Floating sanctuaries: The ethics of search and rescue at sea

Abstract

Search-and-Rescue NGOs in the Mediterranean have been increasingly criminalized. This criminalization has chilled conversation about the ethical dilemmas practitioners face. What, if any, can be the adverse byproducts of rescuing life at sea? In this article, we concentrate on the dilemmas involved in search and rescue (SAR) as rescuers have described them. Our aim is two-fold. The first is to offer a phenomenological account of search-and-rescue dilemmas. The article sheds light on the complexity and nuance of the ethical landscape of maritime rescue, revealing an intricate web of interactions acknowledged by rescuers as posing ethical challenges. The second aim is to offer a conceptual framework for what it is that SAR NGOs are, in fact, doing. We contextualize their actions within the larger terrain of ‘border externalization’, in which states have moved enforcement activities to extraterritorial zones, where human rights law is diluted or inapplicable. We thus argue that the set of norms underlying NGO rescue practices amounts to a strategy of counter-externalization. The ideal here is that a window of opportunity can be created at sea, where human rights or international law protections more broadly apply, but enforcement powers of states are suspended. By utilizing these legal gray zones to the benefit of migrants, rescuers effectively turn extraterritorial zones from spaces of lawlessness into spaces of resistance. The rescue ship thus becomes a ‘floating sanctuary’.

1. Introduction

‘it is a big difficulty for the whole SAR community that we discuss a lot of things only “among ourselves”, have little open discussions with other actors, friends and comrades. Due to the constant criminalisation, but also the political pressure, we are often forced not to openly discuss important questions, which leads to them not even surfacing in a socio-political discussion. They remain in their own discussion spaces, are cut off and invisible’. (e-mail communication).

On 4 March, 2021, an Italian prosecutor issued a long-awaited indictment against 21 members of search-and-rescue NGOs and 3 organizations for aiding and abetting illegal immigration. The charge was part of an ongoing campaign to criminalize acts of assistance and solidarity with asylum seekers and migrants, including acts intended to rescue lives at sea (Cusumano and Villa 2021). Indeed, in recent years, scholars, as well as UN actors and NGOs, have responded with a counterattack, stressing unequivocal support with such rescuers, and countering the ‘smuggling’ narrative (Mann 2020; Mezzadra 2020).

While we believe such scholarship and advocacy is extremely important, the purpose of this essay is different. The criminal-law discourse surrounding search and rescue (SAR) at sea has chilled honest conversation about the real ethical dilemmas that the practice involves. Focusing exclusively on criminalization can prevent us from asking an important question: What, if any, can be the adverse byproducts of rescuing life at sea? This question, which, as will become apparent, rescuers have long asked themselves, draws upon a tradition of critical humanitarianism, in which critique goes hand in hand with action and its assessment in real-world rather than ideal terms (Brauman 2012; Kennedy 2005).

Some might argue that these questions are better left unasked. Debating the possible adverse byproducts of rescuing life at sea is likely to cause major political harm. However careful it is conducted, they might say, such an endeavor will nevertheless prove a godsend to those seeking academic backing for NGOs’ criminalization.

We disagree with this position. Instead, we believe that a candid debate about the ethical challenges involved in SAR is crucial for both academic and political reasons. If SAR advocates and activists seek to deny the existence of the dilemmas, they will miss out on the opportunity to make their voices heard and their perspectives acknowledged. The political debate about the ethical dilemmas of SAR is already going on. The question is who can frame it and whose opinions and experiences inform it. By interviewing members of SAR staff anonymously, we aim to offer a place for reflection on ethical dilemmas where activists are able to articulate the dilemmas from their situated perspective, and help shape the nature of the debate.

We thus make a concerted effort to understand the dilemmas involved in SAR as seen from the perspective of rescuers. By adopting such a phenomenological perspective, we have deliberately chosen not to engage in our own normative analysis of the dilemmas involved in SAR. Instead, our aim is two-fold. The first is empirical and analytical. It involves a mapping exercise, serving as a phenomenology of ethical experience. Here, we seek to identify the key ethical dilemmas faced by rescuers at sea. The article sheds light on the complexity and nuance of the ethical landscape of maritime rescue, revealing an intricate web of interactions acknowledged by rescuers as posing ethical challenges.

The second aim is to offer a conceptual framework for what it is that SAR NGOs are, in fact, doing. We contextualize their actions within the larger terrain of ‘border externalization’, in which states have moved enforcement activities to extraterritorial zones, where human rights law ostensibly does not apply (Benhabib 2020; Giuffré and Moreno-Lax 2019; Shachar 2020). We thus argue that the set of norms underlying practices developed by SAR NGOs amounts to a strategy of counter-externalization. The ideal here is that a window of opportunity can be created at sea, where human rights or international law protections more broadly apply, but enforcement powers of states are suspended (Mégret 2021). By utilizing these legal gray zones to the benefit of migrants, rescuers effectively turn extraterritorial zones from spaces of lawlessness into spaces of resistance. The rescue ship thus becomes a ‘floating sanctuary’.

Our empirical data are composed of 25 semi-structured qualitative interviews conducted with 22 rescuers who have been involved in SAR operations at both the operational and planning levels. The interviews were sometimes complemented by personal e-mail communication and exchange of primary documents. Almost all of our interviewees held positions of responsibility, such as head of mission, operations coordinator, SAR coordinator, public spokesperson, and cultural mediator. At least six of our interviewees were either co-founders or members of the board of their respective organizations. None of our interviewees were contracted seafarers or short-term volunteers; all had a professional bond to their organizations and longer-term commitment to SAR.

Besides their function within the NGO, several other factors may impact rescuers’ experiences and how they make sense of it—such as the timing of activity at sea; whether the rescuers are active primarily on land, at sea, or in the air/over the phone; whether they have experience in multiple NGOs or only in one; the size and professional ethics of their NGO, etc. To the extent that we felt that it was possible to do so without endangering the anonymity of our respondents, we have provided information on these features in Table 1 and when introducing the citations in the text.

Table 1.

List of interviews

		Main area of activity			Timeframe of activity			Leadership position		NGO experience
Interview no.	Date	Land	Sea	Air/phone	2015–7	2017–9	2019–21	Yes	No	Single NGO	Multiple NGOs
Interview 1	9 April 2021	x	x		x	x		x			x
Interview 2	14 April 2021	x	x		x	x	x	x		x
Interview 3	14 April 2021	x	x				x	x		x
Interview 4	16 April 2021	x	x		x	x	x	x			x
Interview 5	20 April 2021	x	x		x	x	x	x		x
Interview 6	22 April 2021	x			x	x	x	x		x
Interview 7	22 April 2021	x	x	x	x	x	x	x			x
Interview 8 (same as 4)	22 April 2021	x	x		x	x	x	x			x
Interview 9	27 April 2021	x	x	x	x	x	x	x			x
Interview 10	10 May 2021	x					x	x		x
Interview 11 (same as 7)	10 May 2021	x	x	x	x	x	x	x			x
Interview 12	14 May 2021		x			x	x		x		x
Interview 13 (same as 6)	1 June 2021	x			x	x	x	x		x
Interview 14	10 June 2021	x			x	x	x	x		x
Interview 15	10 June 2021			x	x	x	x		x	x
Interview 16	21 October 2021		x		x				x	x
Interview 17	4 November 2021	x	x			x	x	x		x
Interview 18	4 November 2021	x	x		x	x	x	x		x
Interview 19	10 November 2021		x			x	x	x		x
Interview 20	15 November 2021	x				x	x	x		x
Interview 21	17 November 2021	x			x	x	x	x		x
Interview 22	18 November 2021			x		x	x	x		x
Interview 23	25 November 2021	x						x		x
Interview 24	29 November 2021	x	x		x	x	x	x			x
Interview 25	30 November 2021	x	x	x	x	x	x	x			x

In accordance with the aims of the article, we adopted a phenomenological approach to interviewing (Seidman 2019). We asked participants to reconstruct and reflect on their lived experiences and sought to understand the meaning they made of those experiences from the point of view of their own subjective understandings (Schutz 1967). We proceeded first through our own contacts, then by ‘purposeful selection’ (Maxwell 2013; Seidman 2019). Although all respondents answered to our questions in an individual capacity, it was important to include people from different organizations in our sample to ensure a plurality of experiences. According to the latest list published by the Fundamental Rights Agency of the European Union, there were 17 different NGOs conducting SAR operations in the Mediterranean (sea and air) between 2016 and 15 December 2020. Of these, 14 are present in our sample.¹ Some respondents belonged to the more ‘humanitarian’ organizations, others to the more ‘solidaristic’ ones. This often-reiterated division highlights the level of willingness to directly challenge state power and to articulate a political vision beyond the humanitarian one, which are stronger in solidaristic organizations (Cuttitta 2018). The activism of all interviewees focuses on the mid-Mediterranean space between Libya, Sicily, and Malta, as well as in the Aegean.

Given the backdrop of criminalization, we have decided to anonymize respondent’s identities and organizational affiliation as well. While this requires a level or generality we would not otherwise choose, we felt that it was necessary as a matter of research ethics in order to protect our sources. All interviewees have been instructed about the purpose of the study and the intended use of the data and have signed a consent form. In addition, we have circulated the draft to all interviewees and given them the opportunity to flag to us any passages of the text that they considered to be problematic. All interviewees were informed of their right to withdraw from the study at any time.

The article proceeds in five sections. After this introduction, Section 2 provides an analytical framework for understanding the ethical dilemmas that rescue staff confront. At the heart of their dilemmas is a bifurcation between the urgency of the imperative to rescue those who are at risk of drowning, and a consequentialist analysis of rescue operations. This contrast may seem all-too familiar from other humanitarian contexts (Bell and Carens 2004; Halley et al. 2018; Weber 2004). But within the context of sea rescue it has an interesting temporal and spatial structure, which we lay bare. Sections 3–6 move to a close reading of the different components of this structure, informed and illustrated by our interview material. Section 7 concludes by interpreting the practices of SAR organizations as a form of sanctuary.

2. The chain and the command

How did our interviewees perceive the ethical dilemmas of SAR? As will become clear in the following sections, SAR rescuers are not a homogeneous group, and their perspectives and responses vary considerably. Nevertheless, the basic structure of the dilemma many rescuers face is the same. It is a dilemma between the ethical imperative to rescue and the possible unintended political consequences of the act of rescuing.

An often-repeated topos in our interviews framed the act of rescuing as an emergency response to an already occurring catastrophe. The act of rescue is based on an ethical imperative to save people who are at an imminent risk of drowning, and that imperative takes precedence over all other considerations. Even though SAR activism has grown increasingly organized, funded, and to some extent bureaucratized over recent years, rescue itself retains at its core a grain of ‘good Samaritanism’: one individual helping another simply because the latter is in distress. Confronted with the drowning person, so the narrative goes, there is no choice to make and there is no dilemma. One must simply lend a helping hand. The act is a response to a command of the conscience.

However, it is not always possible to separate the event of rescue from whatever it was that led to it and from whatever occurs afterwards. The singular ethical moment of rescue can always overflow and be informed by its surrounding context of political and economic power relations. ‘There is always a chain’, as one interviewee put it felicitously, ‘in which the moment of rescue is just one link’ (Interview 1). This chain connects the former domicile of migrants and asylum seekers (often in a region of Africa or the Middle East), transit experiences and countries along the way, the maritime journey, and further transit and destination countries in Europe. Rescuers know that the conditions of possibility for the migrants’ and asylum seekers’ journeys depend on many other actors, whose actions may be questionable on ethical or moral grounds. Migration networks often overlap or intertwine with criminal and trafficking networks (Campana 2018), as well as with abusive state practices, both territorial and exterritorial (Dastyari and Hirsch 2019). And so, it is impossible to assess what happens during rescue without addressing how that event is chained to others along the way.

To rescuers, ‘the chain’ and ‘the command’ present different kinds of ethical choices. On the one hand, there is a large and potentially global infrastructure for movement, in which political and economic interests of all kinds play a major role. When thinking of the chain, ethical considerations appear as questions—should we do x or should we do y? On the other hand, an unseaworthy vessel at sea presents a command of conscience that is qualitatively different. Such a command rests on the individual value of every human life. Rather than presenting itself as a question, it presents itself as an imperative: we must do x. Here, rescue is experienced as inherently valuable—it is pursued for its own sake and not to achieve other longer-term political goals. The emerging set of norms that has gradually crystalized among diverse SAR organizations is largely about how to make these two different registers of judgment sit together. Their relationships are uneasy.

Whereas the moment of rescue is at the center of focus for rescues, it is tied by two separate ‘chains’. One leads back to the port of departure, and from there all the way to a migrant’s home (before she departed). A second chain leads forward to the sought-after destination. Think of the migrant vessel as an object at sea, tied by the two chains to the two different coasts of the Mediterranean. In mid-Mediterranean migration, the first is tied to Libya, the second to Italy (or Malta). The first connects to a territory of danger, and the second to a territory of relative safety. Figure 1 helps illustrate this basic structure.

Figure 1.

The chain and the command.

The ethical questions develop when the rescuers find themselves having to engage with either chain. An investment in Chain 1 opens the question of relationships with non-European authorities, Libyan or other, as well as with non-state actors, including smuggling and trafficking networks. An investment in Chain 2, on the other hand, raises dilemmas concerning relationships with European forces, and perhaps with criminal networks on the European side, European publics, and at times funders. The journey along the line, from left to right, is a temporal progression along the migration route. But the figure also illustrates a spatial distribution of moral experience. By way of rough generalization, the form of the question is associated with land and airspace, whereas the form of the imperative is associated with the sea.

Which specific ethical questions are important to rescuers? The danger of complicity in a criminal system of migrant exploitation by smugglers is one side of the coin. But rescuers are often even more worried by the potential of becoming complicit in abusive practices and unjust policies of European states. These two faces of the quandary take on very different shapes in rescuers’ real-life encounters. At a general level, however, they are experienced in similar ways: as the fear of dirtying their own hands and becoming part of a vicious scheme. The viciousness of smugglers and traffickers is to exploit migrants for profit. The viciousness of European state authorities is to impose violence on asylum seekers and migrants. Both can also be seen as exploiting the role of SAR NGOs in order to pursue their goals.

The danger of becoming imbricated in the exploitation and violence associated with chains 1 and 2 is particularly acute for those rescuers operating from a distance. In recent years, SAR NGOs have started complementing their naval missions with airborne operations aimed at monitoring the Mediterranean. These airborne operations locate distress cases, identify the nearest rescue assets, communicate with different actors to help coordinate a rescue, and report on human rights violations happening at sea. Currently Sea-Watch operates two monitoring aircrafts together with the Swiss Humanitarian Pilots Initiative, the Moonbird and the Seabird. In addition, other NGOs such as Alarm Phone also operate from a distance. Like with airborne operations, their mission is not to rescue but to monitor, coordinate, and bear witness.

Members of these missions face a very peculiar situation: their aim is to save lives, but they have no rescuing capabilities. When they receive a distress call or spot a vessel in distress, they cannot immediately initiate rescue. They cannot directly respond to the moral imperative of the command. Instead, they must decide how to coordinate a rescue to increase its chances of success, while trying to minimize cooptation by both sides of the chain. For them, it is impossible to isolate the moment of rescue from the wider political context. Instead, they must make assessments, define thresholds, and ultimately make difficult ethical and legal choices about which information to share. Theirs is the disembodied power of knowledge, not the embodied power of immediate physical action. We call their perspective the bird’s-eye view.

3. Chain 1: connecting to a dangerous Coast

One of the most criminalized aspects of SAR is communication with ‘the other side’: non-European individuals and groups that may belong to criminal networks, including suspected smugglers and traffickers. Generally, the categories of smuggler and trafficker are not as neat as often suggested in public discourse (Aloyo and Cusumanno 2021; Maher 2018; Sharma 2003). More importantly, these categories are difficult to assess for the rescuers, who are often not able to tell what exactly the functions of persons ‘on the other side’ are. As one interviewee explains:

we have only very limited knowledge of how things work in Libya, and things change very quickly. How should I know if a person is himself part of a criminal network, or just somebody who is being paid or forced to do something by somebody else who is part of a criminal network? The power hierarchies are very difficult to tell. (Interview 7)

If there is any communication between rescuers and smugglers before the situation of distress, state authorities in Europe see this as an indication of collusion. In addition, the idea of transfers happening at sea has been used to delegitimize the work of rescuers, who are accused of functioning like ‘taxis’ for smugglers (Allsopp, Vosyliūtė and Brenda Smialowski 2021). Communication with such individuals, or even with would-be migrants and asylum seekers, may constitute an important part of chain 1. Once such communication is established, the situation of distress at sea may seem more planned and organized, and not already-occurring independently of SAR operations. And if that is the case, it may be difficult to cabin SAR as an urgent imperative, no questions asked. The appropriateness of SAR may look to some viewers more like an immigration policy question.

In our interviews, we could identify two different types of responses to questions on communication. One type of response can be described as a kind of ‘willful ignorance’ about when and where such communication occurs.

I do not know anything besides that there was an alert by somebody—whether it be Alarm Phone, or airborne operations, or the authorities, or another ship, or even Frontex—that there is a boat in distress. As somebody who is on the boat, I don’t want to know more. And this is a conscious choice. That’s all I knew and that is by design. To protect me and to protect people on the boat. Otherwise you can do harm to yourself and to the people you are rescuing. [emphases added]. (Interview 1)

By calling this attitude ‘willful ignorance’, we are not prejudging it as morally objectionable. As the sociology of ignorance has shown, not knowing can often be a very powerful tool for managing risks (McGoey 2012). Hence, one way to interpret our interviewee’s attitude is as a legal technology to protect rescuers from prosecution. However, in the quotation above, the unwillingness to know more does not simply appear to be a ‘legal technology’; it is also an ‘ethical technology’. Apparently, by not knowing the details of communication she may better focus on what she believes is her major role, that is, to rescue. She is relieved of distractions that may make her over-think certain rescue operations, in a way that can hinder saving lives as well as larger emancipatory aims.

The second response is to question the assumption that rescue is only legitimate if accidental. As the same respondent told us: ‘It’s not called “find and rescue”, it is “search and rescue”. Why are we only legitimate when we happen to be floating around?’ In other humanitarian emergency situations, she said, states would be patrolling the border hoping to find people in need of assistance. In the Mediterranean this is different because ‘the situation in the Mediterranean is not acknowledged, not desired and not legitimized as an emergency’ (Interview 1).

This second type of response seeks to deconstruct the notion that communication implies collusion. The accusation that NGOs are carrying out ‘transfers at sea’ implies that there is no real emergency, no real risk of drowning. But our interviewees stressed that knowledge about the location of a migrant boat does not eliminate the risk of shipwreck. Migrant boats are all unseaworthy and can sink at any time, they argued, even in the middle of a rescue (Interviews 1 and 4). The goal of this response is thus to show that the accusations levied against rescuers are based on unrealistic premises about the real-life circumstances they face. Another goal is to challenge the notion that communication is enough to become a morally tainted link in the ‘chain’. According to this view, by merely communicating with persons in need of rescue to help rescue them, one does not become implicated in the upstream exploitation and violence that brought the person to sea in the first place.

A thicker link in the chain may emerge upon physical encounter with presumed smugglers. According to our interviewees, such encounters no longer occur. But in earlier years, around 2016 and 2017, they were often unavoidable. These encounters placed rescuers in a difficult situation. There were moments in which, due to urgent circumstances, rescuers felt that there was no ethical option but to respond to the presumed smuggler’s actions. In such instances, the moral imperative to act in response to the emergency trumps any other consideration:

The question of communication is a very complicated one because of how things work at sea. If you are at sea and somebody talks to you, you respond. If someone approaches you and says: ‘there is a boat’, what do you do? This is exactly the situation that the authorities criminalize, but what should you do? Do you not go there? Or if someone comes to you and says, ‘there are people dehydrating over there, give us water’, you give water. What else should you do? Once they came to us with a pregnant woman and a two-year-old child and asked us to transfer them to our boat. This is exactly the situation that European states criminalize because it is a transfer at sea, but what should you do, should you tell them, ‘no we don’t take this screaming women and child, go back to the boat and we will come in one hour.’ What about your criminal responsibility if something happens to that child and to that woman? (Interview 4)

Nevertheless, even regarding these cases, it is difficult for rescuers to exactly assess who they are interacting with. In earlier years, for example, it was common that unidentified men in fast wooden boats would come to the rescue scene, or be there waiting from the onset, in order to take the engine of the flimsy migrant boat with them. In the media, these were often presumed to be smugglers. Images of such bystanders detachedly observing a rescue have been widely used to delegitimize the work of SAR NGOs, and as purportedly incriminating material in criminal proceedings for collusion with smugglers. But some rescuers prefer to call these men ‘engine fishers’, a terminology that suspends judgment on the role of these individuals in the chain of exploitation. As our interviewees explained, for them it is impossible to know what the true role of these people is. Are they part of a criminal smuggling network? Are they fishermen looking for a source of extra income (e.g. by selling engines)? Have they been forced by smugglers to perform this task of collecting engines, and are therefore themselves exploited? Any of these scenarios could be true. Rescuers therefore feel compelled to adopt a kind of methodological presumption of innocence, calling these people by what they do: they pull engines from the water, hence they are ‘engine fishers’ (Interviews 4 and 7).

Gradually moving to more robust links to the chain that may tie the rescue vessel to questionable practice, consider that some rescuees have been trafficked for sexual exploitation. When this is the case, the links on the victim’s way from their domicile to a European destination may largely be made of instances of rape, torture, and embezzlement (Al-Dayel, Anfinson and Anfinson 2021). Our interviewees were aware of this reality, which becomes most clearly observable in instances when a man says that he is married to a certain woman, but the woman denies such a relationship (Interviews 1, 2, 4, 7–9, and 11). In terms of the analysis of links in a chain, this may mean that one problematic link—the man—is present on board. But here too, the exact role the man plays within the chain of exploitation is unclear. He may be on board for a fee, or he may be playing this role in return for a free journey in earlier legs of the trail. The fact that he is momentarily serving a trafficking network does not in itself mean that he is not legitimately seeking asylum.

Interviewees have hesitated on whether it is appropriate to actively seek to understand what is going on in such misrepresentations of relationship. Some were adamant not to prejudge anyone, and not to slip into the role of law enforcement (which would make them links in ‘Chain 2’). As they explained, they are not responsible and not trained for that, and they want to avoid becoming imbricated in security. Their concern is safety, not security (Interviews 1, 2, 4, 7–9, and 11). These rescuers feel like they must demarcate the boundaries of their responsibility. Note in this context the role played by the distinction between land and sea. European soil is imagined as the territory of law enforcement. The sea, perhaps because it is at least partly beyond state jurisdiction, is imagined as an area where such policy questions and specifically law enforcement need to be suspended:

We are not responsible for knowing whether people were trafficked or not. If somebody tells us that they have been trafficked or tortured, we can make a proper referral. But the limits of what we can do is a referral. We are not on land. There are other organizations who can help. We need to hand people over to the next leg of the journey. There needs to be trust. There needs to be a limit, or a boundary, and you need to set your remit as an NGO. (Interview 1)

We always separate men from women and children. No man is allowed in the safe place. But there can always be the case that someone accuses someone else of being a trafficker. We are not police enforcement and will not do the enforcement part. We will ask the accuser what we can do in terms of safety. We can refer them to protection and potentially help them disembark separately. But to do more on board is impossible. We are not really trained for that. (Interview 8)

In all these instances, opening the isolated space of the ship to the far-reaching chain means a risk of inadvertently becoming part of a vicious scheme. There seems, however, to be a lack of agreement, or perhaps a lack of openness, on the extent to which such complicity is a real possibility. Some of our interviewees denied it altogether, saying that it is ‘a constructed dilemma’ (Interview 2) or ‘no dilemma at all’ (Interviews 1, 5, and 6). According to this view, just by being at sea and helping somebody in emergency, one is not complicit with that person’s exploitation. An ambulance or person providing care to the victim of a car accident is not responsible for another’s negligent driving (Interviews 6 and 11).

Moreover, when it comes to allegations of constituting a ‘pull factor’, rescuers pointed out that this is a narrative that obscures the real drivers of migration, which are systemic in nature. As one interviewee said: ‘It is not a couple of ships that are going to change migration flows from Africa to Europe. The answer to this question is not to be found at sea, unfortunately’ (Interview 6). Another interviewee explained that scholarship has disproved, time and time again, the narrative about ‘pull factors’ (Interview 15). Indeed, available academic research suggests that NGOs have not served as a pull factor (Cusumano and Villa 2021; Heller and Pezzani 2017). It shows that, even if some of their operations might have occasionally helped human smugglers, their cumulative effect has reduced deaths at sea without encouraging departures. For some our interviewees, the fact that NGOs do not seem to have served as a pull factor is sufficient to eliminate the dilemma.

Yet other interviewees did acknowledge the existence of a potential connection between the command and the chain in one way or another. They conveyed how smugglers react to their presence by adapting their strategies. In the words of one respondent: ‘We need to live with that. We are part of a chain. The trafficker, smugglers, flight helpers—whatever you want to call them—they know about our existence, and they are not stupid. It is a pervert system’ (Interview 4). Another interviewee mentioned rumors that smugglers charge higher prices when there are SAR ships at sea, but added: ‘they also charge more when there is good weather’ (Interview 5). He compared the rescuers’ role with that of drug substitution policy, which do not combat the addiction, but manage to do away with the violence and criminality associated with drug consumption.

4. Chain 2: connecting to a port of relative safety

The ethical dilemmas of chain 1—especially the risk of colluding with smugglers and traffickers—are somewhat more familiar, though perhaps in caricatured form, to European publics. They have been fueled by anti-migrant sentiments and by media hungry for headlines. The ethical dilemmas of Chain 2, more salient to many of the rescuers we spoke to, are perhaps less familiar.

For many interviewees, complicity with exploitative practices by smugglers was secondary to concerns about assisting European states in upholding an unjust or inhuman border regime. This concern presents itself in many ways. One is a mirror image of the notion that by their presence NGOs might be fueling or facilitating the migrant smuggling business. With regards to abusive state practices, some members of rescue NGOs expressed fears that their mere presence might be indirectly benefiting European states in maintaining an abusive system (Cusumano and Pattison 2018). Some interviewees detailed concerns that the presence of NGOs releases states of their obligations: ‘Even if they are fighting us, they still rely on us. The NGOs provide the veneer that at least something is being done’ (Interview 4).

Chain 2 becomes relevant, for example, when SAR NGOs must comply with spurious regulations that they disagree with. Such controversial rules were established when, in 2017, Italy promulgated an EU-sponsored SAR ‘code of conduct’. The document made permission for NGO vessels to disembark migrants in Italian ports conditional on collaborating in the fight against smugglers and accepting the presence of law enforcement personnel on board (Cusumano 2019). Most NGOs active in the Central Mediterranean signed the code. But many resented what they saw as an effort to enlist them into becoming an extended arm of the authorities.

Some interviewees also expressed ethical concerns about enforcing policies of migrant selection and categorization (Interviews 1 and 9). Many rescuers are motivated not only by humanitarian but also by political ideals. Activists for open borders are critical of migration control policies that impose strict criteria on who can move (e.g. by differentiating between political refugees and people fleeing destitution or natural disasters) (Jones 2016). And yet, when they are negotiating with European authorities to be able to disembark people in a port of safety, states will ask them to provide information about the gender, age, and nationality of those on board, as well as how many are likely to be unaccompanied minors, asylum seekers, etc. This information is then used in negotiations for authorization to disembark, thus likely impacting the chances of those rescued to be granted permission to stay. The ship is, therefore, tied to the chain of what happens after disembarkation. One interviewee told us:

They’ll ask you how many asylum seekers there are, etc. but I don’t agree with these categories. So we try to find out as little as possible. We only ask the people what we need to know in order to tell the authorities. Otherwise, what they want to tell you, they tell you. You don’t ask questions. (Interview 1)

The ethical questions discussed so far refer to indirect effects of rescuers’ actions in terms of becoming ‘extended arms’ of the authorities—through NGO’s unintended role in upholding policies of border securitization, migrant selectivity, and differential treatment. There are, however, certain instances, where NGO members may feel that they are not only being abused but also complicit with abusive state practices. Thus, interviewees told us about cases where they had been witnesses to state authorities using violence against migrants, or to military ships denying rescue to migrants in distress—but felt compelled to remain silent (Interviews 3 and 5). They did so for fear of sanctions, or to buy themselves more leverage in terms of being allowed to carry out their activities.

All these decisions are extremely controversial inside NGOs and accompanied by difficult considerations (see Dadusc and Mudu 2020). They generate ‘deep disagreements’ within the NGO community and trigger discussions about the ultimate goals of the movement and ‘the kind of politics that SAR NGOs ought to be pursuing’ (Interview 14). ‘We had a very strong debate internally on whether we should report on violence and try to keep authorities accountable. Being a witness to these things without reporting means being a part of them. We considered developing an evidence-gathering protocol. But then we consulted a criminal lawyer and he said that if we did that we could be accused of conspiracy’ (Interview 3). Another interviewee had similar doubts: ‘Are we part of the state system or not? That is the ethical question. But we are not doing this to help the state. We are helping the people. If we are not there, no one will be there. At the end our answer is always the same: we need to comply in order to be able to keep on doing this [humanitarian SAR operations]’ (Interview 5).

5. The command

If rescuers do not act upon the humanitarian imperative to save lives in the face of the emergency, they act against their principles. But if they do act, they may end up implicated in wrongs committed by exploitative migration networks, or by states. How do SAR NGOs respond to this dilemma? Through standard operating procedures and internal codes of conduct reflecting a deep-seated set of cultural norms, they try to isolate the rescue vessel as much as possible from what happens both before rescue and after disembarkation.

Perhaps the most straightforward way to see the work of insulation is in its ‘negative’ embodiments, that is, in rescuers’ attempts to actively push the intervention of the state out of their way. Under the law of the sea, the vessel is a highly regulated environment. Generating an ethically isolatable environment does not mean eschewing such regulation. Rescue teams are aware that states attempt to intervene and eliminate their autonomy by reference to maritime safety regulations and may therefore be very cognizant of them. But they may still try to stay rather autonomous, sometimes by use of ‘flags of convenience’; and in other times by choices of where and how to register—whether as recreational vessel or otherwise (see e.g. Schatz and Endemann 2019).

Contrary to what the use of such flags may suggest, this negative removal from the state’s reach is not merely a matter of ‘convenience’. It is also an ethical matter. Thus, for example, rescuers may be very strict in not letting armed agents of the state board their ship, even if they otherwise agree to an inspection (Interview 5). The weapon symbolizes a possibility of violent disembarkation of rescuees and may suggest to them the potential of state violence. Carving out spaces free of armed state agents of course has a long history within universities, hospitals, and other institutions that may be compared with the rescue vessel. Flags of convenience have a bad reputation, often standing for attempts to eschew statist rule of law; but when the state is a major part of a vicious scheme, autonomy from it may be perceived as upholding international law or higher ethical standards.

Beyond such negative or protective measures directed at the state, there are also positive aspects to how the experience of command is constructed. Some are discursive. One that lies between the negative and the positive is the practice of some NGOs—most notably Sea-Watch—of calling rescued migrants ‘guests’. On its negative side, this label seeks to shake away categories of immigration law. ‘In this space’, told us one interviewee, ‘we do not have asylum seekers or refugees, because that would also mean that we would have people who are not in that group, and we also do not have migrants. We never call them anything but people or guests’ (Interview 1). On its positive side, the term guest suggests a kind of hospitality that implies that certain questions not be asked, but also that we give them a shelter and nourishment (Derrida and Dufourmantelle 2000; Levinas and Lingis 1969).

The practice of eschewing categories associated with migration policy is common also to organizations on the more humanitarian side of the spectrum, albeit in a different way. For example, in its public communication SOS Méditerranée always refers to those rescued as ‘the survivors’, never as migrants or asylum seekers. One interviewee from a humanitarian organization explained this particular form of ‘willful ignorance’ as an ethical choice emanating from the obligation to focus on the command:

We avoid talking about asylum seekers. That is already political. We call them people in distress at sea. I don’t know if they are tourists, survivors, or whatever. I don’t know, this is the job of the authorities, they need to protect the people. To us, talking about ‘migrants’ in the Mediterranean is already politicizing the situation. (…) It [this attitude] is not just playing dumb. Modesty is not part of the humanitarian principles, but is should be. If you know too much, you get crazy. It’s already a big job to rescue and bring people to a place of safety. We need to know some things and read, but ultimately we need to focus on rescue. It’s not only a posture or tactic of being very neutral and getting my way out. It is also a proactive position of modesty and transparency. (Interview 20)

Beyond discursive strategies, NGOs also have internal policies and practices that seek to give precedence to the command. For example, internal policies forbid crew members to inquire into the personal life histories of rescuees or to take pictures. They also establish procedures to deal with specific vulnerabilities if these are pointed out by the migrants themselves, but without alerting law enforcement authorities. Vulnerabilities are thus experienced as generating interpersonal obligations to those who suffer from them. At the same time, there is a real hesitation about framing these obligations as addressed to the state, or in the bureaucratized vocabularies of criminal or asylum law, in which they immediately also become instruments of potential exclusion.

This hesitation exists even in instances where security issues arise on board. In recent years, rescue ships have been subject to long standoffs at sea while waiting to be assigned a port of safety. This wait has been described by rescuers as yet another form of violence, which generates feelings of suffering and despair, and sometimes also of revolt. It also leads to a deterioration of the living conditions on board, resulting in a torturous form of confinement (Stierl 2021; Tazzioli and De Genova 2020). We have been told about several dangerous situations on board during standoffs (Interviews 17–19). One interviewee told us that a group of migrants once put fire on oil tanks in the deck as a form of protest against the excruciating wait. In another case, a violent fight arose between two different groups of rescuees and one person almost got killed. In yet another instance, migrants attempted to enter the bridge and take control of the ship. Importantly, even in these situations, the possibility to alert law enforcement authorities is often portrayed as a measure of last resort that should be avoided. Consider this statement by a member of a large humanitarian organization:

Once we had two young people who fought with one another, and one of them strangled the other until he was unconscious. Also there, we tried to solve the situation through mediation. (…) In part, this is a pragmatic decision, in part an ethical one. As the head of mission, I am responsible for the collective. I cannot tolerate that someone kills somebody else. I cannot tolerate that the food is wrongly distributed. I cannot tolerate that there are fights on board. Although I am a humanitarian, in these cases I usually decide for hierarchy. But only so long as I can bear it [ethically]. My threshold is, for example, that I do not want to be responsible for turning somebody in to law enforcement. I don’t say that on board, but nevertheless it is so. Given everything that these people have endured, what I would not do is to call Malta and say: ‘these two people have tried to put fire on my ship’. That, I wouldn’t do. I know that this makes me inconsistent. But that’s the way it is. I would not deliver them to law enforcement authorities. I would try to avoid that. And if the captain wanted to do that, I would try to convince him otherwise. (Interview 18)

The desire to avoid law enforcement as much as possible reflects yet again the aspiration to isolate the space of the ship, legally and materially. The rescue ship functions as a ‘human rights ex-territory’, a space where the humanitarian imperative rules, where human rights laws apply but state power is held at bay—at least as long as persons on board are not put in dire risk.

6. The birds’-eye view

The structure of the chain and the command assumes a certain spatial division between land and sea that does not completely capture the experiences of all SAR rescuers. Two aspects of the SAR activity in particular escape this dichotomy. One is airborne surveillance, which in recent years has become an important part of SAR in the Mediterranean. Another is communication by satellite phone, as carried out by Alarm Phone. Alarm Phone is a transnational activist network that functions as a ‘hotline’ that anyone can call to make a boat’s location known and ask for rescue. It runs exclusively on volunteers, builds coalitions with activists in the Global South, and is operative in all regions of the Mediterranean (Stierl 2016). When migrants in distress are in possession of a satellite phone, they can directly call Alarm Phone from the sea. Alarm Phone members then offer information, advice, and the possibility of raising public alarm, sometimes discussing what steps to take together with the migrants themselves—a crucial characteristic that differentiates this organization from other NGOs.

Alarm Phone and airborne NGOs are key actors in the SAR process and rescue ships collaborate with them actively. Think of these two aspects as ‘the birds’ eye view’. Whether by use of planes or satellite phones, these aspects of SAR flutter above the schematic reconstruction of experience in Fig. 1. From this position certain epistemic limitations are no longer in place. It is easier to know more. At the same time, the possibilities for independent action are severely limited. Rescue can only be pursued by relaying information onward—to states, merchant vessels, or NGO vessels.

When a ‘spotter’ hovers above the Mediterranean Sea, for example, she or he have longer temporal access to the journey and can potentially observe a vessel long before it is accessed by any other vessel capable to assist (Interviews 7 and 9). When and whether to trigger rescue may thus become a planned decision, rather than an ethical imperative of here and now. Similarly, when Alarm Phone operates an emergency number for migrants in distress to dial to, the call can be received at different stages of the journey. It may or may not be interpreted as an immediate distress case, which will impact the response. At stake is not only the protection of life, but also access to sought-after migration destinations and opportunities for dignified livelihoods (Interview 15). Particularly for organizations on the more ‘solidaristic’ side, the relevant judgment call is particularly complex. It must weigh the probability of gaining access to a safe port against the probability of drowning or refoulement.

On a more general level, from the birds’ eye view it appears nearly impossible to insulate the chain from the command. From this position , therefore, activists experience something much more resemblant of an unabated chain. Ethical quandaries that activists located at sea aim to avoid are consequently uploaded to the flying shoulders of those providing surveillance or communication. But the latter are more amenable to becoming the long arms of states. When they share information with states, their point of view can end up facilitating preemptory rescue operations by Libyan or other non-European forces, which ends up returning migrants to where they came from, blocking access to asylum, and possibly risking ill-treatment there. Such ‘pullbacks’ are one especially problematic aspect of the externalization of border control, which scholars have called ‘contactless control’ or ‘refoulement by proxy’ (Giuffré and Moreno-Lax 2019; Heller and Pezzani 2016; Moreno-Lax 2018; Pijnenburg 2020).

At the center of the dilemma faced by airborne and telephone operations is a double bind that is a quintessential experience of SAR activity from the birds’ eye view. Because they are critical for saving lives and granting access to asylum and opportunities, members of NGOs performing these functions are often accused of complicity with the ‘business model of smugglers’. This is particularly true of those actors engaging in communication by phone with non-state actors outside of Europe. Yet, because they have full knowledge and can alert the authorities, the birds’ eye infrastructure has also been incredibly important for states, including in their questionable activities. They have enabled pullbacks, possibly contributing to violations of the principle of non-refoulement. In both cases, the issue of communication is key. Communication with migrant networks allows rescuers to join the migrant ‘underground railroad’ (Heller, Pezzani and Stierl 2019; Stierl 2020): networks that migrants have established for their own empowerment and liberation. Communication with the authorities is a necessary condition for rescuers to be able to fulfill their humanitarian mission of saving lives.

The birds’ eye view, and especially Alarm Phone activity, is highly contested both outside and within the SAR community (Interviews 9, 12, and 15). Here, the command is closest to entirely collapsing into an unhindered chain. For some activists all that really remains is a political struggle for freedom of movement and liberation, and against it, with little room for good Samaritanism. For others, this amounts to a betrayal of the key principles of humanitarianism and places the legitimacy of the whole SAR movement at risk. As one interviewee put it, by making the whole enterprise about freedom of movement rather than saving lives, political NGOs are ‘de-urgentifying’ the narrative. In his view, they are also undermining the normative justification for their actions before public opinion and distancing themselves from the legal basis that enables a claim to disembarkation (Interview 12). Yet, the situation at sea -- in which the disconnect between chain and command becomes possible -- arguably depends on the entirely different conditions in which the birds’ eye activists work. Without the plane and especially the phone, the impact of SAR activities would be much more limited.

7. Conclusion: sanctuary at sea

Over the last three decades or so, a vast literature on ‘externalization’ has documented and analyzed the ways in which states have established exterritorial border controls (Frelick, Kysel and Podkul 2016; Gammeltoft-Hansen 2011; Ghezelbash 2020; Spijkerboer 2018). As early as 2006, Gibney identified the creation of ‘a thousand little Guantanamos’, spaces where states act free from the constraints imposed on them by courts, international and domestic law, human rights groups, and the public at large (Gibney 2006). Mediterranean migration routes specifically have been major sites of externalization efforts for the EU and its member states, creating ‘maritime legal black holes’ (Mann 2018).

Adding to these earlier analyses, Benhabib (2020) highlights the apparent paradox that externalization strategies actually serve to reinvent territorial sovereignty: because they generate ‘a dual movement of deterritorialization and territorialization’ that threatens the applicability of the 1951 Convention. In a recent essay, Shachar (2020, pp.73–96) suggests two strategies to counter the threat of this dual movement. First, extending jurisdiction to where control is exercised by making human rights and constitutional provisions apply irrespective of the location of border control activities. Second, relaxing the connection between accessing territory and seeking asylum so that protection can be sought and granted extraterritorially.

SAR NGOs enact a third normative response to the threats of state-led deterritorialization tactics. They seek to create a different kind of deterritorialization, one that establishes spaces of resistance against state sovereignty. By adopting internal regulations that suspend as much as possible the applicability of criminal and migration laws, rescuers seek to transform the ship into a floating sanctuary.

To function as a sanctuary, rescue vessels must be able to shield themselves from responsibility for enforcing state’s laws. The way to do so is through the technique of willful blindness—akin to the enactment of internal ‘firewall policies’ (Carens 2013)—aimed at ensuring that only those vulnerabilities are known that can be addressed from a humanitarian (not criminal) perspective. It is an exceptional space, but unlike Agamben’s (1998) spaces of exception, the camp or the zone d’attentes, it is generated by civil society actors rather than by sovereign decree. To be sure, the rescue ship is a highly regulated environment. From a legal perspective, the ship remains under the jurisdiction of its flag state and sometimes under other overlapping state jurisdictions as well. But many of the internal rules and standardized procedures established by rescuers can be read as attempts to create a state of exception where national immigration and criminal laws are temporarily suspended. Instead, rescuers claim to be enforcing the ‘higher law’—international law—where states fail to uphold it (Mégret 2021).

Rescue ships function in a similar way as sanctuaries in that they constitute themselves as spaces where migration categories do not apply and where, for a limited period of time, every person is entitled to stay under a condition of temporary amnesty (on sanctuary, see e.g. Lasch et al. 2018; Lippert and Rehaag 2014; Marfleet 2011; Mourão Permoser 2021). To say that the rescue ship functions as a sanctuary is not to say that living conditions within it are ideal. This, in fact, has never been a feature of sanctuaries. To the contrary, historically the amnesty provided by sanctuaries has always represented a temporary, tenuous, and often torturous condition. A fugitive who took refuge in a medieval church, for example, ‘might be required to pay a fine, forfeit his goods, perform penance, or go into exile, but almost without exception his body and his life were to be preserved’ (Shoemaker 2011, p. ix).

The sanctuary is not an ideal state of affairs, and it is also not a solution capable of bringing an end to injustice for all persons in all places. Also in that sense sanctuary is and remains always a state of exception. But it is also a state of righteousness: A space where a wrongdoing is corrected, where ‘the higher law’ is implemented, and where the spirit of the law is protected and upheld against undue attempts to dilute or deform it. Ideally, the rescue ship seeks to eliminate its own role as sanctuary. If conditions are favorable, it will be granted debarkation as quickly as possible. This too does not remove it from the vaster tradition of sanctuary. When deliverance comes, all sanctuaries will no longer be needed and will thus be eliminated.

What role does the birds’ eye view serve in this construction of floating sanctuaries? As explained above, from this perspective the phenomenology of the chain and the command no longer holds. There is a continuous chain of political and economic power, as well as physical violence. The question where this chain stops, for the sake of the moral imperative to rescue, becomes subject to temporal and spatial planning—now decoupled from the ability to act or intervene directly. This perspective is crucial for such floating sanctuaries to be effective.

This vantage point offers an insight for scholars aiming to understand sanctuaries beyond the specific context of maritime space. All sanctuaries exist only in the context of a complicated environment of larger political and economic power. The birds’ eye vantage point in the context of sanctuaries at sea requires close attention to the technologies and form of civil society and state engagements in the larger environments that make specific sanctuary arrangements possible. It raises our attention as researchers to the fact that sanctuaries do not simply appear as self-constituting islands. If they are indeed ‘islands’ in which power functions differently, that is only due to a myriad of strategies and decisions that precede them.

Acknowledgments

Our greatest dept and most heartfelt gratitude go to all the SAR rescuers and activists who have so generously shared their knowledge and experiences with us. In addition, for their helpful feedback on earlier drafts, we would like to thank the two anonymous reviewers as well Rainer Bauböck, Benjamin Boudou, Joseph Carens, Omri Grinberg, Sonia Fleury, Matthew Hoye, David Owen, Moria Paz, Martin Ruhs, Martin Senn, Maurice Stierl, and Meropi Tzanetakis. This article was presented at the workshop on ‘The Ethics of Migration Policy Dilemmas’ of the European University Institute, at the STS Migtec Circle, at the TrafLab workshop at Tel Aviv University, and at the departmental seminar of the Department of Political Science of the University of Innsbruck. We would also like to thank the participants in these forums for their comments and suggestions.

Funding

This work was supported in whole, or in part, by the Austrian Science Fund (FWF) [‘Migration as Morality Politics’, Grant V 743-G].

Conflict of interest statement. None declared.

Footnotes

1

The three organizations that are not present in our sample are ‘Association Pilotes Volontaires’, 'Mediterranea Saving Humans’ and ‘Lifeboat’.

Author notes

Itamar Mann and Julia Mourão Permoser contributed equally to the work.