Schizophrenic agendas in the EU’s external actions in Mali


 Because of its geographical proximity to the EU, the Sahel region's perceived cross-border security threats of terror, migration and organized crime have become a top European security policy priority. Contributing to existing debates on EU external actions, the article develops the idea that the EU's engagement in the Sahel has become an attempt to construct and confirm the Union's ability to act as a global security actor. Through the analytical lens of ‘translation’ emphasizing the continuous transformation of norms and ideas by actors and contexts, the article examines how EU staff implement shifting policy objectives in their security practices in Mali; the effects these intervention practices produce; and how, in turn, these effects reflect back upon the EU's role as a security actor. The article shows how the EU's actorness and ability to perform security are hampered by the lack of buy-in from their local partners, as narrating success in a context of escalating violence becomes increasingly implausible. Thus, we argue that while the EU's activities in Mali reinforce the idea of the EU as a security actor, the limited character and impact of the EU's activities on the ground also reinforce the idea of it as a limited or even ineffective actor. The article thereby provides fruitful input to discussions of the constitutive effects of everyday practices in interventions and the production of EU security actorness as played out in the Sahel.

International Affairs 96: 4, 2020 despite numerous international interventions, the jihadist insurgents are closer to the capital of Bamako than they were in 2013. 4 Among other international actors (notably France and the UN), the EU has responded to the growing conflict, working through a multifaceted engagement comprising continued economic assistance, diplomatic partnerships, and security-informed initiatives such as the deployment of three CSDP missions in the Sahel to reform the security forces of Mali and Niger.
As violence continues in the Sahel, EU action in Mali is increasingly framed by the country's proximity to Europe, and as a result the region is constructed by EU policy-makers as an intervention space where territorial border control is enacted at a distance. 5 The declared European 'refugee crisis' of 2015 accelerated the EU's inward-looking turn in a foreign policy that seeks to pursue EU internal security objectives through actions abroad. One of the EU staff serving in the capacitybuilding (EUCAP) mission described the situation as follows: When I started working for the EU in Mali, the focus was on the long-term perspective, aiming to help Sahelian countries build future government structures and security institutions. Then everything changed in the summer of 2015 after the European migration crisis.
We were told that we should 'hold our horses' with the long-term perspectives of state reform and change our focus to short-term activities relating to migration pressures and counterterrorism. 6 The Sahel is a key experimental arena for the EU's security-migration-counterterrorism nexus in a rapidly changing political security context. Responding to the Sahelian crisis has therefore become an opportunity for the EU to strengthen its role as a global security actor. This article analyses the effects of the processes through which the EU pursues these objectives.
The EU intervention in the Sahel region has become a focus of growing scholarly debate, within which the region has been dubbed a 'laboratory of experimentation'. 7 The policy shift away from promoting peace towards securing borders has been interrogated, and both the EU's security 'actorness' and its normative claim to doing good in the world have been challenged. 8  criticized the policy shift from a 'security-development nexus' to a 'securitymigration-development nexus', 9 seeing it as indicative of a foreign policy driven by the interests of the European member states rather than by a contextualized analysis of the needs of the African countries in question. 10 The EU as a security actor has been analysed in terms of both discourses and practices. 11 In approaches focused on discourses, attention has been concentrated on explaining EU 'actorness' through the formulation of policy objectives, whereas practice-based approaches have focused on evaluating the EU's engagements on the basis of their ability to achieve predefined policy objectives.
In contributing to debates on the EU's external actions, we develop the idea that the EU's engagement in the Sahel has become an attempt to construct and confirm the Union's ability to act as a global security actor (i.e. its 'actorness'). 12 We argue that EU actorness is constituted (partly) as an effect of the processes through which EU staff seek to implement (at times conflicting) policy objectives in their everyday practice; that the performative act of intervention and the intervention actors are mutually constitutive. 13 The article's overarching research question, therefore, is twofold: how are EU policy objectives translated by EU staff, and how does this process shape the ability of the EU to perform security in the Sahel?
We address this question by focusing on how EU staff 'translate' sometimes conflicting ideas of security in their daily routines when implementing policy objectives formulated in Brussels. The notion of translation, as we use it, emphasizes the continuous transformation of norms and ideas, as opposed to the diffusion of relatively fixed constructs unaffected by actors and contexts along the way. 14 Specifically, we explore examples of how staff in the two CSDP missions and the International Affairs 96: 4, 2020 EU Commission in Mali interpret their intervention engagements as they manoeuvre between conflicting policy objectives, negotiated mandates, contentious security partnerships and shifting organizational pressures and priorities. The article is structured in three parts. First, drawing on critical security studies and the analytical tool of translation, 15 we introduce a conceptual framework within which to approach how individual EU staff translate policy objectives and the 'constitutive effects' of their practices. 16 In the second section, the analysis begins by situating the shifting and sometimes competing policy agendas defined by various organizational units in Brussels. The analytical framework is then applied at the level of policy implementation, via three illustrative examples of how EU officials constitute their own understanding of security as they translate competing policy objectives into practice. The article concludes with a discussion of the effects of EU intervention practice and how this reflects back on the EU's role as a security actor.

Conceptualizing security translation
We approach security as it is defined within critical security studies, 17 as a socially constructed phenomenon, with an emphasis on the intrinsic links between processes of producing security, power and processes of exclusion. Inspired by the 'Paris School', 18 we take a broad approach to how this concept unfolds in intervention practice. This is not to say that security as discourse cannot precede action, but rather that security is enacted and not 'just' spoken; 19 that '"(in)securitization processes" are producing effects through the daily routines of the security professionals'. 20 Thus, while acknowledging discourse as a social practice, this article is particularly concerned with the analysis of how social action produces its own constitutive effects that-in the case considered here-come to shape EU actorness. 21 To trace how security effects are produced in daily practices, we introduce the concept of 'translation', which denotes the ways in which ideas of security are adapted, deflected and modified as they pass through the different layers of an organization. 22  offer scripts into which others can be recruited. In this sense their interpretations are performative: they prove themselves by transforming the world in conformity with their perspective on the world.' 23 Inspired by Latour, we approach intervening actors as signifying agents (i.e. 'translators') actively engaged in the production of meaning and the construction of intervention realities in which action is deemed necessary. The purpose is to show how, in unforeseeable ways, security projects and programmes become real through the work of generating and translating interests by tying in supporters and thereby sustaining interpretations. 24 The concept of translation thus refers to the mutual enrolment and interlocking of interests that produce project realities and security contexts. However, ideas of security do not float around freely. They meet resistance and are shaped by their encounters with stakeholders on the ground ('development brokers'), 25 who have their own agendas and try to bend these ideas of security accordingly; these include EU staff, who also bridge and mould policies and objectives to meet different purposes. Importantly, approaching securitization through the concept of translation allows us to demonstrate that securitization is not a homogeneous, well-orchestrated process that happens primarily at the level of policy-making. It is a highly messy process, in which different staff members, departments and organizational actors seek to exert influence and work according to their own ideas of how the EU should perform its actorness. To make sense of this highly messy process, we distinguish between two separate levels of public policy. 26 In the first there are 'policy choices'-intentions, objectives and decisions made by politicians, civil servants or other authorities through which actionable policy emerges. In the second, 'policy implementation' denotes the processes whereby policy choices are put into action, and which produce effects as interventions unfold. 27 In particular, the article directs its analytical focus onto the processes through which EU security professionals working in the intervention theatre in Mali implement and sometimes resist policy choices formulated by a range of EU institutions in Brussels. In doing so, we offer an analysis of EU security actorness as an intervention practice that is inseparable from the 'constitutive effects' that are produced in the everyday practices of staff.
Methodologically, the analysis is inspired by Nader's notion of 'studying up' to unpack the 'black box' of powerful organizations. 28  people who wield power-corporations, the government, the wealthy, scientists, the police and so on-and the subjective dynamics that shape their practices. This, first and foremost, requires an ethnographically informed approach that encompasses actually occurring intervention practices where concepts such as security gain meaning. In the present case, we unpack the EU by entering the relatively under-studied spaces of military and diplomatic offices, headquarters and military camps. In these sites, deep-seated policy unit and implementing organization practices, as well as the personal relationships, beliefs, values and motivations of individual staff members, can affect how security is practised. 29 The article primarily builds on data gathered during two short periods of fieldwork conducted in Mali in 2019. Granted access to EU staff, we conducted interviews, made field observations, held informal chats, attended security briefings in headquarters, participated in training at civilian and military training sites, and attended meetings at the restaurants and hotels designated 'safe zones' in Bamako. 30 We aimed to explore security translators' activities; to discover how development officers bridge and bend existing instruments to deliver according to the expectations of their Malian partners while adhering to sometimes shifting policy objectives on different organizational levels, and what this translation tells us about the logics and the constitutive effects of interventions. The focus of the analysis-and its conclusions-is therefore on the voices and perceptions of EU professionals in their translation of the EU's security policy as they implement it in Mali. The way in which Malian stakeholders perceive and respond to the role played by the EU in the region is included to only a limited extent, as this has been described to some degree elsewhere. 31

Translating competing security objectives: twisted tools and schizophrenic agendas
In 2011 the EU adopted its first 'Sahel strategy', which hinged on the idea that the decades-long EU development efforts in the region should be supported by a new focus on promoting security. Before that, the EU's engagement with the Sahelian countries, including Mali, had consisted essentially of providing development aid. 32 The opening line of the EU's Sahel strategy states: 'Security and development in the Sahel cannot be separated ... helping these countries achieve security is 29 Cold-Ravnkilde et al., 'Global norms and heterogeneous development organizations'. 30 In addition to the main data collected during two short fieldwork trips to Mali, the article also draws on the authors' prior extensive fieldwork experiences. One has worked mainly on various aspects of the EU CSDP at policy-making level in Brussels; the other has conducted long-term fieldwork in Mali, including within the UN mission in Mali (MINUSMA) in 2016. 31 34 As such, the CSDP missions were initially focused on providing long-term security sector reform to the country as the main objective of their mandates. The EUTM Mali operation has its headquarters located in Bamako, and is a non-executive mission-that is, the operation supports the host nation through an advisory role only. Its principal objective is to strengthen the capacity of the Malian armed forces, by 'training the trainers' and by providing advisory and educational support for Malian security sector reform. The mission has just over 600 staff from 23 member states. 35 The civilian EUCAP mission in Mali was established to support the internal security forces, including the Malian police, gendarmerie and national guard, and relevant ministries, with training and strategic guidance. The EUCAP Mali mission consists of approximately 250 staff, one-third of whom are Malians. 36 The EU also has a range of policies and instruments in play under the aegis of the EU Commission, including its humanitarian aid programme. These various instruments and policies are coordinated by the European External Action Service (EEAS) delegation in Bamako.
In 2015, the 2011 Sahel strategy was reinforced by the EU's Sahel Regional Action Plan 2015-2020. The declared 'refugee crisis' of the same year served to accelerate the formulation by EU leaders of a security approach concerned with pursuing internal security objectives through an external security policy. Thus the European Council, commenting on the 2015 action plan, stated that 'the problems of the Sahel not only affect local populations but increasingly impact directly on the interests and security of European citizens ... thus, bringing security and development to the Sahel leads to a strengthening of the EU's own internal security'. 37 This interest-driven security approach of the EU was reinforced in its most recent 'grand strategy', the Global Security Strategy of 2016. 38 Since 2016, management of migration from the region has also been increasingly supported through the regionalization of the CSDP missions, with a particular emphasis on support for the G5 Sahel Joint Force. The G5 Sahel initiative, established in 2014 with the essential support of France and the EU, was originally presented as a regional framework within which Burkina Faso, Chad, Mali, 33  Mauritania and Niger could pursue security and development, but it soon became heavily focused on security concerns. 39 Thus, with the regional focus expressed in their renewed mandates, the CSDP missions adapted to the EU's focus on emphasizing the links between territorial security and the European continent. 40 Until then, the CSDP had been explicitly framed as a humanitarian-driven conflict management framework to be used outside Europe. 41 Meanwhile, the EU has since the 1990s been attempting to restrict migration from west Africa to Europe by criminalizing migration and externalizing border control through a number of different policy frameworks. 42 Specifically, in response to the European refugee crisis, the 2015 Valletta migration summit established the Emergency Trust Fund for Africa (EUTF), managed directly by the European Commission, to address 'root causes' of migration and displacement in Africa. 43 The fund spent €4.6 billion within a span of three years. 44 In the words of EU staff, money is 'floating to Mali', which receives €214 million annually in Trust Fund money alone. 45 Thus, with the 2015 crisis, the traditional logic of the EU's external approach merged with the inward-looking approach of existing migration policies and border controls-framed as working towards a safe inside that can be protected from an unsafe outside. 46 As outlined here, the EU's activities in the Sahel region can be seen as part of a bigger transformation of EU actorness whereby the external role of the EU and its CSDP is moving away from its original intent to play a soft, value-based and antipower role in international affairs and towards a more narrow focus on EU internal security. 47 This is not to say that EU actorness is now entirely promoting a defensive and inward-looking agenda. Numerous references to value-based, altruistic concerns are to be found in key strategy documents relating to the Sahel. The same holds for the CSDP mission mandates, which continue to be non-executive and concerned with promoting long-term reform. In addition, conflicting agendas exist, even at the level of policy-making, as both CSDP and EU Commission 39 The G5 Joint Force has 5,000 troops operating in the region. 40  the field, they cannot monitor the effects of training on soldiers' behaviour and routines. 55 Furthermore, the risk aversion of European contingents is particularly visible in the geographical organization of troops in Mali. The EU security staff live inside securitized bubbles, with the risk being transferred to their Malian partners. 56 EU staff operate only in defined spaces, with strict limitations on how far they can move beyond their designated areas of operation: We are trying to be present and control the situation-but it happens at a distance. It is difficult to go into town. To do things with the population is impossible. It is not possible for us to go out there and provide security when we are only allowed to stay inside HQ. 57 EUTM's headquarters is based in a former hotel, now hidden behind a massive gabion barrier following an attack on the facility in 2016. EUTM military officers are under strict rules on leaving the premises, and many rarely do so. Local trainees are brought to the EUTM training camp in Koulikoro, 60 kilometres north-east of Bamako, despite the fact that the Malian army found the location impractical, given its distance from the zones of active operations. 58 The EU chose Koulikoro mainly for security reasons, reflecting the political unwillingness to operate outside Bamako's perceived safe zones. The civilian counterpart to EUTM, EUCAP, operates a training site in the centre of Bamako to which trainees are brought from all over the country, also creating physical distance from their daily operations.
The non-executive mandates also co-produce these EU-designated perceived safe zones where Malians must come to receive training. The camp in Koulikoro came under attack just a few weeks prior to our visit; what happened on that occasion illustrates how EU translators perceive themselves, their intervention practices and their safe spaces. The attack involved two vehicles packed with explosives, and a group of armed men who opened fire as they sought to breach the main gate of the camp. One of the two vehicles detonated and the other caught fire after crossing the main gate. When we visited the site the remains of the burnt-out car gave us a vivid image of the recent event. Three Malian soldiers and one civilian were injured in the attack; however, as all the EU soldiers we talked to emphasized, none of the EUTM military officers were injured. To them, their response to the attack was solid proof of a well-designed security space: 'It was a success for our mission-a failure for the terrorist,' 59 an EUTM officer explained to us.
The spatialized everyday security practices of the EU translators thus depend on demarcated safe zones distinguished from the surrounding 'danger zones'. Reflecting the non-coercive and distant approach to interventions pursued by the EU, its staff work from safer positions than the Malian army, which also leads to a certain disruption of perceptions of danger.  60 One EU officer explained his feeling that in Mali, threats were more difficult to assess than in Afghanistan: In Afghanistan, we could sense when there was going to be an attack. Women and children disappeared from the streets and people would stay inside their houses. In Mali it is different. You never know when and from where an attack will hit. But that makes it even more stressful, because you never know who to trust and where the enemy is hiding. Therefore, we take our own measures. We have to treat all situations outside the camps as potentially dangerous. 61 The attackers in Koulikoro evaded checkpoints that the EU had installed just outside the camp in cooperation with the Malian Ministry of Interior, whose perceived carelessness the EU military staff indirectly blamed for the attack. Referring to the weakness of the Koulikoro checkpoint, one soldier explained: We can see there are flaws in their [the Malians'] security measures, like vulnerable points, where the enemy can attack from. But they don't care. And we cannot push them too hard. They have a different perception of work. If we push them too hard, they will refuse to collaborate. 62 Indeed, the fact that the EU officers did not get hit confirmed their self-perception as superior to their partners.
The non-coercive character of the EU CSDP activities is an important aspect of the EU's self-identification as an anti-power type of security actor in interventions, which is emphasized in the mission mandates. In practice, however, translating the EU as a benign security provider in a hostile, asymmetric conflict context is extremely challenging for the individual officers. On the one hand, the distant approach that the non-coercive mission mandate forces the EU to take implies that EU staff are highly dependent on their relationship with local partners. At the same time, as we shall see below, the EU's approach ends up undermining the relationship with those partners by being unable to deliver what is needed.
In sum, the EU's intervention practices in Mali do not simply follow an already established profile as security intervener. The EU as an intervention actor is constituted during and through ongoing intervention practices that contribute to the constitution of certain desired, yet fragile and temporarily constituted, self-understandings, which in turn have implications for intervention actorness and practices. Meanwhile, as we will now go on to explore, intervening actors' desired self-understandings are indeed relational and depend on the recognition of other actors in the same intervention space. Getting partner buy-in and support for a self-interested security agenda may prove a major challenge for translating the inward-looking turn-and, consequently, for the legitimacy of the EU as a security partner.

Diverging perspectives on long-term security sector reform
One policy objective that requires a great deal of translation relates to supporting structural security sector reform, not least because these objectives often differ from the interests of the Malian partners. As noted above, the EU has identified longer-term security sector reform, notably through the rehabilitation of Mali's security forces, as one of the main policy objectives guiding its CSDP missions. The overall goal of EUTM Mali is to rebuild the military capacity of the Malian armed forces, while its civilian counterpart, EUCAP, is tasked with assisting and advising Mali's internal security forces in implementing the government's security sector reform. However, in negotiations between Mali and the EU, the quest for long-term reform is a point of contestation. 63 The Malian authorities resist EU support for structural reforms, which they see as interfering with their national affairs. Many EU staff interpret the Malian stance as both a way of maintaining a corrupt patron-client system and an exercise of sovereignty. Malian state actors consider the security sector a core pillar of state sovereignty, perceiving external EU scrutiny of it as 'invasive and paternalistic'. 64 As one Malian officer put it: 'Sometimes, the European partners can be very patronizing. The leaders act as if they are in command. Our commanders don't like that. Sometimes, they refuse to shake hands with the French commanders.' 65 Malian resistance to EU reform objectives is evident in several contexts within the defence sector. EUTM has encountered significant obstacles in its efforts to overhaul the logistics and human resources management systems of the Malian army (FAMA). For example, EUTM has tried to introduce a modern electronic data-processing system into FAMA's human resources management, covering such aspects as the payroll, and monitoring the actual number of soldiers enrolled across different companies. But the Ministry of Defence has essentially resisted such efforts. 66 Resistance also emanates from the country's military schools, with which EUTM has difficulties in communicating. EUTM depends on the schools to send soldiers to be trained but is provided with only scant advance information about the participants, which makes training difficult. 67 Furthermore, while EUTM aims not only to train soldiers, but to build the capacity of entire companies, the Malian military schools prefer to send individual soldiers from different companies. The EU trainers complain that this hampers the long-term structural effects of training; they interpret these actions by the Malian authorities as a sign of resistance to external interference in matters which are 'close to the sovereign heart'. 68 The EUCAP mission encounters similar challenges, as the national internal security forces tend to select individual police officers from various units, instead of whole companies, to be trained. For their part, the Malian partners International Affairs 96: 4, 2020 often feel that the EU projects are not adapted to the realities on the ground. One Malian police officer explained: Most often, the partners come with their predefined projects. But unfortunately, the projects don't take into consideration the reality on the ground. Here, a military expression says: 'It is always the field that commands.' Now, most police posts in the central and northern part of the country are closed because they are being attacked. The security situation is deteriorating. Still, the partners come with their construction and capacity building projects, but they are not adapted to the needs on the ground. 69 This resistance at different organizational levels reflects how, having being treated as a so-called 'donor darling' by the international community for decades, 70 the Malian authorities have carefully developed a negotiation style where leverage and leeway are achieved not during initial negotiations but in the course of implementation. Evidently, the analysis of the implementation level is crucial for the study of the derived effects of these negotiations.
Instead of engaging in long-term structural reform, the Malian authorities prefer EU training and assistance activities that focus on their urgent priorities and either can be quickly transformed into usable means in combat or can provide medical aid and other technical assistance. Interviews with numerous EU staff working in the CSDP missions confirmed that most of their activities are currently centred on shorter-term technical assistance training that can be implemented quickly by the security forces. Witnessing one of the most popular training sessions held in the EUCAP training camp, we watched police staff being trained to dismantle improvised explosive devices (IEDs)-which the trainees explained they increasingly encounter. In this case EU staff, attempting to adapt to the situation on the ground, ended up promoting a different kind of security from that anticipated in the reform-focused mandate of the CSDP. This example also illustrates that interventions are the outcome of a negotiated compromise between Mali and the EU in a relationship that has been described as dissonant and tense. 71 EU staff are involved in a process of bending and shaping existing programmes in order to speak to and be accepted by their Malian counterparts, who are struggling with an increasingly precarious security situation: in just the last three months of 2019, more than 100 Malian soldiers were killed by jihadist actors.
Meanwhile, the non-executive character of the EU mission is also problematic for the relationship with the Malian partners: 'It is very difficult to meet the expectations of the partners because of the way our mandate is formed.' 72 Contention emerges when EU missions provide equipment only for training sessions, and cannot allow trainees to take the equipment into the battlefield. 'The Malians look at our equipment and they see us as security forces. But we don't have arms to give them. They come to sniper training, for instance, but they have no rifles to 69  shoot with. It makes no sense.' 73 This reflects a broader trend in the relationship between intervention actors and regional security partners. Sahelian state actors use the security tools and technologies provided by international capacity-builders to enhance and consolidate their authority over rival institutions and to increase personal wealth and social status. 74 Security tools and technologies, therefore, mediate relationships between interveners and intervened-upon groups within the space of intervention. 75 In this case the EU fails to meet the expectations of the Malian partners, who as a result are unable to demonstrate their authority and strengthen their social position vis-à-vis other actors in the field.
As Mosse reminds us, so-called intervention 'success' is constructed and depends on the ability to get buy-in and support from project beneficiaries and partners on the ground. 76 The interveners' self-understanding and 'project success' is a relational construct between 'beneficiaries'-or the 'intervened-upon'-and the 'interveners'. Seen through this lens, EU actorness is a relational effect of interventions; that is, it is constituted through relationships with and between other actors in the field. In this case, the EU translators struggle to find common ground with their Malian partners, through which they can construct themselves as successful intervention actors. However, as we shall see in the next example, through the EUTF the EU is providing new security measures that may benefit its partners, albeit in ways that constitute new, precarious, security effects on the ground.

EU Trust Fund intended for migration management used for security purposes
As noted above, before the security collapse in Mali in 2012, EU engagement in Mali and the broader region consisted essentially of development cooperation aid. 77 While the development engagements continue, the European Commission is increasingly allocating development aid budgets to security purposes. Following the 2015 crisis, the inward-looking approach of existing migration policies and border control measures was strengthened in particular with the establishment of the EUTF.
One of the regional Trust Fund programmes is the Groupes d'Action Rapides-Surveillance et Interventions au Sahel (GARSI Sahel), 78 in Burkina Faso, Chad, Mali, Mauritania, Niger and Senegal, launched in 2017 with a budget of €41 million. In Mali, GARSI is intended to help local authorities implement effective 73  territorial control through the deployment of elite gendarmerie forces specialized in countering terrorism and illegal migration. The blurring between civilian and military measures in part reflects the institutional set-up in Mali, where the gendarmerie falls under the Ministry of Interior and therefore by definition is a civilian force. This allows for counterterrorist mechanisms to be 'disguised' as civilian support. The EU staff we spoke to had noticed that the EU development support has become increasingly securitized: It feels a bit schizophrenic. Development funds are being used to arm a civilian mission, but it is not the army. We used to buy rice for the poor people, now we are buying them arms. We have a new agenda, but it is the same [development] tools that are being twisted to reach new targets. But that is not really possible. 79 As an EU development officer stated: GARSI is a difficult mission. The Malians do not acknowledge that they have an issue with illegal trafficking. They do accept development projects that benefit the Malians. But they do not want to take any concrete measures to limit irregular migration. Mali is a proud nation; they are not happy to be patronized.
The disposition of the Malian state should also be read within the context of how, historically and currently, European anti-migration initiatives such as readmission agreements have been subject to intense contestation and popular protest. It is therefore a sensitive political subject and certainly has an influence on relations between the EU and Mali. 80 Thus, the rejection of the anti-migration label by Malian authorities in favour of the further-reaching border control and anti-terror concept also illuminates how power is being exercised in the partnership between Malian government officials and the EU. Delivering on migration in the current security context is particularly difficult. An EU staff member explained: 'In Mali the government plays a double game internally. They tell us what we want to hear to get access to the EU Trust Fund for cutting down on illegal migration.' Meanwhile, the Malian government does not want to further upset civil society actors who are increasingly expressing their distrust of the government and their distaste for what they claim is a neo-patrimonial relationship between Mali's government on the one hand and international donors and security actors on the other. 81 At the same time, the EU staff allow the Malians to play this internal game; staff in GARSI and the EU Trust Fund-financed 'programme of support for enhanced security in the Mopti and Gao regions and for the management of border areas' (PARSEC Mopti-Gao) do not actually think they are doing anything related to migration; they just have to indicate on paper that it is related to this issue in order to fit within the EUTF framework. 82 The example illustrates how funding, earmarked according to donors' shifting policy objectives, is both appropriated by local partners and used by the EU trans- 79 Interview, EU staff member, Bamako, Feb. 2019. 80 Interview, EU staff member, Brussels, Oct. 2018. 81  lators to 'tick the boxes' in terms of delivering on the migration agenda. Furthermore, with increasing numbers of internally displaced persons (seeking shelter from violence elsewhere in the country) on the outskirts of Bamako, curbing out-migration is a not a priority for the Malian state. According to EU staff, it is in fact almost impossible to get the Malians to deliver anything on the migration agenda. Though programmes such as GARSI are supposed to look at irregular migration flows, irregular migrants are difficult to track, as there is freedom of movement within the Economic Community of West African States. Hence, a project labelled as migration management has more to do with countering crossborder security threats through the provision of infrastructure and technology to counterterrorist police units. 83 As such, Malian state actors benefit from raising the profile of terrorism and dangerous migrants to attract funding and investment in the security sector. Indeed, it becomes apparent that militarization, understood as use of force, finds its way in during the course of intervention, 84 in this case disguised as an emergency crisis response.
A similar trend is reflected in PARSEC Mopti-Gao, which aims to improve border management and prevent irregular migration and human trafficking. 85 However, according to EU staff, €21 million of the project's total budget of €29 million was spent on security infrastructure such as boats, planes and border posts. 'We use the projects to build internal infrastructure for the security forces like, for instance, accommodation, communications rooms, equipment, vehicles, and communications apparatus.' 86 EU staff have nevertheless increasingly come to realize that, in fact, Mali has no control over its borders. The EU has funded monitoring centres in the desert where nobody crosses. Meanwhile, the violence the population encounters in these areas has proved to be beyond their reach.
The translation at play in these three examples is significant for several reasons: first, because it is representative of how the EU as a security actor struggles to promote a new inward-looking agenda; second, because it shows how the securitization process channels unprecedented amounts of resources through grant proposals and to meet specific objectives; third, because it shows how the issue of migration is treated distinctively in the EU's Trust Fund mechanism, which is embedded in the Union's new inward-looking turn in its foreign policy. Moreover, the nexus between internal and external security is shaped by staff members who may agree or disagree with the new priorities, and simultaneously meets resistance from Malian partners. The question is, then: to what extent are effects produced in the attempt to translate the inward-looking turn into practice? In the encounter between security translators and brokers on the ground, the framing of migrants as a security threat blends neatly with the framing of terrorism as a transnational 83  phenomenon thriving in 'ungoverned borderlands'. As such, the ability of securitization to extend into new social fields, in this case development and migration, allows the EU and its Malian counterparts to find new common ground-albeit from very different power positions-as the EU's inward-looking approach translates into very concrete, available security measures delivered by the EUTF.

Conclusion: how implementation practices shape EU actorness
Drawing on the Paris School's conceptualization of security practices as performative acts, this article has explored how EU actorness is shaped by the daily practices of EU staff in conflict-ridden Mali. So far, existing debates have analysed EU actorness in the field of security either as a discursive struggle for self-identification or as the (in)ability of the EU to meet predefined policy objectives. Few have analysed EU actors' performative role at the level of implementation, and how these intervention practices, in turn, shape EU actorness. In attempting to fill this gap, this article has constructed an analytical framework within which to trace both the level of discourse and the level of implementation through the concept of translation, which denotes the ways in which ideas of security are adapted, deflected and modified as they pass through the different layers of an organization. As Latour emphasizes, all actors produce their own interpretations, which then become performative. We cannot, therefore, fully grasp EU actorness by examining only the level of policy choice. We also need to direct the analytical gaze towards the site of implementation, exploring how EU staff enact the Union's security role as put forward in its official policy objectives. In order to enact such policy objectives, translation by EU staff is needed. Our three examples illustrate how, through processes of translation, EU actors in Mali adapt, deflect and modify the EU's shifting policy objectives.
First, the non-coercive character of activities under the CSDP is an important aspect of the EU's self-identification as an anti-power security actor in interventions, which is emphasized in the mandates of the two CSDP missions. Caught between restricted mandates and political imperatives, security translators confirm their desired self-understanding as superior by supporting Malian partners from a distance. In practice, however, translating the EU as a benign security provider in a hostile, asymmetric conflict context is extremely challenging for the individual staff members, because the character of the missions leads to loss of control when they are forced to act at a distance from their local partners. Moreover, the distant approach that the non-coercive mission mandate forces upon the EU implies that EU staff are highly dependent on the relationship with their local partners-a relationship that has nevertheless become increasingly tense. This leads to fragmentation, for intervening actors' desired self-understanding is indeed relational: it depends on recognition by their partners in the same intervention space.
Second, a combination of the tense negotiations between Malian and EU staff, with Malian partners resisting cooperation on the structural reform agenda, and a new focus at the EU policy-making level in Brussels, has led to suspension of the International Affairs 96: 4, 2020 latter's long-term goal of public sector, especially security sector, reform. Malian state actors, upon whom the EU increasingly relies to carry out its activities, often feel the EU projects are ill-suited to their context. Meanwhile, they try to pursue their own agendas and use the support they gain from their partnerships with the EU to bolster their strategic position vis-à-vis their European counterparts. Thus the EU's policy objective of externalizing border control in pursuit of an inwardlooking security agenda ignores the pushback from local partners. Furthermore, as the security situation deteriorates, Malian demands for equipment and support increase. Meanwhile, the EU's approach of managing threats from strictly demarcated safe zones, and its inability to deliver the security (and tools) that are needed, only emphasizes its lack of capability to act. In turn, the EU's actorness and ability to perform are hampered by the lack of buy-in from their local partners, as narrating 'success' in a context of escalating violence becomes increasingly implausible. Meanwhile, counter-effects on the ground produce disorder at the very borders that the EU wishes to protect.
Third, in the negotiated process of translating security, EU development and anti-migration initiatives are being repurposed for security objectives. Through the translation process, EU development funds are channelled into anti-terrorist police units framed as enhanced border control, as the result of manoeuvring between pressures from the EU Commission's offices in Brussels to create results on the anti-migration agenda and the priorities of the Malian state. Meanwhile, the boundaries defining who is to be secured, and who is defined as a threat to security, become increasingly blurred. This is a framing that, nevertheless, serves both the interests of the Malian state and those of the EU security providers on the ground.
To summarize, the EU's intervention practices in Mali do not simply follow from the Union's established profile as a security intervener. The EU's actorness is constituted during and through ongoing intervention practices. Intervention practices contribute to the constitution of certain desired, yet fragile and temporarily constituted, self-understandings that in turn have implications for intervention actorness and practices.
The EU's actorness depends, first, on the ostensive act of identifying an external crisis in ways that allow it to intervene and thereby confirm its desired self-understanding as a capable intervention actor. For example, the political labelling of 'illegal migrants' and 'terrorists' in unregulated coexistence in the same 'ungoverned' Saharan space as a European (national) security threat is deemed to necessitate EU intervention in the Sahel. While these political labels are constructed and reproduced in the first instance to legitimize and confirm intervention practices, the secondary effects of managing these perceived threats (as an ongoing intervention practice) reflect back upon the actorness of the EU. Furthermore, the growing popular discontent with both the Malian state and the international interveners creates a disjuncture between the EU's increasingly grandiose vision of security projects and its inability to perform security on the ground. Thus, while the EU's activities in Mali reinforce the idea of the EU as a security actor, the limited