“We know that at MSF the best way to bury a subject is not, as in politics, to set up a commission – because a commission is a functioning body – but to write a report on it. If you want to bury a subject, write a thirty-page report on it.Then you can be absolutely certain that it will be deliberately ignored for the next ten years.”

(President’s Annual Report, 1991-92)

So here is a report on protection.

To embark upon a study of this theme is to enter a field strewn with contradictory representations linked to a highly sensitive issue – the limits of our responsibility – that has generated endless disagreements and debates on our “identity” and the existence or non- existence of a role for MSF “beyond care”. It also confronts us with the reactions – sometimes ironical or sceptical – of those who claim that we “don’t do protection”, while others express surprise and see no cause for argument, or lament the fact that MSF no longer “bears witness”. Whatever the case, it is clear that a need to explore this field is now emerging, and it is significant that it is doing so at the very time when protection has become omnipresent in the discourse surrounding our action – whether in calls for the “protection of civilians” in Darfur or in the growing number of humanitarian organizations which claim to be “doing protection”. In internal discussions at MSF, as in the external sphere, references to the “failures of protection” in Bosnia and Rwanda and to the responsibility to protect (either desired or rejected) are vivid.

So what is the issue and who does it involve? When exploring the subject, our first task is to strip away the ‘drama’In this document, double inverted commas (“ ”) indicate a quotation or a commonly used expression within MSF; single inverted commas (‘ ’) are used by the author. with which it is encumbered, to free it from the intense pressure and confusion that characterise its discussion, particularly within MSF. But it seems that approaching the question of MSF’s responsibilities with regard to protection directly would lead us head on into this confusion. In effect, the word has many meanings and is employed in a variety of registers which enable its constant mobilisation, although we are never quite sure what we are talking about.

Let us begin with the registers. In the first instance, the word forms part of the phrase “protection of civilians”. In international humanitarian law (IHL) this refers to a specific legal framework, a set of rules designed to restrict and attenuate the effects of conflicts (international or otherwise) on a certain group of people (non-combatants) according to the nature and degree of their vulnerability. International humanitarian law therefore discriminates; it creates categories such as wounded combatants, prisoners, civilians in general, the sick, women, children, refugees and those living in occupied territories. All such categories represent “protected persons”, people who are entitled to protection, i.e. to the application of specific “protection regimes”. To be more precise, protection is most often a matter of prohibitions designed to regulate the conduct of the parties to the conflict: parties should not attack civilians, use famine as a weapon of war, resort to torture, take hostages or subject people to degrading treatment, etc. More positive measures include the provision of safe sites through the creation of neutral zones on territory controlled by one of the parties, the evacuation of a besieged zone and, above all, the right to obtain relief. Consequently, the warring parties should not attack civilian hospitals nor hospital staff and members of relief organizations, but they are also obliged to guarantee the free passage of supplies intended for populations (medicines, healthcare equipment and provisions). These various elements highlight the fact that ‘legal’ protection is a broad framework in which relief is just one right among other.For further information, see the texts that created this right (the 1949 Geneva Conventions, particularly the Fourth Convention Relative to the Protection of Civilian Persons in Time of War and the Additional Protocols of 1977), available on the ICRC website (www.icrc.org). The thematic presentation of the Geneva Conventions and their Additional Protocols provides a clearer summary (see “Protection of Civilian Persons and Populations in Time of War”, www.icrc.org/Web/Eng/siteengo.nsf/htmll/genevaconventions). See also F. Bouchet-Saulnier, The Practical Guide To Humanitarian Law, MSF/Rowman & Littlefield, new edition 2007, particularly ‘Civilians’, p. 36-40, ‘Protection’, p. 348- 352 and ‘ ‘Relief’, p. 373-378. However, the original meaning is altered in practice. The ICRC – the only agency with a mandate, i.e. with a specific role under IHL – has a “protection” (“prot’”) department which groups together activities aimed at certain specific categories of protected persons such as detainees and separated families. This name which stems from the institutional structuring of the ICRC, sustains the idea of a clear separation between protection on the one hand and operations on the other (operations being another department – in which, incidentally, activities other than assistance are called “protection” e.g. interceding with the parties to the conflict in an attempt to ensure respect for civilians).

The protection of refugees is another register deriving from IHL (the 1951 Convention relating to the Status of Refugees). It can be distinguished from the first because it has contributed even more to the formalisation of a ‘protection-assistance’ duo. Each term of this duo covers a set     of concrete measures (registration, camp security and the defence of  the  nonrefoulement  [forced return] principle in the first instance, and aid delivery in the second). In this sense, the practice of both UNHCR and the ICRC tolerates an ambiguity between protection as a legal framework and protection as a set of concrete activities.

We now come to the register used by humanitarian agencies which do not have a specific mandate; while related to the legal register, it refers above all to a professional field. This covers activities ranging from the documentation of violence to public denunciation and has led to the creation of “protection officer” posts within NGOs, while the UN has set up “protection clusters” to coordinate “protection activities”. No such measures have been put in place at MSF-France, but the word protection is in fact spontaneously associated with the concept of témoignage (thus suggesting that care-témoignage is MSF’s term-by-term equivalent of the assistance-protection combination).

Finally, we arrive at the political register of international relations, the talks between states and at the UN, where for several years the topic of “protection” (meaning, of civilians) has referred to the responsibility of the “international community” to intervene in a state’s affairs to prevent or stop violations of human rights. A responsibility of this nature includes the possibility of military intervention as indicated by the ambiguity of the term “humanitarian intervention” in English, which (unlike in French) means ‘military intervention with the aim of protection’.

The plurality of registers is complicated by the polysemous nature of the word protection in its most common application. By and large, it refers implicitly to the concept of physical security, but it may also be employed in a broader social context to designate numerous other fields (welfare and legal protection). But of greater significance is its ability to encompass both an action (protecting someone) and the condition resulting from that action (security).

The term therefore has a variety of meanings which may be intermingled or sometimes closely associated in a single sentence. For example, its polysemous quality may convey the impression that activities conducted in the name of protection (as when agencies talk of “doing protection”) almost automatically result in the provision of protection (as a state of security) when this result has not materialised, we are confronted with the ‘paradox of protection which does not protect’. Moreover, since we are dealing with violence we find that the connection between physical and legal protection is always somewhat ambiguous: in many cases, protection of civilians in the legal sense (their right to be spared) actually refers to their physical safety, which means that NGOs must tread very carefully when calling on the international community to “provide protection”. Finally, the separation between the register employed by humanitarian agencies and that used by the international community is by no means as clear as the above description suggests. The recent UN “coherence” doctrine depicts the “protection of civilians” as a common goal, i.e. a goal shared by every actor working in these fields; the activities of NGOs are therefore simply one aspect of the shared “responsibility to protect”,The “responsibility to protect” doctrine emerged in the UN sphere in the late 1990s and was formalised in the eponymous document published in 2001. The doctrine is based on the idea that states have a responsibility to intervene when a state is committing “violations of human rights on a massive scale” or pursuing a policy of “ethnic cleansing”. which is itself just one element in the doctrine’s objective of global security in the broadest sense of the term (economic, cultural, social, physical, etc.). The doctrine effectively ensures a state of permanent confusion regarding the meaning of the word protection and of the word security.The rise of the term “human security” has been as spectacular as that of the term protection. Stretching the concept of security to its limits through an integrated global approach, “human security” postulates that states should accept human life as their chief concern, rather than national security. From this perspective, human security is as much about alleviating poverty as it is about using military means to pacify unstable areas. At a certain level, everything becomes interchangeable. For an analysis of the concept’s malleability, see ‘What is Human Security?’ in Security Dialogue, no. 35 (5), 2004, pp. 345-387.

This exploration gives us some idea of the complexity of a portmanteau-word that has been subjected to various forms of abuse, distortion and manipulation by a multitude of agents. However, this observation is not in itself enough to justify banishing the word from the present study: it could indeed have led us to try and untangle the threads in the hope of arriving at the ‘right definition’. It is in fact the very status of the word within MSF which poses a problem. Besides reflecting the complexity of its usage by other actors, we now use it sparingly and  generally in a negative sense, which indicates our ambivalence towards the concept’s connotations.See Appendix 4 for instances of the word “protection” in Board meetings proceedings and President’s annual reports since 1978 (the list is based on notes taken from these documents and is therefore not exhaustive, but it is nevertheless useful for indicating tendencies).In MSF documents the word ‘protection’, whether referring to the activities of other agencies, the performance of international armed forces or the role of MSF, is often stamped with the seal of falsehood or impossibility.

So if we confine the discussion to the uses of the word, we dispose of the question of whether MSF has a role that extends “beyond care” in war and/or violence by simply not asking it. We have therefore opted for an approach that circumvents the word: we focus on the responsibilities MSF accepts and enacts when faced with the violence affecting populations. In short, we attempt no constructive and premature definition of what protection is or is not, but instead examine the discourse and action arising from various situations – we look back over the history of MSF and see how the organization has expressed its role in these situations, the dilemmas it has faced and the responses it has applied. In the background comes a series of questions. How does MSF’s formulation of its responsibility fit into its working environment? How has past experience influenced this formulation (our view of the limits of what we can, should or should not do when confronted with violence)? How is the responsibility to do something, to act, transformed into action: what threats do we refer to and exactly who is under threat? How do we define the target and content of the action we take in such situations? To which extent is action motivated (at institutional level and among individuals with different levels of responsibility) by ‘morality’ or by the importance of effective aid provision? At what point does violence, initially taken as the working context, become a phenomenon which we attempt to influence (by stopping it, slowing it down or attenuating it)? In opting for this approach, we are fully aware that replacing the word ‘protection’ with the expression ‘faced with violence’ does nothing to reduce the difficulty, for violence itself is a fluid category, malleable and ideologically loaded. However, the shift to another term means that the context (violence), rather than the nature of the action, becomes the fundamental point of reference. We can thus hope to avoid the twin reefs of permanent semantic uncertainty constituted by the various forms of protection and of a pre-established view of ‘what protection is about’. Moreover, each time the word protection occurs, we have tried to clarify the particular register and meaning involved.

As these questions suggest, a description is first needed. Bearing in mind the need to look at concrete practices in the field (systematised or otherwise, aimed at individuals or at groups) as well as discourse, we decided to focus on certain situations which could serve as specific case studies. The first of these is drawn from the past – the hunting down of Rwandan refugees in Zaire in 1996-97. This particularly tragic crisis created numerous dilemmas and enables us to gain a temporal perspective. We then turn to two contemporary situations. Darfur (Sudan) is appropriate because of the gravity of the crisis and its prominence in international debates on protection. North Kivu (DRC) was selected because the programme MSF is running in the area symbolises the organization’s increasing focus on direct victims of violence; it also adds to the list of issues raised by traditional situations involving the treatment of indirect consequences of violence (massive population displacements, the collapse of care provision, etc.). ​​​

The three studies (reproduced in full as appendices) provided the foundation for a broader attempt to identify the evolutions of MSF practice and discourse in the face of violence, beginning with a review of internal policy framework documents (president annual reports, minutes of board meetings) and certain other sources (press releases, reports and interviews). To be sure, the phrase ‘MSF faced with violence’ requires clarification. Following the example of the case studies, we focus on situations of violence linked to a context of belligerence, whether they form the core of the comet or simply its tail. We decided against a focus on ‘social’ violence partly because projects organized around “exclusion and social violence” have been examined in a recent study, and partly because the specific questions they raise would have complicated the aims of the present discussion considerably. In short, we selected the sections of international humanitarian law concerning the rights of civilians and refugees as our frame of reference, rather than those relating to human rights or child protection. However, the frontiers of the real are more fluid than those of words, and we shall touch upon these situations several times during the course of the study. It should also be noted that when speaking of violence we chiefly mean physical violence, but do not exclude the other forms covered by IHL (destruction of personal goods and property, pillage, etc.).

When referring to MSF it must be borne in mind it is not an entity with a single, steadfast identity. The institution is composed of individuals with their own experiences and opinions. It is also a condensation of the collective experience accumulated over time, and the site where concepts of the humanitarian actor and of the relationship to other actors have evolved to create a space of possibilities which is not the same as that which existed twenty years ago. We have therefore studied the different voices the term MSF represents – the public voice, which conveys the impression of a relatively unified body; the self-analytical and policy setting voice employed by key individuals (often the President) when setting the framework of action; the contradictory voices raised in internal debates; and finally the voice from the field, which tells us about concrete practices. Each section of MSF may thus be likened to a polyphonic score. However, we have chosen to concentrate on the documents produced by the French section; a comparative study, although fascinating, would have been too great an undertaking given the vast increase in relevant material. For the purposes of the present study, MSF hence refers to MSF-France; this short-cut does not imply that the French section is representative of developments in the movement as a whole.

Whereas the case studies, with their focus on practices in the field and the process of drafting a message or arriving at a decision, reveal surprising similarities between the practices of the past and those of today, the discourse seems to have evolved significantly. The general synthesis initially centres on the way MSF expresses its role when faced with violence; it examines the general theorisation of the role, the key public positions and the policy framework. Three necessarily chronological sections trace the mutations of this role embodied in changing figures, from sentinel to aid worker. We then explore contemporary practices, comparing them with past practices and the discourse in which they are embedded in an attempt to identify the logic governing their development.