them.

The case study we analyse in this article is taken from a larger corpus of YouTube videos uploaded after the point that Covid-19 had become a global pandemic. We provide a summary of the news report and provide examples of comments that allow us to illustrate the main themes found across our corpus. Before we get to this, we provide two bodies of literature that help us to contextualise and understand the nature of this data. The first creates context as to the situation of the Roma in Romania and in relation to the Covid-19 pandemic. The second looks at existing research on new racism, white victimhood, and nativism.

The Roma in Romania and Covid-19

The Roma have a long history of discrimination and social exclusion in Romania and have always been regarded as a ‘problem’ that needs to be solved (Achim 2010). In Romania, the Roma's history has been marked by appalling treatment, including 500 years of slavery on Romanian territories (Guy 2001; Marushiakova & Popov 2009; Achim 2010) as well as genocide during Romania's alliance with Nazi Germany (Hancock 1997; Friedman 2010; Kelso 2013). The discrimination continues today as we witness camp evictions, deportations, and the social exclusion of the Roma across Europe (Breazu & Machin 2018; Rorke & Mangiacavallo 2018). In Romania, in particular, these anti-Roma measures are often viewed as further evidence that the Roma are a problem that needs solving (Breazu & Machin 2020).

Public discourse about the Roma in the Romanian context has been largely negative. Most media coverage represents the Roma as a community which is either poverty-stricken or defined by extreme, spectacular wealth (Breazu 2020). Yet irrespective of socioeconomic status they are collectively associated with illiteracy, criminality, social disorder, unconventional lifestyle, and an unwillingness to follow rules and regulations (Creţu 2014; Breazu & Machin 2018, 2019). Such representations need to be understood in relation to the long history of discrimination and social exclusion (Tileagă 2015) and to the longstanding ethnonationalism in Romania which excludes the Roma (Turda 2007).

Since the COVID-19 health crises, Roma communities across Europe have found themselves targeted by authorities and vilified in the media, which exacerbated their already precarious socioeconomic position (Matache & Bhabha 2020). The featuring of the Roma in news and social media painted a picture of an irresponsible community that refuses to follow the rules of social distancing and isolation, which intensified the racism and negative attitudes (Berta 2020; Plainer 2020). However, EU reports show that, as of 2020, the Roma live in the poorest areas, often in places with no electricity, gas, running water, or paved roads, which in turn, limit or deny their access to education, job markets, and proper health care (FRA 2020). For the Roma, the lockdowns presented additional challenges, such as lack of available work and help from authorities to cope with daily expenses and needs, unequal access and treatment, and health care, which left many without medicine or proper assistance (Korunovska & Jovanovic 2020). Such socioeconomic inequalities have been generally absent in media representations of the Roma or rather seen as further evidence of their problematic nature (McGarry 2017).

The increase in scapegoating of the Roma in relation to Covid-19 related to a number of specific issues. First it was linked to large numbers of returning migrant workers from geographical regions that has recorded a high Covid-19 related death toll (Creţan & Light 2020). While this included a mixture of high skilled and low skilled individuals, both Roma and non-Roma, news media focussed only on the potential danger presented by returning Roma (Matache & Bhabha 2020; Plainer 2020). This has particular relevance for the YouTube example we analyse here.

Another important factor is what Berta (2020) calls the ethnicization of the pandemic, as the Roma were assigned the blame for spreading the disease based on existing deeply rooted stereotypes about Roma's lack of hygiene and anti-social behaviour. This was the case even though in news media reports such assumptions were never supported with actual evidence (Matache & Bhabha 2020).

Racism, nativism, and White victimhood

The news media have been acknowledged to play an important role in how the broader public understands race (van Dijk 1991). Ideas about racial superiority or inferiority have persisted for centuries, yet contemporary discourses about race are reproduced in fusion with other discourses about culture, ethnicity, religion, nationality, gender, sexuality, or class (Entman & Rojecki 2001). As a complex system of domination and inequality (van Dijk 1991) or as a poisonous ideology (Husband & Downing 2006), racism has been discursively manifested in various forms that go beyond invoking skin colour or other biological characteristics as markers of difference. Racism is not simply a monolithic ideology but must be understood in historical material terms where it has existence in the specific and unique practices of oppression and social exclusion in different societies at different times (Camfield 2016). Racism needs to be understood, therefore, in relation to how inequalities are historically shaped in various contexts and how it intersects with other social and cultural practices that serve to perpetuate it (Doane 2017). Indeed, there is a danger of collapsing all forms of racism into one thing that cuts it off from its ‘historical basis, severity and power’ (Song 2014:107).

Scholars have observed that in many societies open and direct expressions of racism are no longer acceptable (Augoustinos & Every 2007). This does not mean that racism has disappeared but rather that it takes more disguised, managed, and subtle forms (Simmons & Lecouteur 2008; Nelson 2013). Blatant racism, associated with biology, eugenics, and racial inferiority, has, therefore, been replaced by what is known as new racism (Barker 1981; Ansell 2016). In these new forms, racism can be disguised as cultural evaluations where ethnic minorities are said to constitute a threat to national culture, traditions, and values (Bonilla-Silva 2006:29). Racism can also be concealed as economic concerns about minorities taking jobs away, or about their being a burden on the welfare systems (Augustinos & Every 2007). In all such accounts, these are presented in ways that appear rational and justified rather than as prejudice (Augustinos & Every 2007). In critical discourse studies there has been an interest in how such discourses have become normalised along with an overall rise in right-wing populist ideologies (Krzyżanowski, Triandafyllidou, & Wodak 2018; Krzyżanowski 2020).

New racism often intertwines with new nativism (Sohoni & Mendez 2014; Bonilla-Silva 2015). This nativism is bound up with the idea of the rights of an imagined, coherent, native population who are said to suffer economic, social, and cultural injustices because of ethnic minorities, which are regarded as privileged. This imagined group often appears as victims, while migrants and ethnic minorities are said to gain unearned privileges by simply playing the ‘race card’ (Huber, Lopez, Malagon, Velez, & Solorzano 2008). Here the very idea of liberal meritocracy is seen to be undermined as advantages can be gained, not by individual merit, but due to ethnicity (Bloch et al. 2020). This reinforces the idea that ethnic nationals must also fight for their rights (Huber et. al 2008), and this process has been discussed in terms of reverse victimhood or reverse racism (Doane 2006). As part of this victimhood, anger is also directed at the perceived, out-of-touch political elite that permits these injustices against nationals to persist, by favouring the race card played by ethnic minorities against the interests of their own people (Bergmann 2020).

In Romania, this idea of new nativism must be understood in relation to ethnic nationalism that dominated public discourse from the beginning of the twentieth century (Turda 2007; Achim 2010; Tileagă 2015) where the Roma have always been seen as the ‘other’. This has a longer history in their 500-year enslavement or their more recent persecution during the WWII, when the Roma, including most middle-class intellectuals, were systematically mass murdered under the fascist dictatorship of Ion Antonescu (Kelso 2013; Schmitt 2018). The end of communism in 1989, in particular, was followed by a rebirth of ethnonationalism, when racism and intolerance toward the Roma became rampant (Schneeweis 2009). This was fuelled in part by the rhetoric of scapegoating the Roma by politicians when addressing frustrated citizens, dissatisfied with the poverty, lack of jobs, and declining economic prosperity (Breazu 2020). This history is important for understanding the localised discourses of nativism and victimhood found in the YouTube comments analysed in this article.

Data and methodology

In critical discourse studies (CDS), as well as media and communications studies, social media offer a huge potential for data that can allow us to study both the ideologies of people as they post about issues and also how they respond and negotiate with media content such as news programming (Pink 2016; KhosraviNik 2017). In CDA, we need to carry out more research of such social media to understand how racist ideologies are now expressed and exist in our societies in different ways (KhosraviNik 2017).

In this article, we look at comments posted on YouTube in response to a news clip from Romanian television about the Roma and Covid-19 lockdowns. We focus on a case study of a news report about police intervention during the lockdown in the town of Ţăndărei—identified by human rights groups as one case affecting the Roma and where excessive force was used (McLaughlin 2020). We searched for all content posted on YouTube that made reference to Ţăndărei and Covid-19. The search resulted in thirty news clips. For the analysis in this article, we selected one that was uploaded by one of the leading audio-visual stations in Romania (PRO TV) and that gained the largest views (135,735) and number of comments (442). This example allows us to reveal the discourses found in the comments across our broader sample.

The analysis was carried out in two steps. First, a thematic analysis was conducted. All comments posted under the YouTube clips were imported into NVivo, a qualitative software tool for organising, classifying, and sorting data. It also has a variety of affordances that facilitate visualisation of the relationships within the data and the themes they carry. The purpose of this analytical step is to map out the thematic nature of the content following the principles of van Dijk's (2009) semantic macro-analysis. We first started with open coding of the data before moving to axial coding, which refers to the breaking down of major codes into categories and subcategories (Charmaz 2006). This step of analysis is instrumental for mapping out the most salient themes in the data. The first part of the analysis below presents an overview of these themes.

The second stage of the analysis involves an in-depth critical discourse analysis (CDA) of each theme that emerged from the posts in response to a news clip posted on YouTube. CDA is a well-established tradition within the broader umbrella of critical discourse studies, which seeks to examine the role of language in social and political life and in particular how events, persons, things, and processes are represented in ways that legitimise and uphold various inequalities based on race, gender, or access to economic resources (Flowerdew & Richardson 2017). CDA can interrogate texts, or collections of texts, for the discursive themes they carry (van Dijk 1998).

In critical discourse studies, scholars have approached media discourses in terms of the recontextualization of social practices (van Leeuwen & Wodak 1999). The basis of this notion is that the discourses (Foucault 1972) carried by news media, or in other communication, can be thought of as comprising ‘scripts’. These scripts form the accounts of what is going on in the world, who is doing what, for what reasons, with which kinds of priorities. They are comprised of elements such as participants, actions, performance modes, evaluations, presentation style, time, and place. In this sense, media texts or social media posts comprise a kind of ‘script’ about participants who are engaged in actions in specific contexts and who are driven by specific priorities. For van Leeuwen & Wodak (1999) the aim of the analyst is to identify the extent to which such discursive scripts recontextualise actual events and processes taking place. A close analysis of representation in language can reveal where actual events, causalities, processes, identities, and so on have been deleted, substituted, abstracted, added, or re-sequenced. This form of analysis is particularly suited to the analysis of the discursive scripts found in the YouTube comments to account for the Roma, Romanians, and politicians.

The CDA analysis is presented under headings that comprise the themes identified through the semantic macro-analysis. Then in each case we carry out closer analysis of the language and grammar. Before we proceed with the analysis of the comments, we provide a brief account of the news report. The aim here is not to carry out a full analysis of the report but to explain its basic narrative and representations of the Covid-19 situation, the lockdown in the village, and the Roma, which is necessary for establishing the context for understanding the comments we analyse below. The language in the comments and the discourses they carry makes the object of enquiry in this article.

The news report posted on YouTube

The headline on the YouTube post was ‘Ţăndărei was quarantined. Cat and Mouse game between residents and police’. This sets the scene for the clip, that the Roma are not following the lockdown and are evading the police who seek to enforce it.

The news anchor explains that a quarantine has been set up in Ţăndărei, that seven people have died there from Covid-19, and that ‘hundreds of locals’ have been returning from around Europe because of fear of the virus and even more are on their way: ‘a new wave is expected’. As statistics are given about the status of the virus in Romania, we see footage of people in hospital undergoing tests, busy hospital corridors and labs, ambulances transporting sick people to hospital or carrying coffins. We then cut to a street scene in Ţăndărei, where vehicles speed past with sirens sounding and armed police stand in the streets. The voice-over tells us that after desperate requests from the local mayors, the government decided to quarantine the entire town. We then cut to an onsite reporter, dressed in a Hazmat suit, who drives around the village pointing out people, small family groups, or children on bicycles, who are described as not following the lockdown instructions. The clip shows, therefore, that the Covid-19 situation is terrible, with hospitals overwhelmed; then we see the drama in Ţăndărei, where locals are returning from Europe, bringing with them the threat of a second wave, and where they are not respecting the quarantine, hence the need for enforcement.

The report then cuts to a customs checkpoint at the Romanian border where we see a long queue of cars. The voice over tells us they are to be escorted to official quarantine sites. This footage is in fact taken from a Facebook live streaming of a man whom we see speaking and joking.

Look bro, how many we are here. We'll all be quarantined. They'll take me to a hotel of three stars [laughing], as they said. We are over 500 people, only on this side of the road.

No overt connection is made between the village and this scene, but the implication is that this is part of the same sequence of events: of Roma returning from abroad, of them not following quarantine, and the danger of a wave of Covid-19 and the scenes of the busy hospital with coffins. The clip of the man joking for his friends on Facebook, presented in this context as not taking the lockdown seriously, appears to align with the same disregard shown in the village. This man and his comments play an important role in the comments posted.

The news report aligns with previous research on the Romanian news media where the Roma are represented as impossible to integrate into the wider Romanian society due to the extreme unruly nature of their culture, that they comprise a threat to social order (Erjavec, Hrvatin, & Kelbl 2000; O'Nions 2010; Nacu 2011), and that there is almost a war-like relationship between the Roma and non-Roma population (Erjavec 2001; Cheshmedzhieva 2009).

Thematic analysis

We first look at the main themes identified through the semantic macro-analysis. It allows us to consider how comments align alongside or against the discursive script laid out by PROTV's news clip. The comments on this page are for the most part negative, with few voices who speak up for the Roma. We see clear examples of explicit racism, as well as other forms that align with what scholars have described as new racism (Ansell 2016). Running through all of these themes is the discourse of victimhood on behalf of those leaving comments.

We identified six main themes.

1. (i) Roma as an economic and cultural burden. This theme comprises a set of discourses that relate to Roma being unproductive, untrustworthy and criminal, uninterested in mainstream forms of employment, and often dependent on social welfare. This contrasts with self-presentation of those commenting as hard-working, responsible, and tax-paying who receive no such favours.

2. (ii) Meritocracy and race card. This theme includes typical populist discourses about Roma receiving privileges based on their race and not on merit. There is a sense of Romanian injury where those posting emphasize that the hard-working, responsible non-Roma are underprivileged, while the Roma manage to gain unearned benefits and get away with wrong doings, simply by playing the race card.

3. (iii) Corrupt and incompetent elite. Politicians and national and international institutions are said to be out of touch with the challenges faced by ordinary Romanian citizens. Yet this elite favour the Roma, either through incompetence, naivety, or corruption. This includes turning a blind eye to Roma's wrong-doings and presenting the Roma with a range of benefits and favours.

4. (iv) Roma are an embarrassment for Romanians. The bad behaviour of the Roma across the EU has had consequences for ordinary Romanians who also, therefore, become victims of prejudice and discrimination as they seek work opportunities. This discourse is realised as part of a focus on representing the Roma as not being Romanian.

5. (v) Romanian people are not allowed to speak out. Where Romanian people do seek to give true accounts of the situation, they are dismissed as racist.

6. (vi) Nativism, biological racism, and eugenics. While most discourses we find on this forum can be associated with what scholars refer to as new racism, this forum allows for more explicit manifestations of racism that clearly refer to biology and eugenics.

A critical discourse analysis of the YouTube comments

Economic and cultural burden

To begin with, we find that the discourses are typically associated with new racism; the Roma constitute an economic burden on the country and that they are a nuisance due to extreme and problematic cultural differences.

1. (1) It is our fault because we pay child allowance from the state money to all those who have 5 children… Let's put them to work hard… you'll see how beautifully Romania will flourish.

2. (2) I would put everyone in construction and agriculture… Including those from gangs… Why put them in jail so that we pay for everything… Isolation at Home… and work for the benefit of the state… Foot chains… with GPS… and you'll see how good they turn to be after 3 years of hard work.

In comments (1) and (2), we see typical populist discourses about ethnic minorities (Bergmann 2020) as the Roma are represented as a drain on taxpayers: ‘we pay child allowance’, ‘we pay for everything’.

Merging with new racism, the more ‘reasoned’ part of the account, here is the explicit provision of a solution by pressing the Roma into labour, which includes ‘GPS’ and ‘Foot chains’. This social-engineering solution to the burden presented by the Roma runs through our corpus of YouTube comments and is never challenged by others. In this discourse there is no reference to the cycle of marginalisation that keeps Roma outside of the job market. Foregrounded is the victimhood carried by the taxpayer.

We also find linguistic features that are relevant to the polarity set up in the comments, between the Roma and non-Roma. This sense of community and mobilisation created across the comments, which is highly characteristic of social media (Bouvier 2020), becomes clear as we develop the analysis.

In CDA, pronouns such as ‘we’, ‘us’, and ‘you’ can be used to express and manipulate social relations (van Dijk 1998:203). Here the ‘we’ in particular can be undefined and includes a complex and shifting range of contradictory views (Machin & Mayr 2012). Yet in the context of such a sequence of comments, this perception of ‘we’ comprises an important part of the engagement or of what is called the ‘affective connectivity’ (Papacharissi 2015)—meaning a sense of community or shared purpose that is not so much driven by clear and concrete issues and aims but by emotional engagement (Bouvier 2019). The clear and simple polarity between ‘we’ and ‘them’ has been observed to be highly typical of simplified narratives found on social media feeds where there is incivility (Bouvier & Way 2021). In comments (1) and (2) above, we find: ‘we pay child allowance’, ‘we pay for everything’. There is continual reference to this affective ‘we’ and ‘us’, which here are clearly pointing to who is carrying the burden, who are the actual victims.

In these posts we find a further feature of the feed, relating to how the ‘we’ forms a sense of collective mobilisation of the affective sense of victimhood, which is the use of simulated conversations. In comment (1) we do not find a simple statement of an opinion, as in ‘I think we should put them to hard work’, but such discourse is formed as an ongoing interaction with others. So, for example, comment (2) starts with the question: ‘Why put them in jail so that we pay for everything’. Then we have the answer provided in ‘you'll see’. On social media feeds there may be little actual dialogue or clear interactions between those posting (Bouvier 2020). But these simulated interactions form a part of the affective connectivity, that there is rallying of the ‘we’, that the ‘we’ is in dialogue. This forms an important part of the sense of shared victimhood and the ‘voice’ of those speaking out together.

Comment (3) is also about the burden of the Roma but has another important feature.

1. (3) They have just returned from begging and they want to stay in hotels which are at least 3 stars.

In CDA it has been shown how persons or groups can be negatively represented and discriminated against through the attribution of them to specific mental processes (Machin & Mayr 2012). Such attribution evaluates them not only on the basis of what they do, but by imposing intentions and states of mind upon them. Here the Roma are represented across the comments as having this sense of entitlement: ‘they want to stay in hotels which are at least 3 stars’. Such mental processes play an important part in these simplified narratives on social media feeds, where those commenting create a kind of shared mental representation of the evil-doer (Suler 2004). Here there is a sense of outrage as the unproductive, burdensome Roma also feel entitled.

In contrast the non-Roma Romanians are represented as hard working with no sense of entitlement, as we see in comments (4) and (5).

1. (4) Logically. Why am I still in Germany? Aaaa, because I've been working here legally for 6 years and I pay 1200 € in taxes every month?

   P.S. Our company is still running, even on Saturdays, and last Sunday, we worked 5 hours…

2. (5) You're right and I'm in exactly the same situation as you say I've been working for the company for 10 years, including now at this time of COVID. Plus I'm also doing a part-time job to gain extra money, yet these ones with grimy faces do not like to work….

In CDA it is highly useful to analyse what is called the ‘representation of social action’. This simply means looking at the verb processes attributed to different participants. For the non-Roma here, such verb processes foreground productivity (‘last Sunday we worked 5 hours’; ‘Plus I'm also doing a part-time job to gain extra money’) and being good citizens (‘I pay 1200 € in taxes every month’). This idea of productive and disciplined citizens is important in the discourse of challenging the perceived entitlement and unruliness of the Roma that we come to shortly. This discourse of burden also relates to the unruly way of the Roma, who appear not to obey laws, as see in the following comments.

1. (6) Go online and read the press from abroad, the criminals, the prostitutes have returned. Maybe 10% of them will be among the honest ones, the rest of them are the shame of Romania. If people abroad stay at home [because of the lockdown], from where else can the despicable ones steal? I am sorry for those who are forced to bear them.

Again, these comments are formulated as simulated conversation as an instruction to others: ‘Go online’, as if there they find the evidence and facts behind the expressed opinion. Important in comment (6) is the statement: ‘I am sorry for those who are forced to bear them’. While above in comment (3), we see the mental states of entitlement attributed to the Roma, here we find a relatively humane set of feelings, as the comment expresses concern or regret for those exposed to the Roma. One observed element of new racism is that speakers represent themselves as reasonable, decent, and considerate (Capdevila & Callaghan 2008). Many comments in our corpus show that people express other kinds of worries, fears, and support for those suffering due to the Roma. There is a humanisation of the non-Roma here and a corresponding dehumanisation of the Roma.

We also notice the overt biological racism which surfaces through the comments in our corpus, converging with new racism: in comment (6) the ‘despicable ones’ and in comment (7) below ‘this stinking ethnicity’. The reasoning and ‘evidence’ of new racism are present, but so is the more blatant form of the ideology of hatred that this carries with it.

Meritocracy and the ‘race card’

In addition to the discourse of the economic burden and cultural threat, comments point to the unjust breakdown of meritocracy through the way the Roma are given privileges. Here those commenting draw on what Bonilla-Silva (2015) refers to as a kind of abstracted notion of liberal society, where all are judged and have equal opportunities based on individual merit, where no special treatment should be given based on race. We see this in comment (7).

1. (7) Mister, please do not discriminate against them. Apart from this stinking ethnicity, there are also many Romanian criminals [reference to politicians]. If they [the Roma] are “eliminated” , who will occupy the free places at universities, which is discriminatory against young Romanians who want to achieve something in life… it is sad that we have a country led by corrupt and incompetent people, who favour an ethnicity from which very few people rise

This comment takes the form of a simulated conversation, performed with irony: ‘Mister, please’. Here we find the view that young Romanians experience discrimination while Roma are offered free university places. For this person, the news report where the Roma man jokes about expecting a room in a good hotel is related to a wider problem of the Roma playing the ‘race card’ and is more evidence of entitlement.

Here and throughout our corpus, humour is deployed as a discursive strategy. Comment (7) begins with irony and sarcasm as it goes into blatant racism—with the underlying message that it is clearly insane to support the Roma. Scholars have pointed to the important role that can be played by humour and joking in hiding racism behind what is presented as fun and being playful (Billig 2005).

Additionally, those creating the humour position themselves in a superior way as they display their smartness and mastery over discourse (Breazu & Machin 2019), by relying on play-on-words, irony, and sarcasm, along with plain outrageous incivility. The most brutal comments are rewarded with replies of hysterical laughter emojis. The use of the term ‘eliminated’ in comment (7), referencing the idea that the Roma not following lockdown may die through Covid-19, is a form of esoteric humour, meant to be understood by those with inside knowledge. This sharing of humour and knowingness is, therefore, an important part of how the affordances of social media can create a sense of engagement around an affective ‘we’—here the ‘silent majority’—who share what is presented as common-sense knowledge.

What is salient in this reference to playing the race card is that it is presented in a way that gives agency to the Roma, as seen in comment (8).

1. (8) They still shit at the back of their yards and demand better conditions.

Here the Roma ‘demand better conditions’. In other comments, we have seen that they ‘occupy the free places at universities’. These are not represented in terms of them being ‘granted places’ nor the Roma being ‘supported, or helped, with free places’. Foregrounded is the active verb ‘occupying’, which connotes something active. The Roma are not passive nor vulnerable but are highly active making forceful, unreasonable demands. This sense of ethnic minorities taking over has been associated with white victimhood with the implication that the majority, the ethnic nationalists, are losing power. This has been observed as a central part of the idea that they are now neglected by distant and corrupt elites and need to turn to forms of leadership that will return power to them (Eatwell & Goodwin 2018).

In comment (8), repeating across our sample, there are references to lack of hygiene—a well-documented discourse used to represent the Roma (McGarry & Drake 2013): in (8) ‘They shit at the back of their yards’, in (4) above ‘these ones with grimy faces’, and in (7) ‘this stinking ethnicity’. These references to hygiene relate to the unruliness as part of the discourse of the burden that the Roma create, but also relates to a longer history of representations of Roma as being subhuman and inferior (Tileagă 2015). Here we see how new racism, with discourses relating to cultural, economic, and social burdens, with its corresponding sense of victimhood, so easily converges with the blatant hatred of an ethnic group as dirty, unruly, inferior, and despicable. Worse perhaps is that new racism can add an extra layer of logic to such biologist racism—the economic and social burden and the unruliness provide evidence for the inferiority of the Roma.

The corrupt elite who fail to deal with the Roma

Running through the comments, another key element of this nativist discourse (Bergmann 2020) is the idea that the establishment, politicians, and major institutions are either openly supporting this failure in the meritocratic system or are doing so since they are out of touch, through incompetence, or particularly in the Romanian context, due to corruption—the latter relating to many high-profile incidents in Romania (Harris 2018). We see this above in comment (7), which states ‘we have a country led by corrupt and incompetent people, who favour an ethnicity from which very few people rise’. Here the authorities are said to favour the Roma with university places. Important here is the failure to control or punish the Roma as we see in comment (9).

1. (9) The problem is that they [the authorities] do not fine or to put them in prison for 20–30 days so that they don't walk around. They let them do whatever they want. For them [the Roma] there are no rules… People of 2 bucks.

In terms of verb processes, we find that the authorities ‘do not fine or to put them in prison’ and they ‘let them do whatever they want’. This repeats across the comments. Here the authorities, which have the power to do something, do nothing. The interests of the ‘we’ are simply not addressed.

What is foregrounded, therefore, are not structural matters of cycles of poverty and the need for the government to address them, but the failure to discipline the Roma. Such comments imply that something should be done and someone should do something. Part of this discourse of nativist victimhood is that a simple solution is required, often to be supplied by a leader who can claim to align with the interests and wisdom of the people (Bergmann 2020).

Comment (10) includes a further shortcoming of the authorities.

1. (10) Corona established order in Ţăndărei! Otherwise they would not understand, headless people! They do not know what order or cleanliness are. This is how they have lived so far under the leadership of the PSD members!! The social democrats filled their pockets with bribes and let them [the Roma] do whatever they wanted!! Now the bribe stopped!!! And Ţăndărei must obey now and then!! Otherwise they'll go to hell!

The politicians are represented as having ‘filled their pockets with bribes’. The widespread nature of government corruption in Romania has been well documented (Harris 2018). But here this corruption is linked only to the Roma. And Covid-19, therefore, becomes a bringer of order. Now the Roma cannot do what they want to do, or they will become ill and die. Important here is the discourse not that some Roma, due to cycles of marginalisation and poverty, may experience certain life situations, but rather that they simply ‘do not know what order or cleanliness are’.

The Roma as an embarrassment

Across these comments the victimhood of the non-Roma is overtly clear. As we have seen in earlier comments, they work hard and pay taxes. They carry the economic and cultural burden. They are the victims where meritocracy breaks down as the Roma play the race card. And they are the victims of the incompetent and corrupt government and system which allow this to take place, which fails, or refuses, to control the Roma. There are, however, other aspects to this victimization discourse.

1. (11) These lazy ones embarrassed us all over the world. It's an embarrassment to say you are from Romania. Wherever they go in Germany, they make a mess and destroy everything they rent as if an earthquake passed by. There is a city in Germany, Duisburg, where all their tribes gathered with 7, 8, 10 children per family and they destroyed the entire building. They were finally kicked out as the block was going to collapse.

In comment (11), there is reference to the embarrassment caused to Romania as a country and to Romanians. This discourse, which has been present in media coverage of Roma migrants in EU countries, suggests that it is a problem that the two groups become conflated by the populations of other countries (Mădroane 2012; Breazu & Eriksson 2020). This means that Romanians, too, experience discrimination and assumptions that they are untrustworthy, criminal, and so on.

In addition, we find the Roma again represented through active verb processes as agents of destruction, chaos, and crime: ‘they make a mess and destroy everything they rent as if an earthquake passed by’, ‘they destroyed the entire building’; these verbs suggest agency rather than passivity. This destructive culture then becomes associated with all Romanians, for which non-Roma must also suffer.

Romanian people are not allowed to speak out

Another aspect of this victimhood is that it is not possible to complain as this will result in accusations of racism, as we see in comment (12).

1. (12) The Roma representatives were more concerned to give lectures to Romanians instead of educating their own ethnics, while they [NGO representatives] wasted millions of euros from European money. If you mention any of the problems of the ethnic group, there will always be an obscene paid NGO person to bark at you that you are racist.

Here the Romanians are said to be denied any voice, unlike the favoured minorities. This has also been an observed feature of this nativism where political correctness exists to prevent any kind of open criticism of this unjust situation (Eatwell & Goodwin 2018). In this discourse of victimhood, it is the non-Roma who are to adapt, not the Roma, who, we have seen, have the freedom and power to make demands and act in an unrestrained manner. Such silencing of open criticism relates to a broader sense that the elite political system no longer represents the interests of ordinary people and also suppresses truth and free speech. These comments position themselves as the silent majority speaking out.

Nativism, biological racism, and eugenics

We have seen that comments on the feed carry reference to ‘we’, ‘us’, and ‘our’—those who work hard, suffer ruined reputations, and are failed by corrupt and incompetent authorities. This represents the affective community who engage each other in conversation and joking. But in rare instances, the biological or ethnic nature of the ‘we’ is overtly stated.

1. (13) According to the history we founded Rome, you found your fucking hell

2. (14) Hey Indian, unlike you, we are a Latin country!

In comments (13) and (14), we find references to distinctive origins of the Roma and Romanians. Links are made to ancient classical European civilizations compared to the Roma's connections to India. Here we find racism expressed not in relation to culture or the economy, but to biology. In such cases, the ideology behind the managed new racism becomes explicit: we are a different, superior race, with classical credentials. We also find references to genetic differences, raised in regard to how Covid-19 will infect the Roma.

1. (15) Put guards at the entrance and exit of the city and set them [Roma] free. They'll infect each other will die and we'll get rid of them. They're of no good anyways.

2. (16) Haha you think so! They were raised naked, in the cold, rain and begging, and pneumonia is their friend. It is us who will die like chickens and they will survive like cockroaches even in the next pandemic.

Comment (15) relies on humour and sarcasm to create an affective flow, suggesting that the Roma should be simply allowed to infect each other and die. But comment (16) contradicts this to say that the Roma have developed immunity due to their way of life. The difference in viewpoint between the two posts was not debated on the feed. Both show blatant racist hatred towards the Roma: ‘they'll die and we'll get rid of them’ and they ‘survive like cockroaches’. Typical of social media feeds, there is no debate nor necessary coherence, only a loosely defined sense of affect mobilising those posting.

Conclusion

In the title of this article, we signpost the issue that racism is more than hate speech. This was motivated in particular since, as a result of previous publications about racism towards the Roma, we have been approached by various organisations, including Facebook, who were interested in addressing overt racist hate speech, taking what Siapera & Viejo-Otero (2021) call a ‘data approach’ to racism. But research into racism and social media points to the need to reveal and understand both blatant and more subtle forms (Ben-David & Matamoros-Fernández 2016). And ‘data’ racism risks treating racism as a monolithic ideology, failing to understand it in historical material terms (Camfield 2016). Such an approach can miss how it intersects with other social and cultural practices that shape and maintain it (Doane 2017).

This news clip, and the comments posted in response to it, must be placed in a specific Romanian historical and material context. There is the longer history of racism and xenophobia towards the Roma, as the Roma have remained in a socially marginalised situation, largely excluded from civic society, and experiencing high levels of poverty. EU reports have shown that the Roma live in the poorest areas, often in places with no electricity, gas, running water, or paved roads, which in turn, limit or deny their access to education, job markets, and proper health care (FRA 2020). These socioeconomic inequalities and political exclusion of Roma are largely absent from contemporary media and political discourse in Romania. This situation is also represented as one chosen by the Roma who are reluctant to progress and resistant to social inclusion.

We must also place the discourses we find in these comments into the specific context of the post-communist economic crisis in Romania, the growth of ethnonationalism, along with disillusionment with institutional corruption. And, as in many other countries in Europe, there has been a growth in fears about change, economic instability, growing inequalities, and the decline of former ways of life (Eatwell & Goodwin 2018). Such fears can be fostered by political groups, using new racism and its associated new nativism. In our corpus we find evidence of these fears and frustrations. But we see the embeddedness of the longer-term history of racism towards the Roma.

As crises make xenophobia and racism more extreme, or at least more blatant, this gives us a chance to examine what kinds of discourses come to the surface. As Gilroy (2012) argues, even when people are upset, embittered, or frustrated, because of their life situations, it is of importance to ask which ideas they use to account for this. Such ideas tend to be those that have become, even if implicitly, more widely available and shared, and which are embedded in the longer history of racism and xenophobia towards the Roma in Romania, as well as in more recent upheavals and shifts.

While social media allow for contributors to vocalise their opinions, the dominant views found in the comments in our corpus are clearly ideological extensions of what we see in the mainstream media. Commentators take up the media's invitation to be hateful, and social media create discursive spaces for commentators to experience a sense of being part of a community of knowers who see these problems for what they are: the burden of the Roma, the blindness and incompetence of political institutions, the reverse victimhood. It is also striking how in social media we see discursive elements not found explicitly in the mainstream media where comments take the dehumanisation of the Roma to another level, proposing cruel solutions to ‘solve’ the Roma problem, realised through cruel and terrifying humour and sarcasm.