Voice and community in the 2015 refugee crisis: A content analysis of news coverage in eight European countries

Introduction: Humanitarian securitization and the refugee crisis

The refugee ‘crisis’ ¹ turned into major policy issue and news story in May 2015, following two deadly shipwrecks off the coast of Italian island Lampedusa. Faced with the arrival of a million migrants from Middle Eastern and African countries, the greatest influx since the Second World War, ² the European Union attempted to establish a unified framework of refugee reception. This framework would both keep the continent safe and protect the lives of those who fled persecution and war to reach its shores; in its own words, the EU engaged in a ‘search for the balance of humanitarian needs with concerns over sovereignty’ ³ or, what critical migration and security studies refer to as, a framework of ‘humanitarian securitization’ (Vaughan-Williams, 2015).

Defined as a ‘practice of power’ through which ‘a political community’ is invited to ‘treat something as an existential threat’ (Buzan and Wæver, 2003: 491), securitization can be seen as ‘humanitarian’, when, as in the case of the refugee crisis, this threat is simultaneously construed as a referent of vulnerable humanity. As a practice of power, then, humanitarian securitization operates through a contradictory discourse that is supposed to humanize and care for those it also regards as its ‘others’. Territorially, this contradiction was resolved in practice when, in spring 2016, several nation-states situated on the refugee route towards Central and Northern Europe closed their borders and stopped the flow – leaving approximately 57,800 people stranded in Greece. Symbolically, the contradiction was manifested as a constant struggle of shifts and turns over the narration of refugees in news journalism.

Given the relentless rate and tragic circumstances of refugees’ arrival, news across Europe routinely worked to accommodate narratives of security (why and how ‘our’ borders are kept safe) with narratives of care (portraying ‘their’ suffering as a cause of ‘our’ emotion). How these stories were told and who spoke for or about the refugees was thus instrumental in highlighting the causes and circumstances of their arrival as well as the implications of their reception. These news narratives are considered to be, as the European Commission (2011) itself claims, a key source of knowledge through which the continent’s collective perceptions of refugees emerge. The importance of news journalism on the crisis lies, therefore, not simply in its informational value but also in its political value as an instrument for the formation of European publics as moral communities (Chouliaraki, 2013). It is this political value of journalism that our focus falls upon in this essay, as we address key questions of narrative, voice and community in the news. ⁴

Relying on the discourse of humanitarian securitization, our analysis aims at establishing how European news on the crisis manages the inherent tension of this discourse between care and protection. Reflected in the linguistic ambivalence of the term ‘refugee’, swiftly shifting between the positions of ‘victim’ and ‘threat’, this tension raises the question of voice, that is of who has the right to narrate the crisis and from whose perspective. How were refugees narrated in the news? Who spoke about their plight and in which authority? How was their voice contextualized? And what were the implications of these news narratives for the communities we belong to and for the ways we imagine the world beyond ourselves?

By addressing these questions, our essay aspires to shed light on the fundamental moral challenge with which the refugee crisis confronted ‘us’, Western communities of relative safety: our encounter with non-Western ‘others’ (Silverstone, 2006). Our research questions on narrative, voice and community matter then precisely because they enable us to reflect on journalism as a space where this encounter is performed linguistically before it is enacted corporeally. Rather than using a restrictive conception of the border as a territorial line of separation that regulates entry into ‘our’ nations, we adopt a broader conception of the border as a symbolic practice that includes and excludes; in the words of Vaughan-Williams, the border ‘seeks to rhetorically identify and control the (very) mobility of certain people, services and goods that operate around its jurisdiction’ (2015: 6). In its function to ‘rhetorically identify and control’ mobility, bordering entails the symbolic power to name and authorize who is accepted as one of ‘us’ or shares ‘our’ humanity and who does not. European journalism can, in this light, be seen as a practice of ‘symbolic bordering’: a linguistic practice that works in tandem with Europe’s territorial borders, insofar as it regulates the symbolic construction of refugees as human beings – voicing or silencing them, humanizing or vilifying them (Chouliaraki, 2017).

In so doing, journalism-as-bordering also renegotiates the boundaries of ‘our’ own communities of belonging. Depending on their language choices, some stories may enable us to recognize those different from ‘us’ as worthy to be listened to and thus encourage ‘us’ to stretch the imaginary boundaries of ‘our’ community in order to include ‘them’, but other stories may block this potential. Drawing on Hannerz’ definition of the cosmopolitan disposition as ‘an orientation, a willingness to engage with the Other’ (1990: 239), our study of press journalism in Europe aims precisely at identifying the different ‘orientations to the other’ that its news stories articulate. It asks whether European press stories construct ‘cosmopolitan’ communities that are open, hospitable and inclusive of refugees or ‘communitarian’ communities that are closed, phobic and introverted (for the vocabulary: Chouliaraki, 2006).

Drawing on the findings of London School of Economics and Political Science’s (LSE) ‘Migration and the Media’ research project (see Appendix), we address these questions by engaging in a content analysis of broadsheet press in eight European countries, in the peak of the crisis, June–December 2015. We begin with a theoretical overview of existing studies on refugee representations. While we draw upon this literature to build our analytical vocabulary, we also expand on it by proposing a conceptual understanding of journalism as ‘symbolic bordering’, that is as a set of narrative strategies, which perform ethico-political effects of misrecognition and ex/inclusion. Within this framework, our analysis of news stories establishes three different strategies of bordering across our empirical material: bordering by silencing, by collectivization and by de-contextualization. National and temporal variation in our data granted, we conclude that the humanitarian security discourse of the news has serious implications both upon refugees and upon Europe’s national publics. On the one hand, news stories systematically misrecognize refugees as political, social and historical subjects; on the other, in so doing, it simultaneously calls up largely ‘communitarian’ publics: publics willing to consider the humanity of ‘others’ only in order to affirm ‘our’ benevolence but not in order to consider including ‘them’ into ‘our’ communities of belonging.

Literature review

We begin with a discussion of two relevant bodies of research. First, we review literature on the mediation of human mobility, which provides us with the conceptual vocabulary to analyse the dominant representational tropes of migrants and refugees in the news. Second, we introduce a theoretical understanding of journalism as performative practice that uses selective distributions of voice so as to create hierarchies of who speaks and how in the news. Combined, these two bodies of scholarship enable us to produce the framework of symbolic bordering within which to make sense of our data.

Voice as narrative: A content analysis of European news

Our approach employs a content-analytical framework to the study of news performativity across eight European national news cultures. Our data consist of newspapers in eight countries: five of the countries positioned across the route of the refugee journey towards Central and Northern Europe (Greece, Serbia, Hungary, Czech Republic, Germany), but we further included three other major countries, which were themselves some of the preferred destinations (France, UK and Ireland). This empirical choice of countries enabled us to compare local, regional and continental trends and uncover a broad and varied range of narratives surrounding the crisis. Our sample included two influential broadsheet papers per country, each representing the left/right political spectrum. We focused on three crucial moments of the crisis, each of which played a catalytic role in shifting public debate – though not always in also influencing policy-making in EU. ⁶ Each moment further delineated a distinct phase in the narration of the refugee news, what Georgiou and Zaborowski refer to, respectively, as ‘careful tolerance’, ‘ecstatic humanitarianism’ and ‘securitisation’ (2017). For each of these moments, we sampled 20 refugee-related articles per newspaper, which resulted in a 1,200-article data bank overall (see Appendix for details).

While national, political and temporal variations were taken into account, a detailed discussion of differences remains outside the remit of our study. Our focus is instead on establishing shared patterns across data. Given our interest in journalism as a performative practice of bordering, let us reiterate that our analytical categories seek to classify how, in the framework of humanitarian securitization, news stories tactically use voice to provisionally settle the unstable signifiers of victimhood and threat in specific news narratives, and how, in so doing, they make different claims to the humanity of refugees. Our analysis of ‘voice as narrative’ focuses, therefore, on the subjects of voice (absence or presence of refugee voice); the status of voice (how refugees are named and characterized as) and the context of voice (the justifications and consequences through which we are to understand the predicament of refugees). Depending on linguistic choice, as we shall see, each of these categories plays out a different politics of ‘voice as value’, that is each identifies a different strategy of humanization in the news and, in so doing, conjures up particular relationships of recognition and hence of inclusion and exclusion between ‘us’ and ‘them’. We examine each linguistic category in the sections on ‘subjects’, ‘status’ and ‘contexts’ of voice, before we conclude with reflections on the imaginations of community these categories invite us to produce, in the section ‘voice as value’.

The subjects of voice

The question of who speaks in the European press takes a straightforward response: it is politicians, not refugees. In our 1,200-article data bank, quotes from migrants or refugees were present in only 16.6% of the sample, as compared to 66% of politician quotes (national or EU). Differences in newspaper format and article length granted, this trend was consistent across all countries of our sample: refugee voices were a minority compared to politicians (Figure 1). There was significant difference between right- and left-wing press, with the percentage of refugee quotes being higher in the latter (20.4%) rather than the former (14.1%); and there was also a difference between Western and Eastern European press, where the former (39.2%) had a considerably higher percentage of national politician quotes than the latter (27.2%). While such variations illustrate key differences in conceptions of democracy and civic participation across the political spectrum and the historical trajectories of the continent, our focus here falls on the persistent gap between ‘our’ and ‘their’ voice, across these divisions. OPEN IN VIEWER

The temporal perspective complicates without drastically modifying this gap. Following Alan Kurdi’s tragic death, in September 2015, there was a 4% increase in refugee quotes throughout the ‘humanitarian’ phase of reporting: from 17% to 21%. This was accompanied by a concomitant shift in the emotional trajectory of the news, where ‘solidarity’ rose in the ‘humanitarian’ moment from 12.5% earlier to 14.7% and ‘understanding’ from 7.5% to 10% (Figure 2). Still, however, politicians continued to be by far the most vocal presence in the news: 41% of articles had quotes from national politicians, 34% from politicians coming from other nations and 30% from EU officials. Following the Paris attacks, in November 2015, we witness a sharp drop of 9% in refugee quotes, which fell to 12%. This suggests that, in the ‘securitization’ moment, we have a nearly total silencing of the refugee, who had, by now, shifted position from victim to terrorist (Figure 1). This muting was, predictably perhaps, also accompanied by a sharp increase in negative emotion: ‘frustration/concern for possible security risks’ rose from 10.7% to 18%, ‘fear’ rose from 4% to 10% and ‘solidarity’ percentages halved from 14.7% to 7.7% (Figure 2). OPEN IN VIEWER

Even though such shifts illustrate the symbolic instability of the signifier refugee, they also reveal the permanent chasm between Europe’s elite voices and refugees. Independently of whether the latter are victims or assumed perpetrators, they remain consistently marginalized or almost fully silenced across European news. As victims, refugees are humanitarian subjects defined, as we saw, exclusively by their human needs (Malkki, 1996), whereas as a threat, they are enemies of national sovereignty and hence as undeserving of voice (Chouliaraki, 2006). Either way, the omission of their voice from news stories throughout the peak of the crisis shows refugees to be endowed with an ambivalent humanity as both vulnerable and lethal. This is what Malkki calls a ‘mere, bare, naked, or minimal humanity’, which traps refugees firmly within the framework of humanitarian security and misrecognizes their status as political and historical actors (1996: 390). Whether victims or terrorists, refugees are never human.

Beyond this consistent rupture between ‘us’ and ‘them’, however, there is a further division within ‘us’. For national civil societies across Europe had a rather low audibility in the news too (Figure 1): volunteers and activists were quoted in 7.9% and 6.7% of articles, respectively, whereas NGOs in 13.3%. Ordinary citizens were similarly only sporadically quoted: 5.9% of all articles included the voice of national citizens, 5.3% the citizens of other European nations and 1.1% the voice of minority group citizens. And yet, despite this marginalization of civic voice, news stories did make frequent appeals to ‘ordinary people’ and ‘their emotions’ – with ‘frustration’ and ‘uncertainty’ being the highest scoring ones across the ‘tolerant’ and ‘securitization’ moments of the crisis: ‘frustration’ at 18% and ‘uncertainty’ at 12.6% on average, with the exception of the ‘humanitarian’ moment, when both emotions scored lower (Figure 2). This marginalization of citizen voice, then, did not only minimize the potential of popular deliberation within the spaces of journalistic publicity. It also allowed the authoritative voices of politicians (in their direct quotes) and journalists (in their general references to public affect) to dominate the argumentative and emotional spaces within which ‘we’ were to make sense of the crisis.

Through the double silencing of refugees and citizens, the European news did not only misrecognize ‘them’. It also misrecognized ‘us’ as political subjects with a right to voice and with a plurality of views that would have, at least partly, challenged the status-quo – for instance, by allowing NGOs and civil society voices to speak for the cause of refugees. As a consequence, the views of the political establishment assumed a naturalness that, in most, though by no means all, cases rendered the refugee crisis a matter of economic calculation and technocratic management, marginalizing the experiences, emotions and testimonies not only of refugees themselves but also of those of ‘us’ who might have been able to communicate their plight – but see Georgiou and Zaborowski (2017) for key differences in selected countries.

The subjects of voice, to conclude, were regulated through the news’ strategies of silencing, which marginalized the voices of refugees as well as those of European citizens and privileged instead the voices of politicians. While there was fluctuation in the degrees of silencing refugees, with certain moments granting more voice than others, ultimately, their systematic silencing created a hierarchy of voice that produced effects of misrecognition. This hierarchy prioritized the voices of the political establishment and, thus, largely legitimized affective claims to frustration and uncertainty, while devaluing alternative narratives of solidarity by or for refugees.

This hierarchy of voice misrecognized refugees as political actors in that it narrated them primarily as abject bodies unable to participate in public discourse with their own stories, explanations and emotions. If people assume a political status, as Arendt puts it, when they appear in public ‘in the unique shape of the body and the sound of the voice’ (1958: 179), then those who remain voiceless are by definition exiled from the political sphere. As a consequence of this effacement of refugees from journalistic publicity, the news tended to call up a ‘communitarian’ community of strictly national publics that could not challenge the limits of humanitarian security nor acknowledge refugees as right-bearing actors with a complex story to tell.

The status of voice

If the subjects of voice refer to the presence or absence of those who speak, status refers to the authority attributed to those who speak. Who are the refugees? And what is it that they bring to ‘our’ societies? Even though ‘crisis’ journalism was intense, at the time, there was little attention to these questions. Refugees were predominantly described as groups in terms of their category membership: predominantly nationality (62% of all articles) and less so gender (24%) and age (27%). More personalized specifications, such as names and professions had the lowest frequency with 16% and 7%, respectively (Figure 3). OPEN IN VIEWER

This general tendency to ‘collectivization’ remained relatively constant throughout the three moments of our study, from the early ‘careful tolerance’ through to ‘ecstatic humanitarianism’ and the ‘securitization’ ones. In the latter, for instance, speculation that one of the Paris killers had entered Europe through the Greek passage, grouped all refugees in the category of the terrorist, suppressing the option of more diverse and personalizing narratives – for instance, though family- or profession-related information (though this was the case with locally born terrorists). There was, however, a slight increase in the inclusion of specificities during the ‘humanitarian’ peak: 25.5% of articles included names of the refugees and 32% included their age. Given that this shift towards personalization correlated with the timing of Alan Kurdi death it is, nonetheless, possible that it did not reflect a generalized trend for all refugee descriptions but was linked to repeated descriptions of this particular victim and his family, as their names recurrently appeared in an abundance of stories. This hypothesis is further supported by the lack of correlation between the increase in name/age references and references in profession – the latter remained unchanged across all three moments of the study, at 7.7%, 7% and 6.5%, respectively.

The personalization of the boy's tragic death was further combined with an increase in emotions of solidarity (from 17.5% to 25.2%) and empathy (from 17.7% to 22%) for refugees between July and September (Figure 4). While this emotional peak contributed to endowing Alan with iconic status as a universal symbol of the suffering of the innocent and perhaps a sign of ‘our’ own failure to care (Ibrahim and Howarth, 2016), it paradoxically did so at the expense of misrecognizing him. On the one hand, Alan’s death as an icon of victimhood came at the cost of muteness – the absolute silence of the dead; moreover, when his father sought to narrate the family’s tragic fate, him voice was quickly undermined as inauthentic in news debates that framed his as a ‘human smuggler’. ⁷ On the other hand, the potential of Alan’s loss to become itself a ‘voice’ of martyrdom, denouncing the ‘let die’ governance at work in Europe’s sea borders (Basaran, 2015), was subordinated to dominant interpretations of his death voiced by elite actors, national politicians – their voices, at the ‘humanitarian’ moment, held indeed the highest audibility at 40.5%. Instead thus of turning the iconic event of Alan’s death into an occasion for, what Christensen and Jansson (2015) call, ‘cosmopolitan reflexivity’ over ‘our’ border politics, the coupling of personalization and emotionalization displaced critique onto sentimentalism and turned the event into a case of ‘sublimation’ (Gay, 1992): the diffusion of the moral and legal scandal involved in the boy’s death in a plethora of emotional calls for solidarity. What was lost in the process was the confrontation with key failures of ‘our’ security priorities, which, in Basaran’s words, ‘dilutes the legal duty to rescue’ and classifies people in terms of ‘worthy lives that fall within the duty to rescue’ (2015: 1) and others that do not. OPEN IN VIEWER

The status of voice in European news narratives was thus largely regulated through strategies of collectivization, whereby groups remained unknowable and nameless whereas individuals were named but dead. Misrecognition in the status of voice, then, both sustained the news hierarchy established through the subjects of voice, and further effaced refugees by grouping them into general categories or by dispersing the unspeakable tragedy of ‘their’ death onto ‘our’ voices and emotions. If then, in the subjects of voice, silencing resulted in the misrecognition of refugees as political subjects deprived of the right to speak as equals in public, in the status of voice, misrecognition referred to refugees as social subjects. Not only were they being spoken about rather than spoken to or with, refugees also became faceless and characterless figures without capacities, experiences, relationships or emotions; without ‘personhood’ – the symbolic marker of humanity as singular and autonomous social agency (Rose, 1998: 2).

Unable to invite ‘us’ to understand and feel for ‘their’ everyday lives as part of an ordinary lifeworld, now shuttered but once similar to our own, the status of voice in European news failed to produce, what Arendt refers to as, the conditions of ‘imaginative mobility’: ‘the capacity to represent the perspective of others’ in ways that thematize simple commonalities in our separate lives (Villa, 1999: 96). Rather than enabling a cosmopolitan reflexivity around the fragility of everyday life as part of our shared vulnerability as humans (Butler, 2009), European news narratives reproduced instead narratives of radical difference, thereby consolidating the ‘communitarian’ ethos of ‘our’ ways of life.

The contexts of voice

The contexts of voice refer to the historical, socio-economic or geo-political narratives within which the news situated the crisis and invited ‘us’ to make sense of it. Context, in the refugee ‘crisis' news, is thus about explanation of the reasons why refugees had arrived and about conversation on the consequences of their arrival, positive as well as negative. Across our data, there was remarkably little by way of explanation as to what led refugees to abandon home and travel to Europe: 49.4% of all articles mentioned either none or only unclear underlying reasons behind the refugee inflow. Consequently, the news attitude towards developing a rational ‘understanding’ of the predicament of refugees, as opposed to expressing ‘solidarity’ or cultivating ‘uncertainty’, remained consistently low throughout the three moments of reporting, at 11.5%, 13.7% and 6.5%, respectively. While geopolitical reasons for migration were mentioned (for instance, ‘fleeing war’) in 43.9% of the articles, a more explicit and systematic connection between crisis reporting and war reporting, normally kept as separate news sections, would have foregrounded the urgent links between, for instance, chemical gas attacks in Syria and Syrian families’ desire to flee the death zone. This lack of explanatory context lies perhaps behind the modest and largely unchanging percentages of empathy – feeling for the pain of others as-if we could be in their place (Boltanski, 1999), which remain stable at around 10% throughout the crisis, with no peak in the ‘humanitarian’ moment and plummeting to 4.75% in November 2015.

At the same time, the conversation around consequences was overwhelmingly biased towards negative impact, privileging the harmful aspects of hosting refugees: 58% of all stories employed one of four arguments against refugees: geopolitical (terrorism), economic (economic crisis), cultural (antipathy of Islam) or moral (deceit) (Figure 5). While the most frequent were geopolitical consequences (28.5% in total), with the sharpest peak occurring after the Paris terrorist attacks (49%), economic consequences (25% in total) were highest in the first stage of the event, July (31%). These results were framed not so much by ‘fear’, which only reached a peak of 10% in the ‘security’ moment, but mostly by low-intensity negative emotions, such as ‘frustration’, stable at 18% with a predictable fall to 10.7% in the ‘humanitarian’ bracket, and ‘uncertainty’, fluctuating between 13.5 and 11.7% – and again slightly dropping to 9.7%, after Alan’s death. This focus on negative consequences and emotions could be seen as a practice of symbolic bordering par excellence, in that it introduces a logic of military deterrence onto the language choices of the news. The refugee remains here a constantly negative figure: from an economic burden, he turns into a terrorist threat and from a ‘fake’ victim to a cultural outsider. OPEN IN VIEWER

Positive consequences, geopolitical, economic or cultural, were rarely mentioned: 7.6% in total. In the context of an already information-poor narrative on refugee status, on their existing competences and literacies, this proliferation of uncertainty-driven arguments came to consolidate a thoroughly dehumanizing image: refugees bring nothing useful, only problems.

In the absence of utilitarian narratives that could have shown how refugees might contribute to ‘our’ societies, the only argument in support of ‘our’ hospitality was the appeal to the European ethos of solidarity – a deontological discourse of ‘moral obligation’ that placed the imperative to act on vulnerable others without asking back at the heart of civilized Europe (Chouliaraki, 2013): 28% of articles included such a moral claim, with the percentage shooting up to 38.5% in the ‘humanitarian’ moment (Figure 6). This upward trend further correlated with the news’ discussions on measures of reception, that is specifications of the forms of action most conducive to hosting ‘them’, where ‘our’ duty to save ‘their’ lives was found in as many as a 71.7% of all September articles compared to 55.4% in July (Figure 7). While, in this Eurocentric discourse of benevolence, the news inscribed refugees into narratives of victimhood, where they figured as passive sufferers to be rescued, the ‘security’ moment of November 2015 subordinated the ethics of rescue to a militarized narrative where refugees swiftly turned into malevolent actors. Positive moral narratives now dropped from 38.5% to 24.5% (Figure 6) and the surge of self-protective rather than life-saving measures soared: 60.7% in November as opposed to 47.2% in September (Figure 7). OPEN IN VIEWER

OPEN IN VIEWER

Figure 7.

Humanitarian versus defensive actions mentioned or declared, by period.

In a pattern complementary, albeit distinct, to the subjects and status of voice, the context of voice illustrates yet another linguistic trope of symbolic bordering. Together with silencing and collectivization, at work here were strategies of decontextualisation that consistently severed the plight of refugees from its historical circumstances and impoverished public conversation on the implications of their arrival. Refugees were hardly represented as civilians fleeing from death zones in the conflicts in Syria, Afghanistan or Iraq, while the consequences of ‘their’ arrival in Europe were almost exclusively discussed in negative terms. Lacking a narrative of the causes of this arrival or the benefits from ‘our’ co-existence with ‘them’, the news further marginalized emotions of understanding or empathy towards refugees, drawing solely on a moral rhetoric of solidarity in order to forge a positive connection between ‘us’ and ‘them’ – a deontological framework that proved unstable and short-lived in the face of events.

If silencing and collectivization managed to misrecognize refugees as political and social subjects, decontextualization led to the misrecognition of refugees as historical subjects. This is because, not only were refugees deprived of their personhood, as unique individuals with their own live stories, but they were also stripped of the historical trajectories that conditioned their fate as stateless fugitives. Instead, they appeared out of nowhere as ‘ghostly figures’ who brought no good, only trouble. If, as Arendt puts it, the ‘pre-political …condition of history’ is the fact that ‘every individual life …can eventually be told as a story with beginning and end’ (1958: 184–185), then allowing refugees to articulate themselves within a coherent temporality (‘with a beginning and end’) is crucial to endowing them with historicity, with their own specific pasts and aspirations for the future – rather than locking them into the abject present of statelessness and dispossession. Their misrecognition was thus no longer due to their muting or irrelevance to ‘our’ communities, but to their a-historical existence, as beings coming from nowhere and bringing nothing.

By this token, the lack of historicity in the news also blocked ‘our’ potential to reflexively contemplate ‘their’ fates as part of a shared chain of events where the West articulates its own responsibility for, at least part of, ‘their’ suffering. Instead, the proliferation of linguistic patterns of malevolence and victimhood, in European news stories, mobilized nothing but ‘our’ own communitarian reflexes of closure and introspection.

Voice as value: Concluding reflections

Voice is, we have argued, instrumental in endowing the refugee with personhood and historicity – and hence with the potential for recognition. Who speaks and in which capacity, or ‘voice as narrative’, defines whether and how their words will be listened to in public as the words of an equal, or ‘voice as value’. This understanding of voice complements traditional views of talk as a semantic event, concerned with the content of words, by foregrounding the very ‘act of speaking’ as a political event in its own right: ‘in acting and speaking’, as Arendt put it, ‘men (sic) show who they are, reveal actively their unique personal identities and thus make their appearance in the human world’ (1958: 179). It is in this Arendtian sense that ‘voice as value’ becomes a performative act that constitutes the speaker as a human being – as a social and historical actor belonging to a political community.

Our content analysis started from an understanding of journalism as a key site for this performativity of voice. It sought to establish how the language of European news tactically distributes ‘voice as narrative’ and how, in the process, it endows the refugee with ‘recognizability’ (Butler, 2009) – how it mobilises ‘voice as value’. The aim of this analysis has been to explore the ways by which, in articulating voice, journalism shapes not only the refugee as ‘recognizable’ but also the communities of European nations as communities of exclusion or openness; as communitarian or cosmopolitan.

Our assumption has been that, in the context of Europe’s dominant discourse of humanitarian securitization, the news’ distributions of voice might reflect the linguistic tension inherent in the refugee as both victim and threat. To this end, we mapped out the language of 1,200 news articles on the crisis in terms of how they narrated the subjects, status and contexts of voice: who speaks, on which capacity and in which context.

Our analysis, which concentrated on similarities rather than differences across our data, yielded two simple but perennial insights into the question of voice. First, the distribution of voice in European news follows a strict hierarchy – one that relies on specifically journalistic strategies of bordering yet reflects and reproduces broader hierarchies of the European political spheres. Second, this hierarchy of voice leads to a triple misrecognition of refugees as political, social and historical actors, thereby keeping them firmly outside the remit of ‘our’ communities of belonging. Let us reflect on each of these implications.