Reading Twitter as a marketplace of ideas: how attitudes to COVID-19 are affecting attitudes to migrants in Ireland

1.1 Ireland as the test case

The context of our study is the Republic of Ireland in the twenty-first century. Formerly a net-emigration state, Ireland is now a net-immigration state, and it is important to consider how the relationship between migrants and their host communities has exacerbated socioeconomic tensions.

Linguistically, the modern experience of the migrant to the Republic of Ireland has been described in quite a positive light. We have seen Ireland as becoming increasingly diverse and multilingual (Carson et al. 2015), and studies on attitudes to multilingualism are overwhelmingly positive (Modrescu and Carson 2017: 47). This is in stark contrast to the negative attitudes to migration during the earlier economic downturn, from approximately 2009 until 2014 (Denny and Gráda 2013). Recent immigration to Ireland can be divided into two overlapping phases: that relating to the “Celtic Tiger” (1997–2008); and that which immediately preceded and then followed the accession of new member states to the European Union (EU) in 2004 and the Republic of Ireland’s response to refugees and asylum seekers (1999–present). Historically, the pattern of migration on the island of Ireland was one of substantial emigration. Beginning in the 1990s however, this pattern shifted to one of net immigration (Conlon 2009; Gilmartin 2015; Woods 2018). This shift from emigration to immigration was brought about by the Celtic Tiger economic boom experienced by the Republic of Ireland during the 1990s and early 2000s, during which time an “apparently infinite demand for and supply of labour” led to a sudden expansion of jobs (Krings et al. 2013: 38). As a result of the excessive demand for workers, which outweighed the supply capabilities of the endogenous labour market, immigration was required, with both skilled and unskilled migrant workers coming to the Republic of Ireland (Fanning 2016; Woods 2018). This influx of immigration was particularly prominent following the accession of new member states, such as Poland, to the EU in 2004 (Barrett and Duffy 2008; McGinnity and Gijsberts 2018; Voitchovsky 2014). As the growth of the Irish economy necessitated increased immigration in order to supplement the limited native labour force, many unskilled migrant workers filled the shortages in lower-paid roles that had been vacated by Irish workers, particularly in rural areas (Fanning 2016; Gilmartin 2015).

Despite being an inherently diverse and varied grouping of people, refugees and asylum seekers are most often reduced to a homogenous, threatening and undesirable collective (Devlin and Grant 2017). Central to this imagining are processes of categorization and hierarchization of migrants and migrant bodies, as was discussed earlier. Within the Irish context, Gilmartin (2015: 28) has argued that Irish political and public discourse has divided migrants into “those who are encouraged, those who are tolerated, those who are expedient and those who are discouraged” (2015: 28). At the top of this hierarchy of acceptability and desirability are returning Irish migrant and business migrants, while EU migrants and unskilled migrants are accepted for economic expediency. However, asylum seekers and refugees have been “vilified in the Irish media and discouraged through increasingly stringent restrictions on their rights and mobility” (Woods 2018: 168; see also Burroughs 2015; Conlon 2010). Research on the rise of the right in Irish politics (e.g. Cannon et al. 2022; Phelan and Kerrigan 2024), highlights the critical discourses of the ways in which far right parties spread their message.

1.2 Critical discourse analysis and metaphor

The framework through which we will view these data is critical discourse analysis (CDA), in particular following the lead of, inter alia, Lakoff (1996, 2002, Wodak and van Dijk (2000), and Hart (2010). Here, we can view language as social practice (Fairclough 2003, 2011), inviting the analysis that linguistic acts can display deeper attitudes about other social practices. Van Dijk (2015) provides a socio-cognitive analysis of UKIP discourse in the run-up to the 2015 UK general election, in particular that of the party’s leader at the time, Nigel Farage. Social cognition, the shared beliefs of a group of followers of any political ideology, allows implicit connections between unrelated topics and activates, in this particular case, xenophobic attitudes, which leads to the normalization of racist ideology. Taken a step further, racist ideology also manifest in other attitudes (e.g. attitudes to abortion, LGBTQIA+ issues, or religion). Recently, the media has been shown to push anti-immigrant discourse in Europe (Krzyżanowski et al. 2018), furthering the spread of far right ideology.

Wodak and KhosraviNik (2013) argue that despite certain nuances in form and practices, right-wing populism is not a “new” phenomenon, but can be traced back to the Second World War. They go on to claim that the nuance of “new right-wing populism” is the recent surge in support for far right movements and the means by which some of the disparate right-wing populist ideologies have established themselves within many of Europe’s liberal democratic societies as mainstream political narratives. Right-wing populism splits society into “us” and “them”, along religious, ethnic, regional, and national lines (Wodak and KhosraviNik 2013: xx). Although the EU claims to respect human rights in its migration policy, the deaths of migrants have been blamed on “criminal gangs” of smugglers and traffickers, rather than the EU’s harsh securitizing of its external border, its implementation of Frontex, and the ceasing of Mare Nostrum, the search and rescue operation formerly carried out by the Italian navy (Hintjens 2019). Data from the European Social Survey (ESS) seem to indicate generally positive attitudes to migration, despite the rise of right-wing populism, but that extant European populations want migrants who are able to assimilate socially (Dennison and Dražanová 2018).

Most importantly for the current study is how CDA and metaphor interact. Lakoff and Johnson (2003: 3) explain metaphor as “pervasive in everyday life”. Metaphors are everywhere, which allows the “understanding and experiencing one kind of thing in terms of another” (2003: 5). Hart (2010: 113–124) suggests that Conceptual Blending Theory offers a more appropriate interpretation of the linkages between figurative and physical images of migrants. El Refaie (2001) describes the different target domains of a thematic analysis of Austrian newspaper discourse. Here, themes focus on immigrants as, inter alia, criminals (e.g. illegal immigrants) and as a weight or burden (e.g. carrying migrants along). Elsewhere, Berry (2019) describes the financial crises of the early twenty-first century as the motivators for describing immigrants as elites who are treated more favourably than British people. More recently, Olza et al. (2021) and Semino (2021) have shown Twitter to be a rich source of metaphor data on COVID-19. These papers discuss the war metaphors (Olza et al. 2021) and fire metaphors (Semino 2021) that have been used to conceptualize the fight against the deadly virus. Twitter can be a powerful platform for the expression of ideas and for far right political messaging (Froio and Ganesh 2019).

As a catalytic event, the COVID-19 pandemic offers a unique opportunity to show us humanity at its worst and at its best. In this study, we aim to show the types of discourse on Twitter as the first eight months of the COVID-19 pandemic evolved. We then analyse the data from the perspective of metaphor, and show how this catalytic event is inviting far right political ideology and normalizing anti-immigrant discourse.

2.1 Pilot

Our justification for using Twitter as a platform for far right political ideology is rooted in examining ESS data on the correlation of placement on the left–right political scale and likelihood to post political messages on social media (ESS 2020). Using the existing data from the Republic of Ireland component of round 9 of the ESS (collected in 2018–2019), we performed a Pearson correlation test in SPSS to show this correlation was significant (Figure 1 and Table 1). This pilot shows that political social media in Ireland skews right, which seems to be consistent with other political alignment studies of Twitter (see, e.g. Sosnkowski et al. 2021).

Figure 1:

Scatter plot graph of sharing political posts on social media and placement on left–right political scale. Source: ESS (2020).

2.2 Twitter data

We chose 23 January 2020 as the starting point for data collection as this was the day when news outlets reported that a patient in a Belfast hospital was being tested for the virus (McClements 2020), though there was no confirmed case on the island of Ireland until 29 February 2020. The end point of the Twitter scrape was 29 August 2020, the day before the scrape took place. The first search terms we agreed were “#covid19Ireland(*)”, including permutations such as “#covid_19_Ireland”, and so on. We then looked at “Ireland (#covid19)” before settling on “#covid19Ireland” as the most productive hashtag for information related to COVID-19. This selection informed the following searches (without hashtag) of “migrant + #covid19Ireland” and “immigrant OR asylum OR refugee + #covid19Ireland”. Finally, we also searched the following terms without hashtags and other terms in order to collect as many descriptive terms for migrants:

1. Ireland covid (migrant OR migration OR asylum OR immigrant OR immigration OR refugee OR foreigner)

2. Ireland corona (migrant OR migration OR asylum OR immigrant OR immigration OR refugee OR foreigner)

3. Ireland covid19 (migrant OR migration OR asylum OR immigrant OR immigration OR refugee OR foreigner)

Each tweet was read and initially coded for whether or not the tweet expressed any kind of far right discourse or themes (see Wodak and KhosraviNik 2013). Following this process, an inductive coding process (see Thomas 2006) revealed the themes that emerged in the far right tweets. In the sections that follow, we will report the findings along with analysis from the perspective of metaphor for the relationship between COVID-19 and immigrants.

3.1 All tweets

In Table 2, we show the total number of analysed tweets and this initial coding. The total number of tweets increased from January to April as the pandemic took hold in the Republic of Ireland and harsh lockdown measures were introduced on 12 March 2020 (e.g. travel was restricted initially to 2 km radius from home). These measures began to be lifted on 18 May 2020, with continued easing into August. This pattern seems to repeat itself in the total number of tweets with the associated hashtags. The shape of this curve is also seen in the number of non-far right tweets which peak in March and April before receding until an uptick is seen in August when a surge in cases linked to meat factories increased interest in the co-occurrence of COVID-19, Ireland, and one of the migrant-linked terms mentioned above.

As a proportion of the total number of tweets, the far right tweets represent 21.6 % of the total, suggesting that the search terms chosen are valid for this investigation. For our purposes, the 210 far right tweets are pertinent, and so we proceeded to the second level of analysis.

3.2 Far right tweets

The 210 far right tweets were carefully read and coded for the thematic elements within them. Where a tweet mentioned more than one theme, it was coded for both themes. Three themes emerged in the data; these are shown in Table 3, together with the number of far right tweets that relate to each theme. We will now look at each of these themes in turn.

3.2.2 Migrants as criminals

The association between migrants and crime is well established through the terminology associated with “illegal immigration” (see El Refaie 2001; Charteris-Black 2006; Hart 2011), and can be seen in (4) and (5):

1. [An amnesty policy will] incentivise further illegal immigration (April)

2. flooding the country with illegal Migrants (August)

In (4), there is a clear example where the tweet indicates that a move by the government to stop deportations will increase the number of migrants coming to the country illegally. There are two metaphors at play in (5), which contains a water metaphor of migrants flooding the country while identifying migrants as criminals as well. What is perhaps startling about this example is that the poster is intending to defend a woman who was the focus of negative media attention at the time, before they launch into anti-migrant discourse. Framing migrants as criminals is a theme in the data from the beginning (see Figure 2).

Figure 2:

Far right themes by month.