By 1997, it could be said that MSF had seen it all – aid diverted and used to murder people, genocide, manipulation of the term ‘humanitarian’, military interventions allowing crime to flourish, etc. The organisation had been confronted with the most extreme issues linked to violence and war. Experiences such as powerlessness, the call for military intervention, being used as bait and incantation towards the international community all challenged the conception MSF had of its role in such situations. After 1997, MSF’s tone became progressively more reserved, hence the ‘era of disillusionment’ during which the figure of the aid worker gradually replaced the figure of the doctor-witness and defender of populations in danger.

These figures are, of course, broad caricatures. It is obvious that they did not emerge with absolute clarity as one period gave way to another, that the caesuras were not so sharply defined. These should not be taken as pure substratum from which the attributes of other figures are absent. They are employed to indicate tendencies, as are the ‘eras’ (whether of ‘appeals’ or of ‘disillusionment’).

“OUR ROLE IS TO PROVIDE QUALITY AID”

The shift towards the figure of the aid worker stemmed from the return to our core responsibility, the basis for our action – the provision of aid. As we have seen, the organization had been moving in this direction since the mid-1990s – but the process gradually acquired a more formal basis as the new millennium approached, notably through the emphasis on the quality and effectiveness of aid, a development that concerns us here inasmuch as it stimulated revisions to the discourse surrounding responsibility and violence.

The expansion of the medical component of our missions and the improvement of its quality, the concern for operations to be judged according to their quality, were consistently put forward from 1999 onwards (Board meeting, project discussion, January 1999; President’s annual report 2001-02, 2002-03, etc., until the notion of ‘quality of care’ was introduced into the draft statutes of the association in 2005). In parallel, the desire for a “stronger involvement in situations of crisis and armed conflict” was stated (P. Biberson, AR 1999-2000).

As J-H Bradol confirmed, “we wanted to refocus MSF action on the provision of aid in crisis situations” (interview). It is interesting to note that the emphasis on quality came about precisely in instances where the existence of wide responsibilities beyond care were being questioned, as illustrated by the following comments regarding international justice: “Our role is not to foster its development. Our role is to provide quality aid … the clarification of our role, our responsibilities and the limits of humanitarian action … in armed conflicts is essential if we are to improve the aid we deliver” (AR 2000-01). As a consequence of the desire to refocus operations on crises and improve quality, the requirement to be “as close as possible” to violence – not an entirely new development – received particular emphasis from 2002-03 onwards. In the DRC, for example, efforts were made to redeploy operations to the areas where “the most severe incidents occur” (AR 2002- 03), and to improve mobility and response times, all of which resulted in the positive observation that after “years of real difficulties and failures, the opening and stabilisation of missions in violent areas has become a reality” (Board meeting, September 2003); such efforts are still on the agenda today. Wherever security conditions permit, teams will always attempt to achieve the greatest possible proximity. This can be felt in the internal discourse, which no longer refers to the value of témoignage or the solidarity that being present represents. Meanwhile “alibi missions”, the idea of “being there for the sake of being there”, have been demolished.

VIOLENCE AS A MEDICAL ISSUE

In this context, the primary responsibility when faced with violence is to treat its victims. The discourse reveals the emergence of violence as a medical issue and a matter of quality, and not just as a context or phenomenon triggering our responsibilities as humanitarian actors“Wherever you work, you have to succeed. That’s the real imperative. Look at war and what counts as success in war. We were talking about caring for victims of violence: we’ve been around for 35 years, it’s only five years since we started taking a bit of interest in the actual victim of violence!” (interview with J.-H. Bradol).. Responding to violence meant possessing and increasing the means to care for its victims, requirements that led to other major operational developments, notably the treatment of warwounded – “in the years to come, we must all focus on developing this essential component [surgery] of our operations in the field.” (AR 2001-02) – and victims of sexual violence, which was accepted as a specific care aim in late 2003, following a pilot scheme in Congo-Brazzaville in 2000. Existing missions were revised to focus more closely on victims of violence (see the DRC case study). Two missions explicitly designed to work with victims of violence were launched in 2005, one in Haiti, where “serious violence has led us to believe that it is a civil war in all but name” (Board meeting January 2005), and the other in Nigeria. The Nigerian mission envisaged setting up a “health centre for victims of violence in Port Harcourt”, the site of a “low-intensity conflict” (Board meeting, September 2005). The word “violence” was introduced into the draft statutes alongside “natural and man-made disasters” and “belligerence” to qualify the target of our mission (Board meeting, November 2005). In accordance with this operational bias, violence and the treatment of its victim played an increasingly large part in discussions regarding a project’s relevance (e.g. at operations meetings covering the reopening of missions in Sri Lanka, the closure of Ankoro in the DRC, the opening of Buenaventura in Colombia, the opening of Kurdistan, etc.). One significant and parallel development was the use of an “epidemiology of violence”. Resulting from the demand for descriptions couched in medical terminology and supported by statistics, the objective language of epidemiology had initially accompanied the “collection of testimonies”, but later largely replaced it as the legitimate method (both internally and externally) for documenting specific incidents of violence.

In this respect, the evolution of MSF mirrored the broader evolution taking place in political, academic and media circles: the word ‘violence’ was ever more present when describing the world, while the word ‘war’ was becoming increasingly inadequate and could not account for the phenomena that affected many societiesSee P. Hassner, ‘De guerre et paix à violence et intervention’ in J. Moore (ed.), Hard choices . Moral Dilemmas In Humanitarian Intervention, Lanham, Md.: Rowman & Littlefield, 1998. See also Frédéric Gros, Etats de violence. Essai sur la fin de la guerre, Paris, Gallimard, 2006(just as the ‘war - population displacement - refugee camps’ schema was increasingly unable to summarise MSF action in unstable contexts). Even so, the tendency to reformulate violence in terms of war (the “low intensity” conflict in Port Harcourt; the civil war “in all but name” in Haiti) is clear, indicating that violence was still an inconvenient and conceptually difficult category for head office and field teams alike.

THE RETURN OF INTERVENTIONISM AND DEFENCE OF THE FIELD

The “refocusing” on the quality of aid and on violence as a medical issue meant that MSF would in most cases express the fact of violence through medical terminology. In correlation, it gradually distanced itself from challenging political actors, particularly in the most extreme form – the call for “military intervention”, but also the more moderate form – the call to respect civilians.

As we have seen, past experience had already convinced MSF of the variable (and sometimes volatile) impact of its calls to protect civilians. Kosovo provided the first example of an intervention mounted for the very purpose of protecting civilians. Whereas the “humanitarian” interventions of the early 1990s had been concerned with aid delivery, the “military-humanitarian” operation in Kosovo was based on “protecting civilians”: the specific justification for “humanitarian war” (the term used by Tony Blair) was that massive violations of human rights were occurring and had to be stopped. An international intervention of this kind was in a way what MSF had repeatedly called for in Bosnia, Rwanda and Zaire, by issuing (explicit or implicit) appeals for the “protection of civilians”: the international community was finally concerned with more than bags of rice; i.e with “protection”. But as this concern was presented again as “humanitarian”, MSF saw in it the latest manifestation of the appropriation of this field by politicians and felt it necessary to set itself apart: “It has not been easy in this case to determine our position in relation to the war (which has never been described as such … the ‘strikes’) and which is being presented as a humanitarian cause (saving the population of Kosovo). It has been, and remains, highly uncomfortable to be so closely associated with it” (Board meeting, April 1999). As in the Kurdistan case in 1991, the latest slant on the term protection could not deflect internal and external criticism of the new interventionist policy (made much easier by the fact that medical needs were hard to determine) on the specific grounds that there was in fact a “lack of protection” – lack of protection in the legal sense for refugees in neighbouring countries (registration, etc.) or in the camps run by NATO forces, which prompted MSF’s denunciation of UNHCR for its acceptance to take a back seat; lack of physical protection for those returning to Kosovo, for “the NATO presence there is not up to the job of protecting civilians” (Board meeting, June 1999). The criticism was similar in tone to that adopted at the beginning of the western interventions in Kurdistan and Somalia, and indicates the serious concern over the confusion of genres. So there was nothing radically new here. It simply remains to be said that as these events unfolded, MSF continued to use the word “protection” in a relatively straightforward manner; states were misusing it to justify military intervention, but it had not been completely tainted and could still be employed in reference to ‘appropriate’ fields such as the law and security.

MSF AND THE “RESPONSIBILITY TO PROTECT”

During this period, the foregrounding of “civilian protection” as a responsibility incumbent on the “international community” resulted in significant changes. Eager to learn the lessons of the many “failures to protect” that had marked the 1990s, the UN’s new secretary general promoted an integrated vision in which the “responsibility to protect” was but one element in a broad spectrum of responsibilities. In this view, the quest of the international community for global peace and security would include conflict resolution, humanitarian action, military operations to ensure security, international legal action, the reconstruction of states and Third World development. The Brahimi Report on peacekeeping operations (2000) represented one stage in the reshaping of the United Nations, and was swiftly followed by a document entitled The Responsibility to Protect, which formalised the turn to interventionismThe Responsibility to Protect, ICISS (International Commission on Intervention and State Sovereignty), 2001. The report was commissioned by the Canadian government following the UN secretary general’s call for the international community “to arrive at a definitive consensus on fundamental issues of principle and procedure” with regard to the “right of humanitarian intervention” (meaning ‘military-humanitarian’). See www.iciss.ca. See also E. Tronc, Humanitarian protection, MSF International, 2007 (draft).. The attacks of September 11 followed hard on its heels and further altered the international landscape by putting security and the fight against terrorism at the top of the West’s agenda. As a consequence, Afghanistan and Iraq experienced a new (but familiar) type of intervention which still managed to employ elements of the humanitarian registerFor a striking account of the evolutions of interventionism in the light of the evolution of the figure of the American president, see M. Feher, ‘Autoportraits présidentiels’, Vacarme no. 30, winter 2005. The author shows how changes in US foreign policy restricted its detractors to symmetric reversals. This echoes the idea, developed by the author elsewhere, of inversions of the polarities underlying the positioning of MSF. In his view, the abdication taboo arising from the discourse of the international community in the 1990s was supplanted by the intervention taboo as a result of the new interventionist discourse adopted by both the Bush administration and the UN.. MSF’s desire to distance itself from “protection” at this time should be seen in the light of the resurgence of interventionism, although as it turned out the actions undertaken “in the name of protection” in Kosovo (a war that did not involve the UN) and Timor (a commitment which the case of Chechnya quickly negated) had already been supplanted by those of the United States as it pursued its ‘war on terror’. Even before the attacks of 11 September 2001, internal references to international interventionism with the aim of protection were negative:

“I note once again that whenever it is a matter of protecting civilians, a major international mobilisation achieves nothing” (Board meeting, March 2001). On international responses to crises: “The protection provided by states and the United Nations is either feeble or nonexistent.” The safe areas “conceived as an alternative to population displacements are illusory in terms of protection” (AR 2000-01).

As we can note in this quote, the term was tainted by the shadow of failure – a qualification whereby MSF members were encouraged to stop referring to it as a possible or desirable recourse, for it would be “hardly realistic”. In 2003 the organization formally decided that it should play no part in calls for “military intervention”, following the chaos in Liberia and the massacres in eastern DRC (Ituri) which had aroused the “temptation” to do so. In the Liberian case, the option was discussed and discarded – the appeal for “all possible means to assist civilians” came close, but did not cross the line (press release 26 May). “I think it was appropriate to refrain from calling for a military intervention,” commented J-H Bradol (Board meeting, August 2003). This position was later extended: “It is not the role of a humanitarian organization to call for the use of force in order to ensure that populations receive the appropriate aid” (AR 2003-04). With regard to the DRC, MSF published a highly critical document on the minimal “protection” provided by the international intervention in Ituri province (the MONUC mission, followed by Operation Artemis, which had a specific mandate to protect populations). The report (Ituri: Unkept Promises?) was disowned after its release for its negative conclusion and unrealistic demands. What it required of the international community was ambiguous, to say the least: if it ventured towards an appeal for better protection, it was only so that it could then more easily express its central concern, the need to guard against “false promises” and the “false illusion of protection” (sic), thus inferring that the problem was not so much the lack of security as the lack of truth. The report signalled the increasing sense of unease arising from stances on protection, and precipitated the adoption of an institutional position on the issue.

“Could an international military force pacify Ituri and guarantee security throughout the province? I doubt it, and we should not allow that impression to be given” (Board meeting, August 2003). “In our discussions of conflicts we have seen the resurgence of calls for protection which, in my view, are hardly realistic. When MSF calls for an international military intervention to pacify the entire province, it seems totally unrealistic. We are not capable of deploying coherent medical aid to every corner of Ituri province. So why should we imagine, given the complexity of the situation, that it would be feasible for foreign troops to turn up and immediately resolve, as if by a miracle, all the problems of violence, all the problems of ensuring that civilian populations have access to aid? We should not become the propagandists of such illusions …” (AR 2003-04).

In short, from a critique of the illusion of protection fostered by an international community that had little desire to act, accompanied by calls for protection (in the 1990s), we turned to a critique of the illusions that some of our own personnel maintained on the subject of protection. The Ituri report represents a kind of pivot between these two positions. In late 2003, the call for armed intervention was unequivocally described as a “temptation” that should be resisted except in the extreme case of genocide, thus ratifying both the history of MSF and its new positionSee F. Weissman, ‘Humanitarian Action and Military Intervention: Temptations and Possibilities’, 2003 (available on the MSF website, also published in Disasters, Vol. 28 Issue 2, June 2004, pp. 205-215). The evolution of public positions on violence in general reflects the gradual retreat from the call for intervention (and therefore from the discourse of protection), with 2003 marking the turning point.

FROM TÉMOIGNAGE TO ‘SPEAKING OUT’

As we have seen, doubts over the evident relevance of témoignage had run parallel with the practice itself. Following the intense inter-sectional arguments over the plight of the Rwandan refugees in eastern Zaire and the many meetings during which the place of témoignage was reaffirmed, MSF-France continued to question its possible deviations. Témoignage was “one attempt among others to create access, to open up a humanitarian space… to ‘protect’ (I use the word with great caution),” but “speaking out can push us into a kind of automatic militancy” (AR 1999-2000). In all likelihood, the very success and spread of the idea, together with the rise of human rights NGOs that began in the early 1990s, made this line of questioning seem particularly urgent: “Almost the entire ‘humanitarian family’ (which is not a family at all!) now lays claim to témoignage! … Let us beware of consecrations …” (idem)There seems to be an irrepressible reflex on the part of MSF (particularly the French section) to distance itself from any idea that is likely to achieve widespread acceptance (in the humanitarian ‘family’ or in the public domain). It is certainly evident in this instance, as it is in more general discussions of the concept of protection. “The problem is that I can’t bear being among the majority,” said R. Brauman in an interview with M. Potte-Bonneville and S. Grelet (‘Qu’est-ce qu’on fait là?’ Vacarme no. 4-5, summer 1997). “Let us remain troublemakers. It’s an important moral goal,” he remarked in his 1992-93 annual report. This personal trait is reflected at institutional level; there is a kind of ‘ethos of difference’ running through the organization’s history, which all MSF sections refer to – a situation which MSF-France must find annoying, because it then feels compelled to distance itself from this claimed difference.. Thereafter attempts to define its role continued; the President warned against becoming an “organization which awards itself a license to adopt positions on every topic in the name of some hypothetical universal moral doctrine” (AR 2000-01). “It is not for us to express an opinion on whether wars are justified or not, but we have a word to say on the way they are conducted: that is where the humanitarian’s responsibility lies” (AR 2001-2002).

Given these frameworks, which themselves reflected the need to find a voice between militaryhumanitarian confusion and anti-imperialist critiques (from Kosovo to Afghanistan), the public stances on violence in war adopted during the years up to 2003 reveal a focus on international humanitarian law and the rights of refugees as the legitimate basis. In situations involving refugees, references to this body of law were resonant and specific, particularly when highlighting the principle of voluntary return, and were in keeping with the positions adopted in the past. When applied to refugees, the word “protection” was acceptable; unlike “protection of civilians” it was not a suspect term, for it had not been distorted into a synonym for military intervention. In the case of Chechen, Korean and Liberian refugees, MSF repeatedly denounced violations of the “principle of voluntary return” and the “right of asylum”, and issued many calls for protection. When the scandal over sexual exploitation in West African refugee camps broke out, it was recalled internally that MSF had “banged on about the lack of protection for refugees in these three countries (three press releases)” (Board meeting, March 2002). When not about refugees, public discourse on violence was dominated by references to “civilians” (a category whose growth we have already discussed). Until 2003, descriptions of the violence inflicted on civilians were often accompanied by calls for the “parties to the conflict” or “belligerents” to respect their rights and stop targeting civilian populations (Congo-Brazzaville, Chechnya, Ivory Coast, Liberia, DRC, etc.). The violence committed by western armies in their “war on terror” led to explicit references to the “Geneva Conventions” (press releases, December 2001; October 2003).

A marked shift occurred at the end of 2003, fostered by the desire to get a grip on the considerable increase in external communicationsFor example, in October 2003 the Board meetings started including a “public statements” section to scrutinise the content of communications. A similar heading appeared in president’s annual reports from May 2004.. As a review of the press releases clearly indicates, the number of those relating to violent situations fell drasticallyThere were over 30 press releases relating to violent situations in 2003, 15 in 2004, less than ten in 2005 and about ten in 2006, while those relating to medical issues formed the broad majority (20 in 2005) then decreased (ten in 2006).. While the “civilian” category continued to appear in press releases to signal the existence of violence, there were very few references to international humanitarian law or appeals to “respect civilians”.

In short, our external discourse began to centre on the responsibilities of the aid worker in relation to victims of violence; descriptions of the effects of violence on human health were thus increasingly linked to the issues surrounding aid provision. Almost all the public statements in which violence is mentioned emphasize problems such as obstacles to aid, its inadequacy, the lack of access (for humanitarian workers to populations and vice versa), the dangers of the working environment (with a particular emphasis on the safety of teams) and the manipulation of the term ‘humanitarian’. Given their interconnection, a single statement would often mention several of these issues. In short, these statements all ask for something directly related to our activities – a secure space, an improvement of some kind, or perhaps access to a previously closed area – whether in Sri Lanka (obstacles to access; violence against humanitarian workers), Darfur (inadequate aid in 2004; access and lack of security for humanitarian workers in 2005), Uganda, Somalia or Haiti (insecurity which made it difficult for civilians to access MSF care). Only in a few events which we considered particularly severe (the bombing of civilians in Chad and Ivory Coast; forced repatriation) did we issue press release in the absence of threats on our own aid action.

The public position on the violence in Darfur (The Worst is Yet to Come, June 2004, issued in the wake of press releases which had focused on the “needs” of “displaced persons” rather than on violence) provided an opportunity to clarify the shift on communications. The president of MSF explained his view of it as follows:

[On the regretted decision not to issue a statement at the end of 2003]: “In order to increase the flow of aid to Darfur, it was absolutely necessary to alert the public to the gravity of the crisis”; [On the adoption of a position in June 2004]: “I wanted to review our public positionings on Darfur. I talked just now about the usefulness of communicating the gravity of the events. It is used to call for more aid. We do not see this as a responsibility we have to act as some sort of human rights observer in this mission. On the other hand, our responsibility in the sphere of aid is clear” (AR 2004-05).

The president thus disputed the idea that statements on violence stemmed from any ‘intrinsic’ responsibility, suggesting that the basis for them was the need to ensure, improve and strengthen aid delivery. In short, it was a matter of “preserving or restoring conditions compatible with the accomplishment of [our] task”M. Feher, ‘Constance déroutante’, Vacarme no. 31, spring 2005. The author argues that the apparently contradictory positions on Darfur obscured MSF’s doctrinal consistency: the emphasis on action and care as the “only raison d’etre”. He therefore maintains that the organization’s public positions are necessarily motivated by whatever poses a threat to the action – obstacles, corruption (diversion) and appropriation (by western powers) are thus the three motives underlying a statement. While we agree on the current primacy of action, it seems that Feher’s argument, although cogent, overlooks the complexity of the motives linked to violence, a factor we have attempted to illuminate in the course of the present study.. The anchorage, or centre of gravity, had shifted. In the 1970s and 1980s, attention and responsibilities had been outwardly directed; there was an explicit relationship between the humanitarian actor and the human rights violations that occurred in “theatres of war”. By 1985, it had become necessary to look also inward, to answer for our own actions. The external and internal approaches co-existed throughout the 1990s. Gradually, the latter approach, the need to examine our own action and defend ourselves against manipulation, provided the hinge between MSF and its working environment. Rather than roles or responsibilities in the face of violence, it was a matter of the responsibilities applying to our own action.

As a consequence, for many of those who identify with the figure of the aid worker, the idea that there is little legitimacy in investing in issues not directly linked to aid delivery has a certain resonance (given that “we don’t do protection”). Others insist that MSF is a humanitarian as much as a medical organization and thus has broader responsibilities. In a forum such as the Projects Week, it was possible to find a range of positions, all reflecting the uncertainty as to the ‘correct limit’; questions along the lines of “What should we do?” and “How far should we go, given that we have no mandate for protection?” were not uncommon.

The evolution of MSF, a product of the sedimentation of past experience, the effects of our growth, and external determinations (the development of state international action, the proliferation of various forms of violence, etc.) is clearly expressed in the La Mancha agreement La Mancha, final agreement, June 2006. Available on the MSF website.. The document stipulates that the denunciation of violence should be regarded as an exceptional measure (justifiable only when its perpetration is massive and ignored), and states that the promotion of rights is not in itself one of the aims of the institution (in the sense that, unlike certain other organizations, we don’t “do protection”). It also states that MSF cannot claim to provide populations with physical protection. External limits and internal requirements establish a common foundation from which all sections can proceed.

From the sentinel of the 1970s to the aid worker of today, the repositioning on a carefully delimited field is quite clear. In 1985, the “we protect people” stance (part assertion, part ambition) typical of the early 1980s gave way to the qualms that MSF continues to harbour; they are as pronounced in the actions currently underway as they were throughout the violence and appeals to political actors that characterised the 1990s. However, the acceptance of limits springing from the “loss of illusions”, on the capacity of our presence to “make a significant contribution to the protection of populations” (AR 2005-06, referring to La Mancha), not to mention the international community’s inability to provide protection, does not mean that the question is closed for MSF. This will become apparent as we examine some of the permanent features of the concrete practices deployed in situations of violence.