On Danger, Sacrifice and Professionalisation: MSF and the security debate

Michaël NeumanTranslated from French by Nina Friedman.

Since its inception, Médecins Sans Frontières (MSF) has confirmed its commitment to working in war zones. Its staff have faced danger, to a greater or lesser degree, throughout the organisation’s history. In this chapter we examine how this issue of risk has manifested itself, from the founding of the associationThis analysis concerns only MSF’s French section.
 to the beginning of the 2010s, notably in discussions during Board of Directors meetings and annual reports presented to the General Assembly. We will see how the debates and deliberations on how best to protect ourselves from danger have been influenced by the growth of MSF, changes in the political context and the advent of “humanitarian security” within the aid system.

The Early Years: The 1970s and 1980s

Romanticisation of Danger and Rejection of Sacrifice

With most of its founding members profoundly marked by their experiences in the 1960s with the Red Cross in Yemen and Biafra (Nigeria), MSF leaderswere well aware from the outset of the potential dangers. The confronting of danger in the early years was staged, corresponding to an “aristocracy of risk.”Bernard Kouchner, Le malheur des autres, Paris: Odile Jacob, 1992 (1st ed. 1991), p. 322 [translated].
 This romantic view was reflected in the association’s original charter: “Anonymous and volunteers, [its members] seek no individual or collective satisfaction from their activity. They understand the risks and dangers of the missions they carry out.”For an account of the origins of the engagement of MSF’s founders, see Eleanor Davey, Idealism Beyond Borders. The French Revolutionary Left and the Rise of Humanitarianism, 1954-1988, Cambridge: Cambridge University Press, 2015.

Referring to the association’s Lebanon mission in his 1977 President’s Report, Bernard Kouchner paid “special homage to the fifty-six volunteers, men and women, nurses and doctors, surgeons and anaesthetists who, on behalf of Médecins Sans Frontières (MSF), left the tranquillity of France as volunteers to face fear and danger in the name of the brotherhood of humankind and medical action.”Médecins Sans Frontières, President’s Report, 1977 General Assembly [translated].

Despite the departure from the organisation in 1979 of many of MSF’s founders, Kouchner included, this bravado, in the form of a “heroisation” of the narrative, lived on—but with a rejection of sacrifice. The assertion that “We know there will be a price to pay, because nothing big is ever achieved, nothing gets created or accomplished without risk”Médecins Sans Frontières, President’s Report, 1980 General Assembly [translated].was immediately followed by a call for prudence and to heed the advice given by exploratory missions. In 1981, the president agreed: “We are not asked to be heroes, we are asked to do our job, as well as possible, as sincerely as possible, but, above all, to come home.”Médecins Sans Frontières, President’s Report, 1981 General Assembly [translated].

The 1980s were a decade of very rapid expansion for MSF, both for the French section—whose revenues rose from 7.3 to 207 million francs between 1979 and 1989—and internationally, with the creation of the Belgian (1980), Swiss (1981), Dutch (1984) and Spanish (1986) sections. In 1983-4, MSF- France sent 600 people on mission, almost four times as many as in the mid-1970s.

Moreover, by the end of the 1970s, the association began operating under its own flag in refugee camps, and then, where possible, across borders in countries in conflict: Afghanistan, Honduras, El Salvador, Chad, Sudan, Eritrea and Uganda. Its exposure to risk increased significantly.

In this high-risk environment, the association suffered its first deaths, either accidental or from disease. The first combat-related incidents came in 1980. In Chad, a lone sniper targeted a team of three doctors and wounded one. The same year, in Zimbabwe, a car was machine-gunned, but no one was hurt. In Uganda, where security was non-existent, “a team came close to being massacred,” but “was only looted.”Médecins Sans Frontières, Collegial Steering Committee, 18 July 1981 [translated].
 There were also arrests and detentions, some of them lasting months, as in Turkey in 1981.Médecins Sans Frontières, President’s Report, 1981 General Assembly [translated].
 Each incident was considered in isolation, with no wider perspective. MSF was not looking for trends.

In the years that followed, the number of incidents grew, yet there was still no structured response from the organisation. What we now call “security incidents” included teams being caught in aerial bombing raids in Tigray, Ethiopia (1983), a plane carrying volunteers to Mozambique being fired upon (1985) and kidnappings in Chad (Belgian section volunteers in 1984) and Somalia (1987), along with all of the dangers of the Afghanistan mission (bombs, attacks on convoys, etc.). Missions were suspended or evacuated in Sudan, Afghanistan and Mozambique.

The early days of MSF’s mission in Uganda, then in the grips of the 1980 famine, illustrate to some extent the improvised and localised approach to security in an extremely dangerous setting, but of which MSF attempted no political analysis. The teams had to contend with criminal acts, primarily committed by livestock traffickers, as well as the unpredictability of soldiers at roadblocks. Their exposure was greatest on the roads and MSF vehicles were sometimes targeted. “We kept our fingers crossed when we came across highway bandits or Tanzanian soldiers. We took the road early in the morning, otherwise soldiers who drank too much would become aggressive and dangerous, and in the towns we drove fast to avoid getting ambushed,” recalls Rony Brauman, who was in charge of the mission at the time.Interview with Rony Brauman, 6 October 2014.

In 1987, with a growing number of humanitarian workers (in Somalia, for example) and journalists (especially in Lebanon) being taken hostage, MSF began wondering whether “a new trend”Médecins Sans Frontières, President’s Report, 1987 General Assembly [translated].
 was emerging. But the president’s answer was “no”, and there was no change in the narrative on risk-taking.

But while the term “risk management” was almost never used in discussions, a certain consistency in security practices might be remarked upon, characterised by a fledgling professionalism, a high degree of delegation to armed groups, the use of public condemnation and, as a last resort, withdrawal.

Professionalisation, Delegation, Condemnation and Withdrawal

The professionalisation process initiated in the early 1980s stemmed from the will to create a structured and effective organisation: raising funds, developing media contacts, setting up a uniform data collection system and drug lists that were “as consistent and standardised as possible,” generalising the use of radio and telex, and using planes to travel when necessary.Médecins Sans Frontières, President’s Report, 1984 General Assembly [translated].
 The position of “coordinator” was created—a role which required both resourcefulness and diplomacy. The association’s dread of bureaucratisation was countered with arguments about security and the quality of communications: “We have to stop leaving numerous or widely scattered teams out of contact with France in these dangerous countries.”Médecins Sans Frontières, President’s Report, 1982 General Assembly [translated].
 At the time MSF consisted of a very small head office, overseeing missions with which, for lack of resources and technology, it had very limited contact. Information from the field was scarce because it could take days to get to a telephone and letters could take weeks to reach their recipient. In reality, MSF delegated much of its security (and logistics) management to the belligerents, in the belief that they should do their share of the relief work. In Eritrea and Afghanistan, for example, the teams crossed the borders in guerrilla convoys, in a bid to stay safe. This practice remained in use in Angola until the mid-1990s. Such alliances of convenience were not without their difficulties. Logistics did not always follow, communications were erratic, the armed groups could make excessive financial demands and sick volunteers sometimes received poor care. Yet such problems did not cast doubt on either the modus operandi or its legitimacy; this was simply how things were done.

In spite of its rapid growth, MSF was still small and relatively unknown outside France. Meetings with political and military groups in countries where MSF wanted to work were less a time for negotiations than an opportunity to make itself known. MSF’s leaders counted mainly on the mobilisation of the public to increase the organisation’s influence and extricate itself from dangerous situations. One example was the campaign that publicly condemned the pro-Soviet Afghan government over the detention of Philippe Augoyard, a doctor working with AMI (Aide Médicale Internationale), captured on 16 January 1983 in Logar Province. Another could be found in the denunciation of the Soviet Army’s bombing of hospitals run by foreign teams.Four MSF and two AMI hospitals were hit between 1981 and March 1982.
 In 1983-84, together with AMI, MDM (Médecins du Monde) and the FIDH (International Federation for Human Rights), MSF planned to draw up a charter on the protection of medical teams. Ultimately the project was dropped on the grounds that it would have entailed MSF systematically officialising all its activities, which would have been at odds with the organisation’s practices and its aspirations to maintain its unofficial status in countries such as Pakistan (used as a rear base for the Afghan mission), Ethiopia’s Tigray and Eritrea regions.

The ultimate response to risk was to withdraw. Programmes were suspended in Uganda in 1981 because of safety incidents and again in 1982 in Iran due to widespread insecurity and problems with obtaining access to the population. The decisions to withdraw were made at head office, and in some cases, for example Uganda in July 1981, against the wishes of the field teams.

The Turn of the 1990s: Formalising the Rules, the Drive for Professionalisation and Tensions over Practices

Growth and the End of the Cold War

MSF’s French section continued to grow. The number of international volunteer posts increased from 275 in 1990 to 426 in 2000. Head office also expanded, from fifty employees to some 150 during the same period. By the year 2000, MSF had developed into an international organisation with sections in nineteen countries and an increasingly well-established reputation. But, like the rest of the world, it was facing the geopolitical changes brought on by the end of the Cold War.

Withdrawal of Soviet forces from Afghanistan was the first sign of a new era for MSF; in September 1989 it was noted that “the resistance is breaking up.”Médecins Sans Frontières, Board of Directors, 15 September 1989 [translated].
 The Mujahideen made it clear to MSF that things had changed: “They no longer feel that the benefit of having humanitarian teams there is worth the price of effectively protecting those teams [...]. The situation is becoming more and more complex; the increasingly acute security problems are difficult to even think about.”

As their alliances with “freedom fighters” crumbled, MSF and other humanitarian organisations saw new spaces opening up. In post-Cold War conflicts, as in the wars breaking out in Somalia, Liberia, the former Yugoslavia and the Great Lakes region, it was now possible to work on both sides of the frontline. Under such conditions, the practice of “embedding”—considered a stopgap measure, despite its romantic aspects—was gradually becoming obsolete. In a context marked by a horrific succession of mass crimes, MSF was forced to become more self-reliant in terms of security.

MSF and its First Casualties of War

The increasing number of security incidents made 1988 the “year of living dangerously.”Médecins Sans Frontières, President’s Report, 1989 General Assembly
 And according to President Rony Brauman, it was “only a matter of luck” that no one had been killed. He seemed to anticipate the worst, however, and, by the next General Assembly, he was lamenting the deaths of two volunteers killed when their plane was shot down over southern Sudan in December 1989, and of another who was killed in Afghanistan in April 1990. They were MSF’s first casualties of war. MSF-France ceased operations in both countries.

The conflict in Somalia—where, along with Iraqi Kurdistan, the first international military interventions in the name of the protection of humanitarian assistance were launched—ushered in a decade of mass violence and UN interventionism. Countless incidents were reported during meetings of the Board of Directors. Here are just a few examples, to illustrate their variety and impact: “Over the past three months, seven people have been wounded while on mission: three were hit by machine-gun fire by a lone gunman in Mogadishu and four were caught by fire from a helicopter and two light bombers in Sri Lanka,” noted the June 1991 President’s Report. In October 1991, as a convoy of wounded was being evacuated in Vukovar, Croatia, an MSF vehicle hit an anti-tank mine, which undoubtedly had been planted deliberately. Four people were wounded, one of them seriously. In Liberia, in addition to all sorts of violent incidents, MSF was plagued by large-scale looting, as were all the other aid agencies.

What set the crises in West Africa and the Great Lakes region apart was not only the extreme violence against civilians, as witnessed directly by the teams in Rwanda, Burundi and Zaire from 1993 to 1997, but also the gravity of the security issues. The May 1994 Board of Directors meeting, for example, reported the evacuation of an MSF-Belgium team from Butare, “as the hospital had been emptied of its patients, killed by militiamen, FAR [Armed Forces of Rwanda] and the Presidential Guard,”Médecins Sans Frontières, Board of Directors, 20 May 1994 [translated].
 and some of the Rwandan staff executed. Although there are no precise figures, more than 200 MSF employees were estimated to have been killed between April and June 1994 during the genocide in Rwanda.
The conflict in Chechnya in the Russian Caucasus also caused its share of incidents—notably kidnappings for ransom—and four international staff members of MSF-Belgium and MSF-France were abducted in 1996-97. In June 1997, a Portuguese doctor was assassinated in Baidoa, Somalia. He was the first international staff member to have died from an act of violence since 1990.

Humanitarian Security Concerns Contribute to Structuring the Aid System

As Mark Duffield commented, the increase in MSF’s war zone operations was part of “an unprecedented aid industry expansion at every level: geographical reach, funding availability, the agencies involved and the range and complexity of their responsibilities.”Mark Duffield, “Challenging Environments: Danger, Resilience and the Aid Industry,” in Security Dialogue, vol. 43, no. 5, 2012, pp. 475-492.
 The deployment of aid workers into the heart of conflict zones considerably increased their exposure. In May 1992, an International Committee of the Red Cross (ICRC) delegate was killed in an explosion in Bosnia, some twenty international workers met their deaths in Burundi during the period 1995-97 and, in December 1996, six ICRC delegates were murdered in cold blood near Grozny in Chechnya. Humanitarian aid observers began to stop seeing these events as “isolated incidents” but as interconnected. A narrative emerged which acknowledged the danger faced by aid workers in the context of the changing nature of conflicts“Security of relief workers and humanitarian space”, Commission working docu¬ment, EU Commission, May 1998.
 and the declining status of humanitarian personnel, who were increasingly perceived as being ineffective. Deprived of the ideological and strategic framework of East-West confrontation, wars were now, so the thinking went, driven solely by ethnic or religious resentment and economic predation. Adopting the “new wars” discourse popularised by Mary Kaldor and Paul Collier,On this discourse and criticisms of it, see Roland Marchal and Christine Messiant, “Les guerres civiles à lère de la globalisation”, Critique Internationale, no. 18 (2003/1), pp. 91-112.
 many humanitarian actors believed that what characterised these new conflicts was that their primary targets were civilians and those coming to their aid.

Such was the context in which the safety of humanitarian workers made it onto the agenda of international institutions. Take, for example, paragraph 65 of the final communiqué from the June 1997 G8 Summit in Denver, which expressed “grave concern at the recent attacks against refugees as well as against personnel of refugee and humanitarian organizations,”Final Communiqué ofthe Denver Summit ofthe Eight, 22 June 1997, www.library. utoronto.ca/g7/summit/1997denver/g8final.htm, last accessed 22 December 2015.
 and UN General Assembly Resolution 52/167 on the safety and security of humanitarian personnel adopted in 1997.

The increasing number of security incidents led to a veritable paradigm shift within the aid system as the European Community became a prescriber of security management practices. A European Commission discussion paper on the security of humanitarian workers recommended that institutional donors require partners to demonstrate their ability to assess situations, track and investigate security incidents, establish security guidelines and commit to training and briefing their staff.EU Commission, op. cit.

Among those promoting “next generation” security management was Koenraad Van Brabant, an anthropologist by training and a researcher at London’s Overseas Development Institute (ODI). Van Brabant greatly influenced the professionalisation of security processes. As he pointed out at the time (and in this he was only the harbinger of a growing trend), “...as recent events in Rwanda, Chechnya and elsewhere demonstrate, there is a real need for agencies to invest in acquiring the appropriate security skills.”Koenraad Van Brabant, “Security Guidelines: no guarantee for improved security”, Humanitarian Practice Network, London: Overseas Development Institute, February 1997.
 Because, he believed, “assessing risk and determining risk reduction behaviours is a skill that few staff may have, particularly those without professional military training.”Ibid.

All but unanimous in their new geopolitical reading of the period and in seeing the necessity for a fundamental re-think of humanitarian security,Franck Schmidt, “Recommendations for improving the security of humanitar¬ian workers,” International Review of the Red Cross, vol. 37, no. 317 (1997), pp. 152-155.
 humanitarian organisations began recruiting their first security managers. In 1994, the ICRC set up its “security and stress management unit”,The integration of stress was anything but anecdotal. It showed the importance that the psychological health of staff having to contend with violence was begin¬ning to have in the humanitarian sector.
 and the number of humanitarian security initiatives implemented by the aid community soared.InterAction’s Security Advisory Group (1991), Inter-Agency Security Management Network (1994).
 The humanitarian security market was booming, fuelled by former military personnel returning to the private sector with the post-Cold War downsizing of Western armed forces.See, in particular, Larissa Fast, Aid in Danger. The Perils and Promise of Humanitarianism, Philadelphia: University of Pennsylvania Press, 2014.

Formalising the “Golden Rules” and Difficulties Complying With Them

The annual report presented at the 1990 MSF General Assembly included for the first time a section devoted to “security problems”, “brought to the forefront of our concerns.” It called for a number of measures: smaller missions (to limit exposure); curative mission objectives (inasmuch as exposure must be gauged against a project’s medical benefits); more detailed and systematic briefings and regular visits by board members. A sense of the association’s collective responsibility began to emerge, as evidenced by the many debates and controversies over security issues in Somalia alone in 1991.
Following attacks in southern Sudan and Afghanistan, in 1990 the operations department established a set of “golden rules”. See, in particular, Gérald Massis (ed.), Manuel de l’administrateur/logisticien, Médecins Sans Frontières, 1990. Rather than technical recommendations, these were general principles reiterating part of the framework laid out by the president that same year: understand the context, communicate MSF’s work, prioritise curative activities in dangerous environments and “never count on humanitarian immunity”. Furthermore, they explicitly reaffirmed head office’s authority over the field regarding the decision to withdraw—something that was gradually to take on more meaning as advances in communication technologies afforded MSF managers in Paris increasingly regular contact with their missions. The year 1992 saw an all-important addition to the golden rules which established a red line whereby any team being specifically targeted must be withdrawn.

Concerned by this change in the nature of their responsibilities, and reckoning that they regularly violated the withdrawal-if-targeted rule, the programme managers made known their disquiet to the Board of Directors. Meanwhile, militias in Somalia were targeting infrastructure and aid workers and MSF recruited armed guards to ensure their security. This was a shift from earlier practices in Afghanistan, Eritrea and Angola as the de facto political authorities no longer provided protection. MSF teams—now managing small military units—saw their position weakening in negotiations with this new breed of employee. Despite extremely heated debates both before and during the deployment of armed guards, the decision to continue the Somalia mission won out, “given its practical utility” and the lack of “alternative solutions.” Médecins Sans Frontières, President’s Report, 1992 General Assembly [translated].

Equally controversial was the decision not to pull out the teams from Burundi in the second half of the 1990s. In 1995-96, the number of attacks and threats against humanitarian agencies escalated in the country Médecins Sans Frontières, Board of Directors, 24 November 1995 [translated].
 The situation was dire, as massacres were occurring not far from MSF teams, frustrated by their inability to provide assistance, and missions lived from one evacuation to the next. The Board of Directors held numerous discussions on the question of the risk to staff and whether to continue activities. At a June 1996 meeting, right after the assassination of three ICRC delegates, someone asked, “Why must we stay when seventeen foreigners have already been killed?” Médecins Sans Frontières, Board of Directors, 28 June 1996 [translated].
 To the question of whether the activities warranted the taking of such high risks, the operations director answered that it was sometimes necessary to take risks even when one was not treating “a lot of people.”33 On the other hand, the deputy operations director responsible for emergency missions said he “couldn’t imagine [himself] briefing someone to go to Burundi.”Ibid.
 Why were people being killed, they pondered? Was it because of where they were, or because they were humanitarian workers? The debates provided no satisfactory answers to these questions.

Those in favour of staying put advanced two arguments: the scope of the population’s needs and the teams’ willingness to continue their work. The communications director, himself a former programme manager, took the operations director to task, criticising her “sacrificial” approach. She, in turn, went on to encounter someone even more “sacrificial” when, on a visit to Burundi, she attempted to withdraw the teams from the north against the head of mission’s advice. She had to inform the departing teams herself, “because the Human Resources people no longer wanted to do it.”Interview with Brigitte Vasset, former director of operations of MSF-France, 1 December 2014 [translated].

During the Board of Directors meeting in June 1996, President Philippe Biberson argued for staying: “All this addresses a real need, and leaving would mean abandoning people.”Médecins Sans Frontières, Board of Directors, 28 June 1996 [translated].
 The Board members voted eleven to four in favour of Operations’ decision to keep the team in the field—subject to the team’s agreement. Although the decision was ultimately taken not to pull out, at times the association appeared to be feeling its way forward in the dark, with no roadmap to guide it.

The “golden rule” on being targeted was thus shattered in Burundi. Programme managers there had for some time viewed it as inadequate, given the high degree of exposure—especially in the case of projects where humanitarian workers were subject to direct attack on a regular basis.Médecins Sans Frontières, Board of Directors, 30 June 1995 [translated].
 In fact, teams were often kept in place even when field missions were embroiled in recurring violence, including in situations where it would have been impossible to get them out, such as Kigali in Rwanda during the spring of 1994 or Sierra Leone’s Freetown in the winter of 1998. However, what was striking in those chaotic times was the intensity of the discussions and the involvement of the Board members, some of whom monitored missions, conducted field visits (for example, in Somalia and Yugoslavia) and shared their analyses on their return.

Resistance to Professionalisation

As an organisation that had defined professionalisation as an historical necessity,For a partial analysis of this professionalisation movement, see Claudine Vidal and Jacques Pinel, “MSF ‘Satellites’” in Jean-Hervé Bradol and Claudine Vidal (eds), Medical Innovations in Humanitarian Situations: The Work ofMédecins Sans Frontières, New York: Médecins Sans Frontières, 2011.
 MSF might have been expected to embrace technological and bureaucratic advances in security—especially given that, in the mid-1990s, it supported more than it challenged the normalisation and professionalisation process that had accelerated significantly within the aid sector since the African Great Lakes crises. Despite strong pressure to change its practices, however, the association attempted to resist the increasingly technological and professional direction that humanitarian security was taking. In 1991, President Rony Brauman had underlined “the limitations of a global discussion on security”Médecins Sans Frontières, President’s Report, 1991 General Assembly [translated].
 and, two years later, expressed his scepticism toward the “rather approximate reports on the closing off of the world and a new international context where humanitarian action is becoming more and more difficult and less and less accepted.”Médecins Sans Frontières, President’s Report, 1993 General Assembly [translated].

MSF could not escape the discourse of “new wars”, and many felt that it was a different—and more dangerous—world. But the association’s leaders voiced their distrust of the trend toward professionalised security. “The most talked about subject after the humanitarian blues is the security of humanitarian workers! Some organisations are offering their volunteers (can we still call them volunteers ?) security training—what to do when you’re taken hostage—by retired military personnel hired as security experts. Some NGOs actually advocate sharing information and communication networks in sensitive areas with official intelligence agencies!”Médecins Sans Frontières, President’s Report, 1998 General Assembly [translated].
 fulminated in 1998 President Philippe Biberson, an advocate of MSF maintaining its own security management approach.

Analyses became more formalised with the creation in 1995 of MSF’s Centre de Réflexion, which published research into major crises—the Populations in Danger series—and contributed to coordinators’ training. “Environment Week,” first held in 1995, was not a technical training course, but was devoted to analysing the aid agencies’ environment (hence its name), the political dynamics of conflicts—an understanding of which is essential to operations management—and security. The Board of Directors had a specific place for “qualified, well-known figures” such as Jean-Christophe Rufin and political analyst Guy Hermet.

At the same time, the importance of protective measures and their reinforcement—such as calling in experts in the case of kidnappings—was discussed regularly.See, for example, Médecins Sans Frontières, Board of Directors, 11 July 1997; 30 October 1998; and 29 January 1999.
 There was no shortage of security procedures, quite the opposite in fact. The security chapters in successive editions (1990, 1994 and 2003) of Aide à l’organisation d’une mission (Guidelines for Setting Up a Mission) were full of them. They talked about the importance of ensuring that employees and equipment be clearly identified with the MSF logo, the “essential” role of the radio and the need to prepare an evacuation plan. Programme managers were sometimes alarmed at how insulated the teams were becoming and by “the walls and the barbed wire that were going up” around living and work places, without the context seeming to justify it.Interview with Marc Gastellu-Etchegorry, former director of the emergency unit of MSF-France, 3 February 2015.

Terror and the Temptations of Exceptionalism and Bureaucracy

Mounting Fear

The 2000s were marked by sustained growth for humanitarian organisations, in terms of both resources and ambitions, due in particular to the Afghanistan and Iraq invasions and the considerable humanitarian funding that accompanied them.Fabrice Weissman, “Quelle place pour les organisations humanitaires en situation de conflit?” L’état du monde 2015: Nouvelles guerres, Paris: Éditions La découverte, 2014.
 Yet it was during those years of growth that the narrative of the narrowing humanitarian space and increasing dangers to staff escalated.“Introduction” in Claire Magone et al. (eds), Humanitarian Negotiations Revealed: The MSF Experience, London: Hurst & Co., 2011.
 The deadly strikes in 2003 on the UN and ICRC headquarters in Baghdad were considered emblematic of an unprecedented rise in deliberate attacks against humanitarian workers. Fear was being fuelled by the difficulties facing relief organisations in the Middle East and Sahel, due to the expansion of radical jihadist groups and repeated kidnappings for ransom. While in the 1990s the upsurge in attacks against humanitarian workers was associated with the deliberate targeting of civilians in conflict situations, a view emerged at the turn of the 2000s condemning the targeting of humanitarians as such.

To use the words of Larissa Fast,Fast, op. cit.
 a narrative of “humanitarian exceptionalism” was being constructed, portraying humanitarian workers as heroes and martyrs. One specific illustration was the designation of19 August—the anniversary of the attack on the United Nations in Baghdad—as World Humanitarian Day to honour “those who face danger and adversity in order to help others.”See http://www.un.org/en/events/humanitarianday/, last accessed 22 December 2015.
 Statistics would prove a powerful ally in this victimisation device. Starting in the early 2000s, relief and research organisations conducted numerous quantitative studies on violence against humanitarian workers, all of them concluding that increasing insecurity was a scientifically established fact.See Chapter 4, p. 55.
Educational institutions began offering degrees in security management. In 2000, the ODI published authoritative guidelines entitled “Operational Security Management in Violent Environments,” in which author Koenraad Van Brabant devoted 350 pages to defining good humanitarian security practice.See Chapter 5, p. 71.
 In December 2004, the UN created the Department of Safety and Security (UNDSS) with, as its first director, a former senior Scotland Yard officer. The trend toward professionalisation was justified, not only by “new threats”, but also by the need for humanitarian organisations to meet their legal obligations as employers.

In the 2000s, Médecins Sans Frontières was also experiencing steady growth. The budgets for its missions rose from €59 million in 1998 to €219 million in 2010, with 600 international field posts that year compared to only 400 ten years earlier. From 1996 to 2012, the number of national staff7 under contract grew from approximately 3,000 to more than 5,500.

Those growth years were also years of grief and anxiety for MSF as the organisation suffered a series of assassinations and kidnappings. In July 2000, a French volunteer was abducted in Colombia and held for six months and, in 2001, MSF-Holland’s head of mission in Chechnya was also abducted and released some weeks later. In August 2002, the head of the Swiss section’s mission in Dagestan was kidnapped.See Chapter 8, p. 127.
 Held for almost two years, his release caused a public dispute between MSF and the Dutch government, which took the organisation to court to demand payment of the ransom it claimed to have paid. Between 2004 and 2008, six international staff" members were assassinated while on mission. Five members of the Dutch section, including two Afghanis, were executed in June 2004 in Afghanistan’s Badghis Province and an MSF-France logistician died in 2007 when her vehicle was ambushed in Central African Republic. The following year, two international staff" members of MSF-Holland and their Somali colleague were killed when their car was hit by an exploding roadside bomb in Kismayo in Somalia.

Those events lent credence to victim discourse and statistics on the worsening security situation, which were echoed by MSF. “It is very important to remember that between 2000 and 2005, 271 international humanitarian workers were killed, [and that] the number of high-risk situations, hold-ups, abductions and physical attacks experienced by our teams continues to grow,” lamented in 2006 the Boards of MSF-France and its partner sections (MSF- USA, MSF-Australia and MSF-Japan), who were becoming increasingly involved in decisions on the conduct of the social mission. Indeed, the internationalisation of MSF also contributed to the rise in security concerns. In 2006, the partner sections became responsible for paying their national personnel, who had previously been under contract to MSF-France. More employers meant more legislative frameworks governing their legal liability (“duty of care”)See Chapter 5, box: “Who Benefits From ‘Duty of Care ?’”, p. 82.
 for personnel security. In response to this increasing legal pressure, the contracting sections sharpened their requirements.

Such was the background for debates at General Assemblies and meetings of the Board of Directors, which were dominated by at least three issues: the persistent unease created by the discrepancies between the rules instituted in the early 1990s and what was actually happening in the field, the role of head office and the Board of Directors in evaluating risk and making decisions and the legitimacy of transferring risk to national staff or to nationalities at lower risk.

Were the “Golden Rules" Obsolete?

Note that while references to the principles contained in the “golden rules” were everywhere in the discussions, the term itself disappeared from mention.

What was the explanation for the attacks against MSF and the ICRC, particularly in Iraq and Afghanistan? A source of concern to MSF leaders was the use of humanitarian rhetoric by Western powers that created a “deadly confusion” between NGOs and foreign armed forces. This was especially the case in Iraq and Afghanistan, and even more so after the Taliban claimed responsibility for murdering five MSF members in Afghanistan’s Badghis Province in June 2004, accusing MSF of “work[ing] for American interests.”See, in particular, Fabrice Weissman, “Military Humanitarianism: A Deadly Confusion,” in the International Activity Report 2003-2004, Paris: MSF.
 Yet the President’s Report for the year 2000 reveals the critical distance that Philippe Biberson wanted to maintain from a narrative blaming such confusion for all of the danger: “If NGOs team up with the military, then they will logically be considered a party to the conflict and targeted or prohibited from staying. [...] But we can also come up with loads of situations where our security depends on what people imagine our political sympathies to be and on the number of guards around us. In reality, and since time immemorial, it is not independence that conditions access to victims, or even the safety of our teams. Most often, it is negotiation (and logistics.).”Médecins Sans Frontières, President’s Report, 2000 General Assembly [translated].

However, what seemed to be worrying the association was that “extremist groups have clearly stated that humanitarian workers would be targeted” in Afghanistan and Iraq.Médecins Sans Frontières, Board of Directors, 24 April 2003 [translated].
 During the Board of Directors meeting following the 2003 assassination of an ICRC representative in Uruzgan in Afghanistan, association president Jean-Herve Bradol expressed his concern at “having to do briefings where we have to warn departing volunteers that there are people who have it in for us,” adding that “ten years ago, it was clear to us that this was the red line that would make us give up.”Ibid.
 “We have no desire to be martyrs to the humanitarian cause, it would be absurd,” he continued a few months later.Médecins Sans Frontières, Board of Directors, 29 October 2004 [translated].

In this he was following in his predecessors’ footsteps. He recognised that the conflict between what was being said and what was being done had to be elucidated. Yet, as we have seen, that conflict already existed back in the 1990s.Médecins Sans Frontières, Board of Directors, 25 October 2002 [translated].
 As the entire MSF movement was in the midst of heated debate over the future of activities in Iraq at the time of the US invasion of the country, Bradol confirmed: “The truth about our current risk exposure policy is that it seems to assume that people are being killed or seriously injured on a regular basis.”Médecins Sans Frontières, Board of Directors, 24 April 2003 [translated].
 He challenged this shift all the more since in his view what it achieved was not justified by the operational results. Hence, it was precisely because he believed in the usefulness of delivering aid in Baghdad when the US invaded Iraq that he supported keeping teams on the ground. He was outraged that those opposed to his thinking could suspect him of wanting to deploy teams “simply in the name of an ideal”—“We send teams in when we believe concrete assistance can be provided, which is the case in war-torn towns and cities”.Médecins Sans Frontières, Board of Directors, 28 March 2003 [translated].
 After the killings in Badghis in Afghanistan, he criticised “poorly thought-out, oversized” operations.Médecins Sans Frontières, Board of Directors, 30 September 2004 [translated].

Trust Procedures and Be Wary of People?

These questions prompted a new round of discussions about the respective roles of the head office, the Board of Directors and field volunteers in taking risks. At a Board of Directors meeting on Central African Republic a little more than a year after the death of a volunteer logistician there, the programme manager asked that the directors collectively assume mission-related risks and make a greater contribution to operational discussions.Médecins Sans Frontières, Board of Directors, 29 August 2008 [translated].

The following year, when the issue of directors’ participation in discussions on security came up again, a member of the management team (who had been operations director from 1986 to 1998) recalled “administrators visiting the missions in an effort to bring a fresh (but not inexperienced) eye. They asked naive questions, sometimes painful, and then they went back to report to the other directors, to share and have exchanges with the field.”Médecins Sans Frontières, Board of Directors, 27 February 2009 [translated].
 It is hard not to see this comment as barely veiled criticism of administrators less involved than some of their predecessors in the 1990s.

Successive presidents cautioned everyone against the importance being accorded to procedures and the tendency to centralise decision-making at head office, to the detriment of individual judgment, despite the fact that “the most important protection is our positioning, our understanding of the context and our ability to establish relationships.” From that perspective, it was emphasised, “the Board of Directors is more inclined to trust in people, rather than a system and procedures, to make decisions.”Médecins Sans Frontières, President’s Report, 2003 General Assembly [translated].

In his final President’s Report in 2008, Jean-Hervé Bradol distinguished between the responsibilities of the association—to ensure that there is “a certain type of efficacy of action” and that “particular attention is paid to the misappropriation of our resources,” and to draw the line when “a political group that could relatively realistically put its threats into action announces that it intends to target humanitarian workers and assassinate them”—and the decision of each individual on exposing themselves to risk.

“Remote Control” and “Profiling”

In more and more settings, “trust in people” means trusting in national personnel, to whom the day-to-day management of activities is delegated, while international managers visit as often as possible. This operating mode—known in humanitarian parlance as “remote control” and often considered a major compromise and downgrade of the conventional method of interventions— was the price MSF had to pay to continue working in places such as the Russian Caucasus and later Somalia. Remote control raises questions about the status of national staff (are they MSF, like the others?) and the specific risks that their involvement in the situation, be it their social, emotional or political ties, might lead them to take.

The setting up of remote control coincided with a discussion on the place and role of national staff that went far beyond the issue of security. Under discussion was a proactive policy intended to raise the status of national employees through better remuneration and access to expatriation, training and association membership. It should be noted that MSF only started keeping records of its local employees in 1994, and, except for a few anecdotal cases, it would take the organisation almost another ten years to seriously take account of their security. These concerns came to a head a few years later as expressed by President Marie-Pierre Allié in 2009: “[W]e should be thinking about the risks our staff are taking. It seems to me that, when we have only national staff in the field, we don’t take into account something very specific: that their personal involvement with the local population may push them to take greater risks than we would wish for them. We should be careful not to underestimate the risks they take.”Médecins Sans Frontières, Board of Directors, 27 February 2009 [translated].

The growth of radical Islamic groups—notably with the emergence of Al-Shabaab in Somalia and Al-Qaeda’s mounting influence in the Maghreb and Arabian Peninsula—contributed to more heated debate on the threat of kidnap and “profiling”; that is, recruiting volunteers based on their gender, religion, nationality or skin colour. Thus, it was explained, given that certain profiles are less exposed, “MSF might consider choosing international staff" ‘compatible’ with the situation, for example, by ‘Africanising’ teams working in the Sahel.”Médecins Sans Frontières, Joint Board meeting, 27 January 2012. Far from considering withdrawal an option—except in situations where the association is directly faced with the death of its international volunteers, like in Somalia and Afghanistan—MSF views the adjustments to its modus operandi as a pragmatic response to allow it to continue to work in settings where it is highly exposed.

The technical and procedural components of security are expanding and becoming increasingly centralised. While it is difficult to determine objectively what the dangers are and whether they are worsening, it is an established fact that fear—as a social construct—is mounting. The fear of abduction looms particularly large in MSF’s attention to what it now labels “highly insecure environments”. In light of this, the use of technology may appear to be a solution that reassures. The 2003 edition of Aide a lOrganisation logistique dune mission makes the logistics coordinator responsible for “limiting the risks taken by the teams through ensuring that the means and methods implemented for security are present, reliable and correctly used,” thus testifying to the increasingly compartmentalised and technology-dependent nature of security management in the missions. In the early 2010s, the operations director recognised “the pressure to professionalise security management.”Médecins Sans Frontières, Joint Board meeting, 6-7 December 2013.

Expressions adopted from leading publications in the sector—such as “risk analysis” or the “acceptance, protection and deterrence” security triangle—are coming into widespread use in MSF’s missions and ever-increasing in-house security training courses. Previously unwilling to appoint a “security focal point”, the French section—which had been an exception, including within the MSF movement—finally yielded in 2013.

***

 

Danger, Risk, Security and Protection : Concepts at the Heart of the History of Humanitarian Aid

Bertrand TaitheThis article is based on two ESRC-funded research projects: UN Data ES/ L007479/1 and Selling Compassion ES/I031359/1. It was presented as a paper at the Humanitarian Studies Association conference in Istanbul in 2013, in a paper version at MSF-France in Paris and at the Humanitarian Congress in Berlin in 2014. Thanks to all the collaborators and colleagues involved, Roger MacGinty, Roisin

The concepts of danger, risk, security and protection—none of which are self-evident or simply observable realities—require a broad historical frame to make sense of their meanings in current debates. They have been borrowed, shaped and reinvented in the discharging of humanitarian policies as ways of engaging with aid work. Humanitarians have consistently made their deployment in the face of danger a badge of honour (this use of a nineteenth century notion is deliberate, as it conveys the origins of numerous contemporary issues). In the face of danger, while taking risks, aid workers have always paid attention to their security and sought to define how their work could be made safer, often combining practical measures on the ground with more discursive claims to provide and obtain protection for and through their work.

Yet the evidence shows that in some acutely violent places security was at times minimal or indeed almost non-existent and humanitarians’ demands for protection were flouted. Evaluation and management of risk to establish security measures as well as calls for protection turned out to be, for the historian, essential tools for representing and comprehending, not only the world in which humanitarians work, but also the humanitarians themselves.

This chapter comprises three parts. The first provides a brief history of the tools at the centre of risk and protection measures, the second, a longer view of the relationship between security and protection, and the third, how the two concepts have been set, since the 1990s, as a dialectic challenge to humanitarians. It concludes with a reflection on how these concepts have reshaped the notion of the “humanitarian field” in relation to humanitarian work.

Evaluations and Risk

The Legal and Insurance Thinking Behind “Risk”

Humanitarian aid was for the most part deployed throughout the nineteenth century to mitigate the consequences of disasters and industrial accidents, principally fires, shipwrecks, floods and mining disasters. The notion of risk is closely linked to that history, on several accounts. The concept of risk assessment stems from the legal and insurance policy language adopted in the mid-nineteenth century, when the term was adopted by loss adjusters and actuaries whose task it was to anticipate the full extent of risk-taking, even in dangerous but nevertheless insurable occupations.A. H. Smee and Thomas G. Ackland, “On the Assurance Risks Incident to Professional Military and Naval Lives; and the Rates of Extra Premiums Which Should be Charged for Such Risks. Being Extracts from a Joint Report, made in May 1890, to the Board of Directors of the Gresham Life Assurance Society,” Journal ofthe Institute of Actuaries, vol. 34, no. 4, 1899, pp. 358-385.
 Their main challenge was to establish who was taking the risk and to what extent exposure to danger represented evidence of negligence. Was danger preventable and, if so, by whom and in what timeframe ? Was an employee taking a risk a wilful or negligent act ? Did people step unknowingly into the path of danger ?

These rather obscure points of law are of importance to the humanitarian sector on two levels. Firstly, the history of risk in the wider social context is rooted in the history of legal as well as insurance policy thinking (including social insurance policies), and secondly, the manner in which humanitarians and organisations think about danger relates to the legal history of accidents and criminal negligence trials.Bill Luckin and Roger Cooter (eds), Accidents in History: Injuries, Fatalities and Social Relations, Amsterdam: Clio Medica, 1997.

There are numerous examples of a voluntary compensation culture whereby funds were raised to compensate the losses of innocent victims and rebuild the lives of survivors, providing the model for reconstruction and rehabilitation relief in times of war. In 1871, the English Quakers decided to disregard soldiers (even though they were viewed as the primary humanitarian subjects) to focus instead on rebuilding civilian lives after the ravages of the Franco-Prussian War.William K. Sessions, They Chose the Star: Quaker War Relief Work in France, 1870-1875, York: Sessions Books, 1991.Like other humanitarian workers in the United Kingdom, they did not venture unequipped into the field and, similar to most voluntary societies or relief funds, they relied on evidence drawn up by accountants. They were particularly dependent on reports compiled by the actuary, a new profession in accountancy. Actuarial reports were specifically designed to provide “scientifically” grounded evaluations of financial liability and measure financial risk to ensure effective and accountable management of relief funds raised for a specific purpose. The term “security” in this economic model referred to the assets held against such risk.Graham Benjamin et al., Security Analysis, New York: McGraw-Hill, 1934; Ingvar Laurin, “An Introduction into Lundberg’s Theory ofRisk”, Scandinavian Actuarial Journal no. 1, 1930, pp. 84-111; William O. Douglas, “Vicarious Liability and Administration of Risk I.”, Yale Law Journal no. 38, 1928, p. 584.Funds committed to relieve the sufferings of widows and orphans had to serve over the long-term because they were pledged to provide until the natural death of the beneficiaries. Consequently, the long lives of beneficiaries were viewed as a risk to the fund. A further risk, a sort of “moral hazard”, emerged when the funds were deemed to be too generous as, in the nineteenth century, providing relief was to run the risk of fostering dependency and through dependency to create unlimited liability.Joseph Brown, “The evils of the unlimited liability for accidents of masters and railway companies, especially since Lord Campbell’s act: a paper read before the Social Science Association”, Butterworths, 1870.Charities were therefore keen to provide enough for long enough, but not too generously. Combined with legal requirements, these moral and practical financial considerations defined risk.

Security issues were to be understood in relation to those arising from limited or unlimited liability. In Anglo-American common law, the codification of danger is related to the notion of peril.William Colebrooke, “Negligence in Imminent Peril”, American Law Register (1886), pp. 617-632.In many ways this term defined the concept of preventable danger, the responsibility of individuals in relation to each other and the relative degree of negligence of each party involved in an accident. In the case of accidents, the original “doctrine” of “last clear chance” defined the responsibility of one negligent party towards another as the responsibility of an individual to assist a person in danger if at all possible (without incurring excessive risk) or if the danger could have been foreseen.J. S. S. “The Doctrine of Last Clear Chance in Virginia”, Virginia Law Review, vol. 40, no. 5, 1954, pp. 666-680.The legal obligation to intervene in order to prevent harm was defined in court primarily as a duty to assist those in peril.Huey B. Howerton Jr, “Tort Liability for Failure to Assist Others in Peril”, Mississippi Law Journal no. 16, 1943, p. 379.In many respects, these legal concerns (the French “non-assistance apersonne en danger”)Will D. Davis, “Doctrine of Discovered Peril”, Baylor Law Review, no. 6, 1953, p. 61.
 related closely to the origins of humanitarianism that merely extended their remit—but not their legal framework—to the global stage.

If the notion of danger and obligation to intervene were thus established early on in Roman and common law, the relationship and unnecessary exposure to danger have a rather more complex and controversial history. Known as the “humanitarian doctrine”,The word “humanitarian” refers here to the spirit of the doctrine—one concerned with the welfare of victims—rather than to humanitarian work in our current understanding. This nevertheless illustrates how plastic and widespread the con¬cept ofhumanitarianism was before it became the preserve ofhumanitarian organ¬isations or international humanitarian law.a new concept emerged in the early twentieth century in Anglo-American common law which established that to take risk needlessly was not necessarily an admission of full responsibility. In other words, when a person puts themselves in danger out of negligence and the danger itself was created by the negligence of others (typically a moving vehicle or an industrial accident), the two negligent acts do not cancel each other out and the victim might still seek redress or compensation. In humanitarian doctrine, the negligent party at the origin of the danger to which other negligent parties would subsequently expose themselves was still the source of the accident. In practice, this meant that careless employees exposed to unnecessary risks by their employers could still seek legal redress against them. In medicine, the notion of risk was often raised in relation to insanity and the danger a patient might present to themselves or others, making risk assessments a common prerequisite for internment in secure hospitals.

Humanitarian efforts of the late nineteenth century and the contemporary humanitarian matrix originated from this capitalistic social context. The logic and structure of these early efforts were in line with the practices of their promoters.

Risk Exposure: From Insurance to Humanitarianism

The founders of Western humanitarian aid—such as the bankers and lawyers of Geneva, or, in the British Empire, Lord Sutherland and his Stafford House Committee—used a language and logic acquired from their legal and financial practices. The archives of Stafford House are to be found at the Staffordshire County Record Office (SCRO) in the private papers of Lord Sutherland. See Sarah Roddy et al., Selling Compassion, forthcoming.
 They were applying their professional standards to the management of resources mobilised for humanitarian relief. The humanitarians responsible for dispatching medical relief to France during the Franco- Prussian War in 1870-1871, the Ottoman Empire in 1878 and South Africa during the 1899-1902 Boer War adopted the same cautious approach as they would for their own investments. For example, the administrator of the Stafford House Committee, which raised resources for a wide range of humanitarian operations and ran field hospitals during the 1877-1878 Russo-Turkish War, demonstrated full accountability, precision, prudence and responsibility.

In practice, this meant that the Committee, which in 1877-1878 funded fifty medical staff, maintained twenty hospitals, handled three evacuations by train and treated over 75,000 surgical cases, exercised careful management of its funds while appraising the nature of its work, the duration of its operations and its exit strategy according to the rules of risk management.Stafford House Committee for the Relief of Sick and Wounded Turkish Soldiers, Report and Record ofthe Operations ofthe Stafford House Committee, Russo Turkish War, 1877-78, London: Spottiswoode & Co, 1879.
 One of the most important risks for the Committee was its image and the fund’s reputation. Notably, it had to fend off allegations of corruption that arose from working too closely with Ottoman politicians. Evening Standard, 2 September 1877; SCRO D593/P/26/2/7.

Yet this language of prudence was also melded with recognition of the dangers and risks inherent to war. The two were perfectly compatible because intervening during a war was not expected to be without danger. In many respects, perils and dangers belonged to a different semantic and cultural register, highlighting and valuing danger as an opportunity to reveal individual valour, masculinity, compassion and character. This language of danger was to be found in travel narratives often recounting the risks taken by travellers, missionaries and vicarious humanitarians and their overcoming of danger. The figure of the heroic explorer standing alone in the face of great danger is significant among the humanitarian leaders of the late nineteenth century— from Dr Livingstone to General Gordon See Max Jones, The Last Great Quest: Captain Scott’s Antarctic Sacrifice, Oxford: Oxford University Press, 2004; Max Jones et al., ‘Decolonising imperial heroes: Britain and France.’ The Journal of Imperial and Commonwealth History, vol. 42, no. 5, 2014, pp. 787-825.
 or the more controversial Roger Casement.Andrew Porter, “Sir Roger Casement and the international humanitarian movement”, The Journal of Imperial and Commonwealth History, vol. 29, no. 2, 2001, pp. 59-74.
 In the twentieth century, polar explorer Fridtjof Nansen, who embodied the humanitarian agenda of the League of Nations, was himself an adventurer in the same noble tradition of danger-seeking individuals.Bruno Cabanes, The Great War and the Origins ofHumanitarianism, 1918-1924, Cambridge: Cambridge University Press, 2014, pp. 133-188.

This relationship between travel narratives, heroics and humanitarian aid has not entirely disappeared. I came across this respect for the skills of the traveller during a recent interview with Jacques Pinel, the pioneer of logistics at MSF, when he informed me that, in the early days, he primarily recruited experienced and adventure-seeking globetrotters and backpackers, the so-called lroutards’, as logisticians.Interview with Jacques Pinel, 7 March 2013.

Some of the field experience so prized in humanitarian circles is still expressed through rhetorical tropes and sometimes tones of orientalism dating back to the late nineteenth century. Courage is, to this day, a highly valued humanitarian virtue. It is not surprising then that many of the humanitarian ego-narratives emerging from this period were framed as “adventures”.For instance R. B. Macpherson, Under the Red Crescent; Or Ambulance Adventures in theRusso-Turkish War of1877-78, rare book club reprint, 2012. The language of adventure recurs throughout humanitarian writings; see for instance Jean¬Christophe Rufin, LAventure humanitaire, Paris: Gallimard, 1994.
 In the dry language of reports and in first-hand accounts, humanitarians would highlight danger in self-reinforcing terms, often dictating a new economy of relief and practice. As Rebecca Gill Rebecca Gill, Calculating Compassion, Manchester: Manchester University Press, 2013, pp. 63-5.
 has shown and I myself have explored, Bertrand Taithe, “Horror, Abjection and Compassion: From Dunant to Compassion Fatigue”, New Formations, no. 62 (2007), pp. 123-36.this language of emotive engagement was compatible in practice (although not in discourse) with the “cold calculation” and reasoned action called for by Henry Dunant in his A Memory of Solferino? J. Henry Dunant, A Memory of Solferino. 1862, Washington D.C.: American National Red Cross, 1939.

Yet such reasoning was by no means precise. While by the 1880s it was fairly well established what the statistical chances were that a miner’s widow might live to the age of seventy, estimating the chances of being killed in a war was still a somewhat inexact science. Calculating the precise number of war victims was no less straightforward. The discursive work of Jean-Charles Chenu, the French statistician involved at the origins of the French Red Cross, Jean-Charles Chenu, De la mortalité dans l’armée et des moyens d’économiser la vie humaine, Paris: Hachette, 1870; Claire Fredj, “Compter les morts de Crimée: un tournant sur l’identité professionnelle des médecins de l’armée française (1865-1882)”, Histoire, économie & société, vol. 29, no. 3, 2010, pp. 95-108.illustrated how war itself remained a largely unknown quantity. For instance, as the Crimean War (1853-1856) was fought (on land at least) on a peninsula accessible only by ship, it should have been relatively easy for the authorities to calculate casualty numbers, given that they knew how many soldiers had been sent in and how many returned. But it would take them over three years to produce detailed statistical accounts and narratives. Still today, war casualty accounting is a source of debate and controversy. Even in times of open warfare, it is no easy task to classify exactly what is the result of direct violence rather than accident, what destruction takes place by design rather than so-called “collateral damage”.

Humanitarian concerns reshaped the perception of war, not simply as danger faced in battle but also, and often predominantly, as biohazard. Evidence shows that during the pre-World War One era humanitarian workers were far more likely to succumb to diseases contracted by soldiers and refugees than to physical violence. The nature of that risk remained framed within the medical literature. Yet biosecurity was already a major concern, with early Pasteurian medicine’s vaccines offering only limited protection. The risks to surgeons and health workers during surgery were extremely high in the early days of modern humanitarian aid (so-called hospital rot, septicaemia, blood poisoning and other forms of cross-contamination). They undoubtedly lessened during the twentieth century, thanks to new standards in cleanliness and asepsis, but the risk of becoming contaminated during typhus, typhoid, cholera and plague epidemics remained real until the 1940s. In the 1870s, a substantial percentage of medical staff" would fall ill and some would die in any humanitarian operation involving large numbers of civilians and soldiers. For instance, by the end of June 1878, one-third of the thirty-nine medical staff serving directly under the responsibility of Stafford House Committee manager Barrington Kennett had been taken ill with typhus; two died of the disease, but none from the conflict. Stafford House Committee for the Relief of Sick and Wounded Turkish Soldiers, Report and Record of the Operations of the Stafford House Committee, Russo Turkish War, 1877-78, London: Spottiswoode & Co, 1879, p. 40. For the same period, seven had died of the forty-five working for the Red Crescent; in the British Red Cross, three out of fourteen were ill.
 By historical standards, Ebola and the risks it poses to aid workers would appear to be more of a throwback to the past than the emergence of a new category of humanitarian risk.

Security and Protection

Humanitarian Sanctity?

The historical relationship between medical humanitarian aid and wartime ‘secours aux blesses’“Aid for the wounded”.
 (as the original name of the Red Cross originally entailed) requires looking back at history to consider how notions of “security”, “danger” and “risk” were encoded and used and how humanitarians, who have always been working under fire, related to those notions to make sense of their practices and cope, individually as well as collectively.

The 1870-1871 Franco-Prussian War provides instances of the enhanced status afforded to hospitals as sanctuaries and the repeated violations of this status. The right of sanctuary constituted a form of protection in as much as the Red Cross emblem signalled an international healthcare space in the midst of war. Part of the moral apparatus of the Geneva system, this implicit internationalisation of conflicts remains to this day. Yet the use of the emblem was prone to abuses and sanctuary to violation. In 1870, German authorities accused the French public of abusing the system by claiming right of sanctuary for individual houses converted into makeshift ambulances, for example in Le Mans. A house protected by the Red Cross emblem could not be billeted with enemy soldiers, but Germans objected to “hospitals” which contained only one or two wounded soldiers. French authorities accused the German high command of shelling hospitals despite the legal protection afforded by the Red Cross flag.Anon., Les violations de la convention de Genève par les Français en 1870-1871, Berlin: Charles Duncker, 1871; J. M. Félix Christot, Le Massacre de l’ambulance de Saône-et-Loire le 21 janvier 1871; Rapport lu au Comité médical de secours aux blessés le 7 juillet 1871, Lyon: Vingtrinier, 1871; Charles Aimé Dauban, La Guerre comme la font les Prussiens, Paris: Plon, 1870; Bertrand Taithe, Defeated Flesh: Welfare, Warfare and the Making of Modern France, Manchester: Manchester University Press, 1999, pp. 169-73.
 Throughout the 1877-1878 Russo-Turkish War, field hospitals funded by the Stafford House Committee witnessed numerous violations of the Geneva Convention. In January 1878, the hospital where Drs Beresford and Stiven were working in Rustchuk (now the city of Ruse in Bulgaria) was systematically shelled:

"There was no doubting now the intention of the Russians as regarded our hospital, as shell after shell fell in our vicinity while we were busily engaged in placing the patients under the protection of the centre wall of the first ward. So great was the panic caused by the first three shells that were fired, that all the patients that were able to walk took flight in the open plain, where the snow is at present lying over three feet deep, and not only they but all the domestics and other officers of the hospital, so that Dr Beresford and myself were quite alone with some 80 patients to do what best we could for their safety. We went to our work, nevertheless, and lifted the patients in our arms and placed them on mattresses under the wall. The Russians kept on firing till sunset up to which time they had fired between 30 and 40 shells at our hospital, eight of which entered into the different wards of the hospital."Stafford House Committee for the Relief of Sick and Wounded Turkish Soldiers, Report and Record of the Operations of the Stafford House Committee, Russo Turkish War, 1877-78, London: Spottiswoode & Co, 1879, p. 50.

Stiven handed over to the British press the names of the Russian batteries guilty of the transgression (Menschikoff and Esmurda) along with those of the officers in charge in an attempt to name and shame and bring about some symbolic redress. Similar anecdotes exist for all conflicts since 1870, confirming the lack of substance behind claims of sanctuary or, at the very least, their contested status in the midst of fast-moving tactical war operations. MacPherson, op. cit., p. 17.

In reality, negotiating neutrality and the use of recognisable emblems were never that straightforward. As a general rule, conspicuous respect for humanitarian neutrality has always been part of a wider strategic master plan, based on reciprocity and/or on the need to establish the legitimacy of the combatant parties. It was undoubtedly a significant propaganda victory for a new world power when the Japanese army received praise for its admirable treatment of Russian prisoners in 1904-1905. Teresa Eden Pearce-Serocold Richardson, In Japanese Hospitals During War-time: Fifteen Months with the Red Cross Society ofJapan (April 1904 to July 1905), New York: W. Blackwood and Sons, 1905; Philip A. Towle, “Japanese Treatment of Prisoners in 1904-1905: Foreign Officers’ Reports”, Military Affairs: The Journal of Military History (1975), pp. 115-118.

These tools of negotiation were key, particularly when the parties involved made their claims to sovereignty via the protection and responsibilities for delivery of humanitarian aid. One of the key outcomes of early twentieth- century wars was the embedding of humanitarian relief as an auxiliary of recognised military health hierarchies and structures. Humanitarians regularly wore special uniforms and took on specific social roles, which, while civilian in nature, were associated with treating the wounded, the good handling of prisoners and even the rituals around the disposing of corpses following wars or natural disasters in China. Caroline Reeves, “Sovereignty and the Chinese Red Cross Society: The Differentiated Practice of International Law in Shandong, 1914-1916”, Journal of the History of International Law/Revue d’histoire du droit international, vol. 13, no. 1, 2011, pp. 155-177.
 The trade-off for these roles was the safety and neutrality granted to medical staff", despite their proximity to the military. But revolutionary and insurrectional warfare granted no such privilege. Civil wars provide many instances of violations of the neutrality of casualties and challenges to the concept of humanitarian sanctuary. Humanitarians themselves often took sides, rejecting any notion that their work should be neutral. Volunteers to the American Medical Bureau Field Unit in the Abraham Lincoln Brigade were explicitly adjuncts to the International Brigade movement during the Spanish Civil War in the 1930s.Maria del Carmen Pérez-Aguado et al., “Medicine and nursing in the Spanish Civil War: women who served in the health services of the International Brigades (1936-1939)”, Vesalius (2010) Suppl: 29-33.Nevertheless, in conventional conflicts, a person could be partisan and still claim Genevan neutrality. During the Russo-Turkish conflict these humanitarians’ predecessors were able to claim neutrality (largely in vain) under the Geneva Convention, despite the fact that they only assisted Ottoman soldiers and subjects. But there was even less hope of the principles of the Convention being evoked with any success in the case of an extremely cruel civil war. In the Spanish context, risk-taking was often portrayed as the nature of the engagement of volunteers and a testament to the solidarity of non-combatant forces with fighting units.

Risk-Taking Humanitarians

While danger and risk affect individuals differently, the impact on organisations is far more consistent. An organisation cannot be brave, only its members can, and safety was always a concern for those kept at arms-length from danger. Even in the earliest accounts of danger recounted using the most heroic language, organisations and their administrators alike sought to negotiate the safe passage of humanitarian practitioners. In most cases, their safety was entrusted to third parties (who could be called “brokers” or “gatekeepers”), These terms are used in their current anthropological meaning to denote power over access and exchanges.
 government and local figures of authority. While “characteristics of courage, devotion and endurance” 3rd Duke of Sutherland, preface, “Report and Record of the Operations of the Stafford House Committee for the Relief of Sick and Wounded Turkish Soldiers”, p. 4.remained paramount, they were unfailingly framed within a context of alleviated risk.

The nature of humanitarian work and the conditions in which it was, and still is, practiced often confronted workers with exceptional forms of suffering and posed new personal risks. The notion of risk to self was always of fundamental concern. While the first 1864 Geneva Convention pre-dated the “psychological turn” of the late nineteenth century, danger in the battlefield had to be understood fairly broadly, since both overwork and excessive compassion could become forms of risk. Indeed, the first instance of burn-out in humanitarian context is depicted by Dunant, who portrayed it as the character failing of a sentimental do-gooder. Primarily originating in accidents—railway in particular—the late Victorian notion of trauma paved the way not only for disaster medicine and emergency relief but also for the treatment of psychiatric trauma, which shares the same origins.See Mark Micale and Paul Lerner (eds), Traumatic Pasts: History, Psychiatry and Trauma in the Modern Age, 1870-1930, Cambridge: Cambridge University Press,
2010, pp. 2-27.Constant exposure to risk and suffering were often presented as two sides of the same damaging context. Diaries and memoirs almost invariably recounted moments of considerable anxiety and sometimes informal support networks, but seldom formal debriefing processes—even for wounded medical personnel and prisoners. Convalescence was often the term used to describe recovery from the exertions of humanitarian work. Most of the suffering was framed in religious tones that made eschatological sense: “We passed safely out of this valley of the shadow of death”, reported the surgeon at the hospital in Kars, the site of a decisive but bloody Russian victory during the 1877-1878 Russo-Turkish War. MacPherson, op. cit., p. 119.

For much of the nineteenth and twentieth centuries, this religious or spiritual undertaking of dangerous activities, a kind of often humble but heroic economy of risk-taking in the face of increasingly violent events, dominated and existed alongside the accounting and actuarial understanding of risk that the humanitarian enterprise represented. This accommodation of two logics, seemingly contradictory, was facilitated by the rudimentary bureaucratic processes adopted by humanitarian agencies and the autonomy made necessary by distance from headquarters. Retrospective accounts and letters from the field present evidence of both logics in the same documents. It is striking that risk was not quantified and security remained a loose and deflationary concept. In conflict situations, people were at risk, security was relative and whoever was granting safe passage might be unable to do so the next day. All employers could do was to rely on vague assurances from third parties—and hope for the best.

Furthermore, contractual agreements with volunteers were radically at odds with contemporary employment law. Since the close of the nineteenth century, the liability of the employer in civilian society across Europe had been presumed over that of the employee (only the employee’s negligence had to be proved as the cause of the accident rather than, as in the past, simply presuming it) but the rule did not apply to volunteers. This loophole remains in practice, even if many voluntary organisations since the 1870s have paid fulltime humanitarians a per diem, salary or stipend. In this specific contractual arrangement, which is more of an informal convention than actual law, the volunteer is an associate rather than an employee of a humanitarian agency.

Of course, even then, this was largely a fiction and many employees did not volunteer for risk-taking. The distinction between a hired employee and a paid volunteer remains obscure. For instance, Henry Dunant’s Solferino coachman had not bargained on risking his life when he took his passenger, nor were the more menial ambulance employees in Kars expecting their fate to be abandoned to the Russians. The culture of risk-taking and danger-facing was, to some extent, self-mythologising and self-glorifying, more revealing of how humanitarians told their stories than of the realities they faced in the field or the risk they imposed on others. In that regard, there are a lot of commonalities with the early ages of MSF. See Chapter 2, p. 21.

Some years after the Franco-Prussian War, the German lawyer Carl Luder produced an emotionless survey that won the Augusta Prize for the best book on humanitarian work. A severe critique of self-congratulatory humanitarian narratives, it echoed scathing military criticisms of the relevance and legitimacy of humanitarians in conflict situations.M. C. Lüder, La Convention de Genève au point de vue pratique, théorique et dogmatique, Erlangen: E. Besold, 1876.
 Combatants rejected the encumbrance of amateur humanitarians while they were winning and found the benefit of their reliefwork unreliable when they needed it. Bertrand Taithe, op. cit., pp. 75-90.
 Yet Luder had failed to appreciate, as have many military commentators since, how entrenched the legitimacy of the Geneva principles had become in the practices of modern war, even though, or perhaps even because, these had often been ignored during the war itself. Furthermore, narratives on the Franco-Prussian War were intended to establish the precedent for later interventions.

At this early stage of the provision of international humanitarian aid in situations of war, the promoting of a new emblem, the new set of principles drafted in Geneva and new legitimacy rooted in an unprecedented delivery of aid and the emphasis on the voluntary and gratuitous nature of humanitarian work were all part of historical humanitarian narratives. This quest for legitimacy was also founded on the modernity of humanitarian aid, the promises of international lawJacques Meurant, “Inter Arma Caritas: Evolution and Nature of International Humanitarian Law”, Journal of Peace Research, vol. 24, no. 3, 1987, pp. 237-249.
 and the enthusiasm inspired by new compassionate attitudes. Yet it became a narrative process early on. By the time the first Nobel Peace Prize was awarded to Henry Dunant in 1901,André Durand, “Le premier Prix Nobel de la Paix (1901): Candidatures d’Henry Dunant, de Gustave Moynier et du Comité international de la Croix-Rouge”, Revue Internationale de la Croix-Rouge/International Review of the Red Cross, no. 842, pp. 275-285.
 the mass of these narratives—by then a dense forest of pamphlets, books and plays (such as Wilkie Collins’s The New Magdalen published in 1873 and set in the Franco-Prussian War)—developed legitimacy, as the expression of civilisation in wartime. These accounts affirmed and asserted the effectiveness of the protective, almost talisman, status of humanitarian aid flags for staff as well as their “beneficiaries”, despite overwhelming evidence to the contrary.

The Dialectic of Security and Protection

This broad historical frame for interpreting the foundations of humanitarian notions of risk and security is intended to show how intrinsically bound these are with origin myth and ideals of humanitarianism. First appearing in the conflicts of the late nineteenth century, the evaluation framework of violence was both statistical and sentimental, enshrining protection and its violation. It relied as much on probability calculations as on emotional responses to danger. The reliance on emblems to afford—and not simply represent—protection lived on in a world dominated by increasingly sophisticated attempts to evaluate and monitor risk and avert danger.

Protecting with Numbers

This statistical intent to quantify risk and exposure along with the scale of violence has been central for modern humanitarians. Current efforts to quantify what is war, using the Uppsala index or various other statistical indicators, do not make much sense in “real-time” and lend themselves to retrospective controversies.See Chapter 1, p. 1.
 The attempts to qualify the Vietnam War or Biafra as genocide illustrated how the two modes of evaluation could correlate and how emotional responses could call on statistical evidence in political discourse. The notion of humanitarian intervention, which, as historians such as Davide Rodogno have shown, has its roots in the engagement of Western powers with specific Christian groups in the Ottoman Empire, was still alive at the onset of World War Two. Davide Rodogno, Against Massacre: Humanitarian Interventions in the Ottoman Empire, 1815-1914, Princeton: Princeton University Press, 2011; Alexis Heraclides and Ada Dialla, Humanitarian Intervention in the Long Nineteenth Century, Manchester: Manchester University Press, 2015.
 The challenge humanitarian intervention poses to the principle of state sovereignty was known as much then as now. The revival of principles and practices after the end of the Cold War renewed utopian dreams that humanitarian aid might be more than an emblem of good intentions and actually deliver effective protection. Of course, the evidence shows that it failed repeatedly, because humanitarians were unable to protect themselves, let alone others. From the failure to maintain “safe havens” during the Bosnian War to the massacre of (national) humanitarian workers in Rwanda, there is no shortage of examples in the 1990s not only of the demonstrable inefficiency of UN protection but also of humanitarian emblems.

If the ability to protect fell far short of the aspirations that were sometimes assigned to aid, security concerns in the meantime grew in reverse proportion. Within organisations, the assimilation of danger to risk, risk management to security, security to safety, collapsed hitherto specific and separate cultural and linguistic categories. Until the 1990s, risk avoidance was perfectly compatible with the notion of exposure to danger. Orbinski, former international president of MSF, relates in his “ego-narrative” a revealing exchange with MSF worker “Joni” in Somalia in 1992 that echoed some of the trade-offs of 1990s humanitarianism. Surrounded by the sound of machine gun fire, and despite the ambient danger, the humanitarian worker argued that there was no risk: “If we get killed, NGOs will withdraw and there will be no one to pay protection racket or wages. They want us scared and alive. So you should be scared and happy because it means you can work.” James Orbinski, Le cauchemar humanitaire, Marne-la-Vallée : Music & Entertainment Books, 2010, p. 99.

Implemented to gauge and assess danger, many of the measures taken since the 1990s have shunned this logic of heroics. The economic trade-off between perpetrators of violence and humanitarian workers remains but cannot be the security approach employers adopt when deploying staff in the field. Insecurity became an inflationary concept, as humanitarian organisations grew and came to realise the full extent of their duties and liabilities as employers. Perhaps necessarily, humanitarians have come to terms with the size of their operations and their bureaucracy has generated human resources policies commensurate with their funding, possibly closing the gap between the fiction of volunteering and their duties as employers. In this workplace as in others, maybe there should be signs stating that “employees have the right to work without threats” and not to be exposed to a “hostile environment”. Perceptions engendered from data have progressively taken precedence over those acquired in the field as risk evaluations become the focus of security analyses. This seems a dangerous drift, however, since it is that fiction of volunteering that was the condition required for “charity under fire” to exist in the first place.

The Field “Under Control”

There is an obvious paradox in security evidence garnered in the field when, arguably, security processes are increasingly dominant in defining the field itself. The late Lisa Smirl argued, when considering the impact of machines and humanitarian spaces, and the role of the ubiquitous 4WD and bunker-like compounds in the shaping of humanitarian perceptions of the world, the field as experienced by humanitarian aid workers is constituted of and mediated through various filters and boundaries. These are not only visible but also implicit. Lisa Smirl, Spaces ofAid: How Cars, Compounds and Hotels Shape Humanitarianism, London: Zed Books, 2015.
 What humanitarians call the “field” is often lived in daylight, limited by strict curfews to what amounts to only half a day in equatorial and tropical countries. As the tinted windows in fast cars provide a filtered perception of the world, the design of humanitarian programmes that include risk assessments from the outset filter humanitarian perceptions of the field. Since the advent in the 1990s of cost-effective rapid communications, the increasingly common deployment of reliable satellite phones and even the Internet, the field has become more elastic as it is shared across continents between headquarter-based programme managers and their security apparatus and the staff on the ground. The issues of remote control and the politics of risk management identified by Mark Duffield have grown from this sense that events can indeed be portrayed as accurately from headquarters as from the actual field—no longer a first-hand experience but one of shared data management. Mark Duffield, “Risk Management and the Fortified Aid Compound: Everyday Life in Post-Interventionary Society,” Journal of Intervention and Statebuilding, vol. 4, no. 4, 2010, pp. 453-474; Mark Duffield, “Challenging Environments: Danger, Resilience and the Aid Industry,” Security Dialogue, vol. 43, no. 5, 2012, pp. 475-492.

Generating data to encode the experience of staff" in the field empowers nobody in particular See Chapter 4, box “Security Incident Narrative Buried in Numbers: The MSF Example”, p. 67.
 but raises new notions of risk and danger at headquarters and country capital levels. The vertical transmission of information from the field to head office is then reciprocated through the sometimes real-time transmission of guidelines and security guidance. It is frequently implemented through the issuing of consolidated, often more extensive, rules complying with the inflationary modus operandi of risk assessments.See Chapter 5, p. 71.

When distance was part of the humanitarian “adventure” and it could take weeks for orders and guidelines to get through, the field was constituted in a very different manner, largely through a sequence of small negotiations. These ensured some degree of security, primarily with more or less reliable gatekeepers and intermediaries who were subject to individual assessments varying from individual to individual. The same setting could be viewed as safe or unsafe by two successive teams, each with their own evaluation of the risks. In reality, these encounters of a profound human nature continue today and so do their differences. Yet gatekeepers and local security brokers have been downgraded as to their importance to security—to trust too much risks becoming a form of negligence for humanitarian organisations and their personnel alike.

This is not entirely new, of course, as the notion of unreliable and dubious intermediaries benefiting from their roles as go-betweens for the rich incomer and hostile locals dates back to colonial times. Crooks and folk heroes, such as Hampate Ba’s character Wangrin, Amadou Hampate Ba, The Fortunes ofWangrin, Bloomington: Indiana University Press, 1999; Ralph A. Austen, “Colonialism from the Middle: African Clerks as Historical Actors and Discursive Subjects”, History in Africa, no. 38, 2011, pp. 21-33.
 have been the focus of many books exploring the role of intermediaries in the colonial setting. Humanitarians directly inherit rules from the colonial and imperial era when they employ staff whose role is to be on-going intermediaries for transient and profoundly ignorant international staff. Ann Laura Stoler, “Rethinking Colonial Categories: European Communities and the Boundaries of Rule”, Comparative Studies in Society and History, vol. 31, no. 1, 1989, pp. 134-161; Benjamin Nicholas Lawrance et al. (eds), Intermediaries, Interpreters, and Clerks: African Employees in the Making of Colonial Africa, Madison: University of Wisconsin Press, 2006.Risk-evaluations are all inclusive and potentially destructive of the notion of trust upon which many humanitarian encounters rely. They do not allow broad spaces for negotiation and, through their transposition from one location to another, the filling in of risk assessment forms demands an ever-higher degree of pre-conceived suspicion.

Of course, the dramatic evolution described here is historically contingent. It reflects other forms of power concentration and the tightening of chains of command supported by technological innovation. On one level, it is rooted in the notion that responsibility should always imply control and that control is rooted in management. It also implies that “duty of care”, as enshrined in employment legislation, confers overwhelming decisional rights and that, conversely, the individual is neither responsible nor careful. Arguably, this loss of autonomy is not restricted to the humanitarian sector as it is widely prevalent in Western societies. It relies on fictions of control and accountability as heroic as yesteryear’s fictions of daredevilry with their assumptions of frictionless mechanisms and fool proof guidance. Humanitarian staff create solidarity in their work in the field through small acts of resistance to the smooth machinery of securitisation, and disobedience frequently becomes a badge of honour. Small transgressions become the stuff of the field as they compensate to some degree for the curtailing of operational and personal space while exposing daily the vacuity of excessive narratives of danger.

In the humanitarian context, extreme securitisation rooted in worst-case scenario and all-eventuality training comforts critiques of humanitarian aid, more often than not originating from within the aid sector itself, which, since the 1980s, have accused NGOs of being self-serving and self-obsessed. Much of the humanitarian literary output of a critical nature explored by Lisa Smirl and others in books such as Emergency Sex and Other Desperate Measures examined in a contradictory manner how NGOs harbour irresponsible and infantilised staff. Kenneth Cain et al., Emergency Sex and Other Desperate Measures, London: Ebury Press, 2005.
 Within this discourse, humanitarians are over protected and yet, due to their powerlessness and lack of “genuine” expertise, they are exposed to the traumatic consequences of humanitarian aid. This analysis is blatantly circular and indeed narcissistic, but it flows in a continuous stream of self-doubt, relentlessly eroding the legitimacy of humanitarian action.

Far from being an additional marker of increased professionalisation in the field, the progressively dominant approach to security and securitisation in the field amounts to a gradual erosion of the humanitarian worker’s sense of agency and responsibility. It also signals the recognition that humanitarian NGOs, similar to large multinational organisations, do not feel sufficiently confident to leave crucial negotiations to their field staff. As a hierarchy carries the burden of responsibility and duty of care, it seems to say, the freedom of distance and autonomy cannot be left to lesser levels in the organisation.

Conclusion

Since 1864, modern humanitarian aid has consistently generated and responded to evaluations of need and risk and this work has been framed by exposure to danger and security concerns. A number of assumptions regarding risk, volunteering and danger have enabled humanitarian workers and NGOs to function in the face of danger. Its emblems embodied this will to work under the protection of the law but they seldom provided protection alone. What enabled humanitarians to work in acutely violent places was a combination of myths about courage, character and adventure, associated with negotiations at the bedrock of very careful risk management. The shift to a “post-heroic” age has profoundly altered how humanitarians relate to each other. The balance between individual agency and collective responsibility, between volunteering for and exposing staff to danger, was always delicate. It tipped toward responsibility and duty of care with the advent of bureaucratic decision processes (rather than the bureaucracies themselves, which have always been a necessary and valuable component of humanitarian aid organisations).

Arguably, bureaucratic decision processes have depended on the ability to translate field concerns into security guidelines, which emerged early on in some organisations and somewhat belatedly in others. The culture specific to each NGO would undoubtedly nuance the broad-brush argument presented here; Michael Neuman’s chapter charters much more precisely the evolution of risk perception and security management within MSF.

In the broad historical perspective, documenting violence affecting humanitarians was framed in the heroic logic of volunteering and possible sacrifice. It invited the ghostly presence of international humanitarian law, but this legal framework provided no more safeguards or guarantees than the adopting of a heroic logic. In a world that has applied the logic of humanitarianism to justify interventions—military and diplomatic—since at least 1860,51. Lebanon intervention by the Franco-British fleet in response to Druze massacres of Maronite Lebanese. Istvan Pogany, “Humanitarian Intervention in International Law: The French Intervention in Syria Re-Examined”, International and Compa¬rative Law Quarterly, vol. 35, no. 1, 1986, pp. 182-190.humanitarians have remained bound to the negotiations that have enabled them to define the duration and nature of their stay or their “terrain”, now called “the field”. The imagining of a global order based on protection bypassed these concerns to foster a focus on security processes, a gathering of data which did not necessarily feed into any planning but rather into even more gathering of data.See Chapter 4, p. 55.This securitisation based on numbers and guidelines prescribes behaviours that have always been part of the landscape of humanitarian aid and which appear to have taken on an existence all of their own.