Contested legitimacy for anthropologists involved in medical humanitarian action: experiences from the 2014-2016 West Africa Ebola epidemic

Abstract The growing involvement of anthropologists in medical humanitarian response efforts has laid bare the moral and ethical consequences that emerge from humanitarian action. Anthropologists are well placed to examine the social, political, cultural and economic dimensions that influence the spread of diseases, and the ways in which to respond to epidemics. Anthropologists are also, with care, able to turn a critical lens on medical humanitarian response. However, there remains some resistance to involving anthropologists in response activities in the field. Drawing on interviews with anthropologists and humanitarian workers involved in the 2014-2016 West African Ebola epidemic, this paper reveals the complex roles taken on by anthropologists in the field and reveals how anthropologists faced questions of legitimacy vis-à-vis communities and responders in their roles in response activities, which focused on acting as ‘firefighters’ and ‘cultural brokers’ as well as legitimacy as academic researchers. Whilst these anthropologists were able to conduct research alongside these activities, or draw on anthropological knowledge to inform response activities, questions also arose about the legitimacy of these roles for anthropological academia. We conclude that the process of gaining legitimacy from all these different constituencies is particular to anthropologists and reveals the role of ‘giving voice’ to communities alongside critiquing medical humanitarianism. Whilst these anthropologists have strengthened the argument for the involvement of anthropologists in epidemic response this anthropological engagement with medical humanitarianism has revealed theoretical considerations more broadly for the discipline, as highlighted through engagement in other fields, especially in human rights and global health.


Introduction
Since the 1980s there has been a significant increase in infectious disease outbreaks globally, which has necessitated an expansion in medical and public health humanitarian response (Smith et al. 2014). Alongside this, there has been growing involvement of anthropologists in medical humanitarian response efforts. Anthropological interest in medical humanitarianism lies in the moral and ethical consequences that emerge from humanitarian action: in concerns with sickness and health; the suffering body in emergency situations; biomedical assumptions of universality; and the structural causes of such crises (Abramowitz and Panter-Brick 2015;Redfield 2005;Stellmach et al. 2018;Ticktin 2014).
The anthropological relationship with humanitarianism is complex, though. On one hand, anthropology is well positioned to contribute to medical humanitarian action with its 'concern with the holistic study of humanity in relation to social, political, cultural, and economic contexts, as well as the breadth of its studies done internationally' (Henry, McEntire, and Blanchard 2007, p. 111). With the majority of disease outbreaks being zoonotic in nature, Brown and Kelly (2014) also argue that anthropologists can reveal essential 'human-animal-nonhuman entanglements' during epidemics, to understand how human engagement with the material world, institutions, and animals facilitate the movement of pathogens (Brown and Kelly 2014, 283). Thus, as Janes et al. (2012) pointed out, whilst epidemiologists focus on mechanisms of transmission, anthropologists can 'speak of the scalar and multifaceted dimensions of influence in social systems -from an individual in a community to the wider political economy -that guide, constrain, or otherwise affect disease risk" (p. 18).
On the other hand, anthropologists take a critical position, turning the lens on medical humanitarianism itself, especially when the imperative to intervene bring concerns of neo-colonialism, neoliberal economic policies that weaken health systems, and the lack of local and historical knowledge in developing humanitarian responses (Abramowitz and Panter-Brick 2015;Fassin 2012;Henry and Shepler 2015;Ticktin 2014). As many anthropologists have argued, an anthropological lens can lay bare local political and economic tensions, as well as historical circumstances such as political repression, war and corruption that have engendered distrust of those in power due to legacies of past misrule and political upheaval that linger today (Atlani-Duault and Kendall 2009;Dhillon and Kelly 2015;Enria et al. 2016). The realities and urgencies of humanitarian crises, however, often do not allow humanitarian actors to foresee or plan for the consequences of their actions in the context of such complex social systems and histories. Thus, when taking such a holistic, social and long-term view of medical humanitarian activities, Abramowitz and Panter-Brick (2015) suggested anthropologists keep in mind that 'critique requires care ' (p. 8).
Scientific and humanitarian consensus about the unprecedented nature of the 2014-2015 Ebola outbreak in West Africa triggered inter-sectoral and interdisciplinary responses to robustly address the need for both better coordinated and more appropriate responses and future preparedness in the region (Abramowitz and Bedford 2016). The humanitarian community's need to learn lessons from the Ebola response as a whole also galvanised other ongoing efforts to systematise and integrate social science knowledge into global outbreak response. Venables and Pellecchia (2017) showed that 'anthropologists and the methodological tools of the discipline became a key approach for grasping and untangling the complexities of such an event' (p. 2). However, questions surrounding the legitimacy of anthropological involvement in epidemic responses have been raised by the humanitarian and public health community. Specifically, the methods applied by anthropologists are seen to be slow-paced and time-consuming (Beshar and Stellmach 2017;Sáez, Kelly, and Brown 2014), and the critical perspectives taken by anthropologists are seen to be unhelpful for urgent and contingent realities of epidemic (Abramowitz and Panter-Brick 2015;Ticktin 2014). In order to understand how anthropological knowledge and action can be legitimised for medical humanitarian action we draw on interviews with anthropologists and humanitarian workers involved in the West African Ebola epidemic to explore the realities and contested legitimacy of deploying anthropologists as part of medical humanitarian responses.

Methodology
In order to gain an in-depth understanding of the experiences of anthropologists on the ground during the epidemic as well as outsider perspectives about their work, we interviewed an equal number of anthropologists and humanitarian workers who had been deployed during the 2014-2016 EVD outbreak.

Participant recruitment
Participants were recruited through a snowball sampling method. Our first interviewee (an anthropologist) was recommended by LSHTM researchers who had been involved in the Ebola response and the second (humanitarian worker) was recruited during the Outbreak Intervention Symposium in October 2016. This symposium gathered both anthropologists and humanitarian workers who were highly active during both the West African Ebola and other contemporary epidemics. These two first interviewees helped us identify all other key informants who were humanitarian workers and anthropologists from their current network or from a past outbreak response mission. All of the respondents were overseas based or had just come back from an overseas-based position. Respondents from the humanitarian sector all had commensurate experience working in infectious disease response with INGOs and anthropologists all had previous regional expertise and were involved in research projects such as clinical trials, infectious disease ethnographic studies or health promotion research during the outbreak they described.
Attendees were invited for an audio-recorded telephone interview with FP in either English or French and asked to suggest other potential interviewees. Eighteen (18) interviews were conducted with nine anthropologists and nine humanitarian workers. Two of the anthropologists interviewed were West African (Sierra Leonean, Guinean or Liberian) and the remainder had European nationality. The humanitarian workers were selected from those whom had worked with anthropologists during an epidemic. Of these, two were also trained anthropologists but their primary role was delivering response activities (see Table 1).

Workshop
Following preliminary analysis of the interview data, the authors organised a workshop at the Wellcome Trust in London to present the findings. To collate feedback and further insights into the successes and challenges of involving anthropologists in field activities during medical humanitarianism activities, four audio-recorded, round table discussions were held with the thirty-seven anthropologists, NGO and humanitarian staff, technical advisors, experts, and academic researchers in attendance, including some of the in-depth interview participants. We identified workshop participants building from our research respondents cluster, who then recommended relevant professionals from their network ranging from key actors in the policy making arena to infectious disease research experts and health practitioners. We did not have a restrictive inclusion criterion for the participants, however as this research project was limited in time and resources, we could only extend the invitation to international experts living in Europe.

Limitations of the study
There are limitations to this study that should be considered. As we used a snowball sampling method, we potentially missed key informants. We are aware that these informants are close to the international response and policy making arenas and could be less representative of other geographical zones and professional backgrounds. However, these key informants contributed to capture power dynamics as well as policy, institutional and political contexts in which humanitarian workers and anthropologists had to work collaboratively.

Data analysis
Recordings of the interviews and group discussions were transcribed, translated into English (if necessary) and imported into NVIVO 10 software (QSR International). Analysis involved a framework analytic approach (Ritchie and Lewis 2003) focusing on questions of contestation and legitimacy. The coding structure was constructed after reading interviews with five anthropologists and three humanitarian workers using the interview questions as well as recurring, emerging themes and applied to the remaining transcripts. For the final synthesis of findings from all research activities, interviews were then re-read following the London workshop to highlight data relevant to new themes that emerged there.
Ethical approval for this study was granted by the London School of Hygiene and Tropical Medicine Ethics Committee; all audio-recording was done with participants' permission.
This study is part of a wider project to strengthen the UK's capacity to respond rapidly to outbreaks of infectious disease around the world by establishing a Public Health Rapid Support Team (PHRST) through a UK Department of Health-funded collaboration between Public Health England and an academic consortium led by the London School of Hygiene & Tropical Medicine (LSHTM).

Findings
The findings present three different constituencies with which anthropologists had to negotiate legitimacy (humanitarian sector, communities and global academic community) are three time points: entry to the field, crises in the field and long-term engagement with communities.

Negotiating entry to the field
For the anthropologists negotiating entry involved addressing epistemological tensions about the topics and geographies that would be open for anthropologists to comment on, including not only 'community' issues, but community perspectives on the humanitarian response itself Humanitarian responders saw anthropological expertise as essential to grasp the complexities of experiences and traditions at the local level given the heterogeneous social context of the epidemic -straddling three very different countries-and that the mode of transmission of Ebola relied heavily on social relations: person to person, bodies to bodies. However, with little experience of involving anthropologists in epidemic responses, most anthropologists reported that they spent much of their time at the beginning of the epidemic advocating for involvement, either on the ground or at the headquarters level. This included navigating socio-structural challenges of coordinating across the unprecedented number of international and local organisations involved. Whilst a few anthropologists were already working in-country as the epidemic commenced, the domination of international coordination mechanisms meant that most anthropologists were recruited through the European headquarters of NGOs or international bodies once crises emerged. Because of this, very few West African anthropologists were deployed, and those that were interviewed reported feeling discriminated against by western organisations. As one said, 'They [humanitarian organisations] were offering some very good incentives for people to come, they could have extended the same incentives, they could have gone out proactively to look for Sierra Leoneans' (Anthropologist 2). As anthropologists from the region, who were more likely to straddle citizenship, cultural, and linguistic boundaries they would have been the ideal interlocutors.
For all the anthropologists interviewed entry to the field raised concerns about their legitimacy as anthropologists within the response team.  A key point raised related to the ethics of humanitarian knowledge generation, particularly the danger of excluding a social perspective in the way responses are designed and adapted. As one responder, who was an advocate for anthropological involvement, explained, In practice, while such tensions were not fully resolved, easing them enough to get anthropologists into place involved convincing agencies that community input could be collected and distilled in a way that fit the needs of the international response. This required: 'a lot of face-to-face advocacy work, bilateral meetings, group meetings […] for the use of social science and the use of this kind of qualitative data -but then also making sure that it really could be applied and operationalized in a timely fashion' (Anthropologist 8).
Moreover, as the epidemic progressed, so did social resistance to response activities, causing humanitarian responders to suffer a crisis of legitimacy of their own in West Africa and on the global stage. The emergency became reconceptualised as socio-political rather than purely medical, reinforcing the urgent need to involve anthropological expertise. In Liberia, for example, a seminal moment came when humanitarian programmes were forced to reckon with the social outrage that followed the cremation of deceased patients by medical workers to avoid contamination. As one humanitarian worker saw it, 'When we look at Monrovia and the corpses burnt, if there had been an anthropologist present at the beginning of the planning …we would have found another solution' (Humanitarian Worker 2). Accordingly, one of the most salient and practical justifications that helped anthropologists gain entry to the field was to play a 'firefighting' role for humanitarian responders to address urgent socio-political issues and eventually one of 'cultural brokering' more generally, to contribute to effective and smooth-running risk communication programmes.

Negotiating legitimacy between responders and disease-affected communities
In the roles described above, anthropologists were expected to manage and negotiate relationships between humanitarian responders and disease-affected communities, each of whom had particular interests and needs. Anthropologists, in effect, became legitimators for each constituency they were working with and representing. The epistemological tensions with responders, which had manifested early on, thus followed anthropologists into the field, but evolved in new ways.

Firefighting
Many anthropologists said they were explicitly requested to join the response activities to help address or mitigate incidences of violence against the response. As one anthropologist put it, this positioned them as 'firefighters' whose role was to put out political 'fires' that threatened the security of response teams by calming disease-affected communities and listening to their grievances.
Humanitarians and anthropologists alike recounted multiple stories attesting to the importance of this role, as in this account about communication surrounding the forced transfer of suspected cases to treatment centres: 'I remember one community where I had to go and brokerpeace because the young men were ready to attack the military. "If they come back here, we're going to attack them" [the young men said] and I had to go and talk to them and calm them down and say, "Look, they'll kill you, they'll put you in prison. Is that worth it? (Anthropologist 2) In this case, this anthropologist and the people around her felt her engagement was pivotal in changing the tone of the response, pointing out that the responders, '…eventually learned […] They started working with the chiefs, they started going to the local communities instead of just sitting in Freetown making orders, threatening people, they started communicating' .
Other anthropologists, however, were more circumspect about their power to truly change peoples' minds and practice, reasoning that: Our role was to advise doctors, but we had our own research agenda at the time, on the rights of the citizen. […] it was necessary but difficult to find your place […] it was hard, there was fear and there were things that were difficult to see. In the end people were more concerned about survival than asking questions. (Anthropologist 6)

Cultural brokering
In less extreme circumstances, most of the anthropologists were invited to act as 'cultural brokers' or interlocutors. The need for this role, as described by both anthropologists and humanitarian workers, stemmed from a poor understanding by response teams of local contexts, languages and practices, and the lack of time for medics and epidemiologists to gain this understanding.
Acting as an interlocutor involved providing information in two directions: to international humanitarian responders who had limited knowledge of the context; and to local communities who were in the process of learning about the epidemic and its implications for their lives. Situated between the response teams and the communities, anthropologists thus acted as both interpreters of context (for the responders) and interpreters of science (for the communities), as described in the following quotes: "When I started, I noticed that many of the epidemiologists were not going out or they were afraid, or they didn't understand. I used to go with them and try to explain to them the people and help them with how to observe and interpret things that they were seeing to promote better communication. In the same way, I also tried to explain Ebola to the people and explain why people were coming there and the difference in symptoms" (Anthropologist 2).
"We were meeting the people in the villages where MSF or other teams had been rejected to enter. We had a focus group with different people, with men, youth, women, and then people who have already someone in the Ebola treatment centre. We were listening to them, asking what their fears were, or their understanding about the disease or the response, and how they think things should be done. … and then we did a feedback with the partners who were involved in the Ebola response" (Anthropologist 4).
However, there were differences between anthropological and humanitarian concepts of culture. From the humanitarian perspective, community views were synonymous with non-biomedical 'cultural beliefs' including social practices and rituals that risked impeding response activities, which anthropologists were best able to understand and modify. The involvement of anthropologists was important to be able to understand these, since, 'When you're dealing with a community that is centuries old and has a fixed belief around a religion […] you need a much more cautious approach to understanding how that community sees health and well-being' (Humanitarian Worker 9).
Such an understanding could be used to adapt responses to local 'context' , since in the words of one humanitarian worker, 'the most important thing is to always keep [in mind] the rituals and customs of people and incorporate them in the care we are delivering. This shows respect and facilitates trust. This should be reproduced in all kinds of medical response. Acting like their beliefs are not valid is the worst thing you can do' (Humanitarian Worker 3).
Anthropological knowledge was also expected to be used by humanitarians to change local behaviours and reduce community resistance and reluctance, through anthropologists' 'ability to rapidly understand and not make assumptions about people's knowledge, beliefs, frameworks […] to engage with those communities to be able to create that culture shift […] that behaviour change is the biggest operational challenge' (Humanitarian Worker 9). From the perspective of humanitarian workers, the role of anthropologists during the epidemic was thus to provide 'contextually appropriate' information to communities to ensure they complied with efforts to contain the epidemic, including reporting suspected Ebola cases and conducting safe burials. This also included: 'helping design locally and culturally appropriate messaging, educational materials, disseminating information down to communities. I think that they have an important role in essentially taking the needs of a response and sort of translating. I don't mean literally in a language sense, I mean, kind of culturally translating those needs into messaging, into educational material' (Humanitarian Worker 5).
Anthropologists commonly critiqued humanitarian responders for expecting communities to adapt behaviour or expecting anthropologists to find solutions to make response efforts acceptable. Narratives from the anthropologists interviewed revealed how they worked to legitimate the viewpoints of communities and try to change the ways that the response worked, through voicing their concerns to the responders. Nevertheless, some agreed that in certain circumstances it was useful for anthropologists to support behaviour change in communities, with one arguing: 'It is important [for responders] to know the customs and civilities, for example to offer condolences, to shake hands or not, how to joke […] Knowing all these codes helps [responders] know how to dismantle them when they pose problems in an epidemic context' (Anthropologist 1).

Negotiating legitimacy between responders and the international research community
Alongside the communication and translation activities described above, anthropologists were also involved in knowledge creation, which engendered tension between humanitarian and academic communities.
A fundamental question around legitimacy was whether anthropologists should be 'doing' anthropology at all, with anthropology fundamentally seen as a slow, in-depth, primarily research-based undertaking. A more contentious issue was whether or not anthropologists should be undertaking original ethnographic research during the response. The key concern, voiced by many humanitarian responders, was that anthropological research takes time and would not be of use to the emergency response, particularly if it detracted from other urgent needs. Public health researchers outside the anthropology field, however, tended to have a more nuanced idea of how the discipline could contribute: Further to this, the language used to describe ethnographic approaches, which are specific to anthropology, such as 'participant observation' was seen as unhelpful. As one humanitarian worker pointed out, 'I would recommend that social scientists get their hands dirty by doing and not observing.' (Humanitarian Worker 3). While several anthropologists felt this was a misrepresentation of their approach since rigorous observation involves constant questioning, testing assumptions and sharing ongoing interpretations to refine theory, they also recognised the need to prioritise early dissemination of actionable findings. As one anthropologist put it: "It's important to write up and share and disseminate messages. It is also important to work with the communities on the ground in that moment" (Anthropologist 5).
In practice, whilst conducting 'risk communication' activities, which clearly fit the needs of the response through community engagement, most of the anthropologists reported they were also able to collect 'data' that could be useful to the response. For instance, anthropologists were able to collect information on community concerns about response activities: 'We were going to the villages, we were having all these meetings and we were always accompanied by two persons from the region who knew the people, who speak the language. Ethnographic approaches were most often used to gather people's views and concerns about the epidemic and the outbreak response. As one anthropologist reported: 'Basically, as soon as the trial was being set up, I had a team of local social scientists, mostly from Kambia. I say social scientists, they became social scientists in the process and they were brilliant. We would go out into the community and do ethnographic observations every day basically and just sit in the most popular hangout places where most [motor]

bike riders hang out. We' d sit in the coffee shops and the palm wine bars and those kinds of places, and just chat to people about what they thought about the outbreak, what their experiences had been and listen out for any rumours or fears or concerns' (Anthropologist 7).
Other anthropological methods used during the epidemic included mapping techniques. One anthropologist used such a technique to explore local power dynamics. This involved 'doing that groundwork about, generally who are the power [brokers], who are the stakeholders, who are the people who are the influencers both formally and informal, breaking open that assumption that people who are officially the representatives of community are indeed the representatives of community, which is almost never the case' . (Anthropologist 7). These data were used to understand effective ways to communicate information about the epidemic, response activities and research activities as well as listen to rumours and concerns from a wide range of community stakeholders.
Such techniques reveal that despite concerns by humanitarian responders the research conducted by anthropologists during the epidemic was applied and utilised to inform response activities as quickly as was possible. This suggests that these kinds of data collection were valuable to the work of knowledge brokering. However, the need for rapid access to information by responders impacted on the quality of knowledge generated due to the lack of time for detailed analysis, and the focus on making the information decipherable for the front-line staff.

'…. what we were trying to do was get the information in a very timely fashion, with lots and lots of caveats thrown in about biases, et cetera, et cetera, but at least to give us something that could help better shape interventions' (Anthropologist 8)
This concern with the quality of the data collected using rapid methods raised concerns amongst academic anthropologists about whether such approaches were legitimately anthropological, that is using ethnographic techniques and critical reflection. As one anthropologist argued: 'Because anthropology is a discipline that creates knowledge, that has time to develop theory, but it's not only applied. Those who apply anthropology are like social workers or other kinds of social scientists, but not anthropologist' (Anthropologist 4).
However, by and large, the anthropologists interviewed defended the rapid approaches they undertook in the field to make their work more accessible to a non-academic audience: Some anthropology colleagues, particularly those who are more academic were very critical of this approach. That we were taking a huge amount of knowledge and trying to synthesize it down on to what became two pages of names but I can quite vigorously defend that position because the people we were working with who were asking for information were in no way positioned to [

Distilling and communicating findings
A key contribution of field anthropologists during the outbreak was to review and distil existing social literature on the intervention context for humanitarian responders. This particular role was accepted as legitimate at the international and local level and essential for response activities. As one humanitarian worker stated: We either have to teach anthropologists to understand what response options are, so they can interpret that information …This is about getting the information that you need operationally, reviewing where you're still having challenges, going back and checking why you're having those challenges, getting more information, interpreting that again. More in a cycle rather than a oneoff data collect. (Humanitarian Worker 1) It involved distilling and communicating findings, collating geopolitical and historical information about Ebola as well as local understandings of disease. The different platforms set up during the West African Ebola epidemic were crucial sources of information for both anthropologists and humanitarian workers, whose priorities were operational. Whilst some of the anthropologists had in-depth knowledge of the country they were working in, others sought information provided through websites including the Ebola platforms, anthropological literature about the country, and discussions with other anthropologists who had previously conducted research in the country. Knowledge gained from involvement in other epidemics was also utilised. This information was collated by anthropologists and presented in a format that was useful for informing the response activities. One anthropologist said 'We want to be using as many different sources and different types of data as possible, to build up the most granular picture that we can, around not just the social science-socio-cultural issues, but also around some of the behavioural surveillance work, for example, looking at how we dig that into the epi[demiology]'. (Anthropologist 8) The data collated was utilised by anthropologists in-country to understand issues that arose during the epidemic as well as to provide information to different actors in the response, including the military. This information was tailored to be accessible to those involved in the front line of the response.
'They had what they called combined crisis team meetings which were cross-government and including the military. They were very military-led meetings where you each add a point, and it has about three minutes. They happened maybe twice a week or something… I went into some of them in an ad hoc way depending on if I had a report that was about to come out, then I' d go in and very briefly, brief the headline of that, so that people were aware of it. Also, we had sent it through to the operations room in Freetown, and then again, the headlines were briefed in by people there for each paper' (Anthropologist 9).

What I was largely doing was actually tapping into LSHTM and the Ebola Response Anthropology
Platform in bringing in that academic and regional expertise almost translating it for my kind of audience, military. How can I convey those messages around the spread of Ebola in a simple way, but in language and things that would be more digestible to the military? Then what did it mean for what they were doing. Ended up writing a number of papers on the issue, which was part of an introductory one with what people's perception of healthcare and illness, and international healthcare workers and that kind of thing. We did one on stigma, we did one on the role of civil societies (Anthropologist 9) Forums for discussion were important in ensuring rapidity of dissemination or the provision of briefs as well as sharing between social scientists. One anthropologist noted that forums for discussion were particularly useful in an emergency context: 'I think also with the context changing and it being an emergency, everybody's moving so quickly that they're not always taking the time to talk and see what other organizations are doing. We had, I think it was, weekly social science group that met in Monrovia which was made up of anthropologists, psychologists, people working in community engagement from different organizations. That was the chance to share ideas and find out what other people were doing. That was quite useful'. (Anthropologist 5)

Discussion
This paper presents a case study of anthropological involvement in a medical humanitarian response to an unprecedented Ebola epidemic in West Africa. We argue, as others have, for the need to involve anthropologists and other social scientists in a range of activities to support preparedness, response and health and other system strengthening (See . Anthropologists' role in humanitarian responses to disease outbreaks is likely to increase given the need to understand social, political and economic complexities that fuel disease transmission Abramowitz and Panter-Brick 2015;Stellmach et al. 2018). As discussed, anthropologists can describe local social networks as well as determinants of health and patterns of gender, age, and kinship, within social, economic, political and cultural systems (Henry and Shepler 2015;Janes et al. 2012;Lloyd-Smith et al. 2009). However, the interviews revealed logistical, moral, and epistemological tensions that the anthropologists negotiated to gain access to, and work in, the field due to their responsibility for three constituencies: the medical humanitarian sector, the disease-affected populations, and academia. Negotiating tensions between these required anthropologists to gain legitimacy in each constituency, if we understand legitimacy as 'the normative belief by an actor that a rule or institution ought to be obeyed. It is a subjective quality, relational between actor and institution' (Hurd 1999 p.381). Legitimacy theory reveals how institutions pursue different strategies of legitimacy, including procedural legitimacy (an institution is legitimate through the principles of right process), substantive legitimacy (policies or rules to are justified on the basis of shared broader norms and values), and output legitimacy (public assessment of the relevance and quality of the institution's performance).

Legitimacy as, and legitimising of, medical humanitarian responders
At the start of the West African epidemic there were questions about the usefulness of anthropology during an outbreak, which led to the delayed deployment of anthropologists during the West African epidemic. As noted by , integration of anthropologists during the humanitarian crises is often 'delayed, inconsistent, and distant from the centre of decision making and resource prioritisation' (p. 330) (see also Janes et al. 2012). This suggests a question of legitimacy of anthropology within the medical humanitarian sector, which was only extended to anthropologists once the medical humanitarian community faced a number of crises and anthropologists were engaged to help manage them. However, once in the field, anthropologists were often asked to support health communication, social mobilisation, or provide analysis of 'culture' (see also Abramowitz and Bedford 2016). As Fassin has described when deployed "as an anthropologist" (his inverted commas), his role was 'to help them understand what had caused the difficulties they encountered during a particular mission, expecting me to give them "cultural keys" for interpreting "resistance from the population' (Fassin 2011, p. 40). This suggests that the expected role for anthropologists in medical humanitarian response is to ensure legitimacy of the response effort itself by ensuring that approaches to response are acceptable to the communities rather than trying to convince, persuade or coerce 'communities' into accepting it and its hegemony. As interlocutors, those that were deployed were also asked to act as conduits of accountability for humanitarian responders. Whilst taking on these roles, anthropologists now argue for a more legitimate position within response early on that 'needs collaboration, local involvement, and joint solution finding that meets both socio-cultural needs and humanitarian response protocols' (Wilkinson and Leach 2015, p. 12). Finally, from a critical perspective, anthropologists during this epidemic rendered clear the normally invisible social and political structures that support the humanitarian sector's claim to power, legitimacy and ability to influence (Mosse 2004).

Legitimacy of anthropologists in affected communities
Whilst struggling for legitimacy within the medical humanitarian sector, anthropologists also had to manage their legitimacy within the disease-affected communities. During the early part of the epidemic, anthropologists in the field and elsewhere argued that the crises that emerged were due to lack of community consultation and a fundamental disregard for community concerns. Whilst this could be attributed to the unprecedented scale of the epidemic and the complexity of the response, it also reflected a wider epistemological issue underpinning response structures that focus was on technical interventions that prioritise speed of response over community participation in decision-making. This would suggest, in legitimacy theory, a conflict between output and processual-type legitimacy strategies.
Whilst recruited to act as firefighters and interlocutors, the anthropologists deployed had some success in listening to, and acting for, the disease-affected communities and ensuring that responders understood local concerns, and that crises were not 'cultural' but rather social, economic and political. Anthropologists engaging in two-way communication and asserting some influence ensured that some of the fundamental structural governance inequalities of the response were addressed. For example, by revealing the dehumanising aspects of the response activities such as quarantine and medical burial practices (Fairhead 2014), as well as the negative rhetoric about ' African' cultural practices (Abramowitz and Bedford 2016;Venables and Pellecchia 2017). This led to significant changes in the practices of the response, including ensuring safe and dignified burials.
Disease-affected populations are often not involved as agents in conferring legitimacy because of their less powerful positions in humanitarian governance arrangements and the lack of formal representation and decision-making in humanitarian response. Certainly, there is evidence that disease affected communities were initially excluded from the input of humanitarian responders' legitimation claims in West Africa, either procedural or output based. Engaging with disease-affected populations and taking seriously their demands on the response required building legitimacy with this previously non-powerful constituency. As Abramowitz (2017) argued, anthropology came to 'serve as a semantic marker of solidarity with local populations, respect for customary practices and local socio political realities, and an avowed belief in the capacities of local populations to lead localized epidemic prevention and response efforts' (Abramowitz 2017, p. 421). Thus, anthropological engagement with disease-affected communities allowed new legitimation practices as well as the management of tensions with the existing legitimation strategies and accompanying norms of the humanitarian community. However, it is important to notes. as pointed out by some of anthropologists interviewed, the involvement of national or regional African anthropologists may have strengthened legitimacy with local communities.

Legitimacy in academia
The experiences narrated by anthropologists in the field during the West African Ebola epidemic raises the legitimacy of anthropology itself, when conducted in medical humanitarian responses. For the anthropologists deployed, there was a moral imperative to act. As Redfield argues '…when facing acute episodes of human suffering, anthropologists also tend to measure moral failure in destruction and death. When disaster strikes they often display humanitarian impulses and expectations, desiring action and imagining a global response' (Redfield 2005 p. 8). The anthropologists deployed agreed to prioritise needs the of epidemic response due to the rapidity of change, and heightened fear and distrust in the disease affected communities, rather than long-term ethnography involving 'proximity, intimacy, and critique' (Benton 2017 p. 503). This is in line with other humanitarian efforts that focus on short-term solutions and the longer-term concerns of anthropology (Redfield 2005).
As well as those interviewed for this study, other anthropologists voiced concerns about approaches to communication, risk and uncertainty and the continued focus on a biomedical and epidemiological approach, which did not take into account social, cultural, economic, political, and religious factors (Atlani-Duault and Kendall 2009;Jones 2011;Leach 2015;Wilkinson and Fairhead 2017). Anthropologists were able, though, to challenge cultural blaming through the provision of anthropological knowledge, either from locally collected data sources or external sources such as The Emergency Anthropology Initiative, the Ebola Response Platform, and the Réseau Ouest-African SHS Ebola Network.
These initiatives provided what  referred to as, 'social science intelligence' , providing recommendations for community engagement, and safe and dignified burials, which was particularly useful for anthropologists in the field (Venables and Pellecchia 2017). Further to this, anthropologists offered critical perspectives that contributed to understanding social responses to, and ethical and human rights perspectives on, quarantine and isolation (Calain and Poncin 2015;Pellecchia et al. 2015), guided community engagement (Marais et al. 2016;, provided advice in the application of biomedical models (Chandler et al. 2015), and understanding on how structural violence intersected with the epidemic (Fairhead 2016;Leach 2015). The anthropologists interviewed here, and published elsewhere, explored intended and unintended consequences of response activities, tracked myths and local power structures, using anthropological techniques (Abramowitz, McLean, et al. 2015;Enria et al. 2016;Leclerc-Madlala 2014;Palmer et al. 2014).

Theoretical contributions to anthropology as a discipline
Anthropological engagement with the Ebola outbreak provides a lens to examine the methodological and moral liminalities of anthropological research in the context of medical humanitarian emergencies, and thus provides a number of theoretical contributions to the discipline more broadly. In particular, anthropological engagement in this field highlights tensions in professional identity and practice. These tensions align with theoretical insights about anthropological contributions to human rights administration (see Jean-Klein and Riles 2005). These authors explore the unique contribution of anthropology to this field and highlight ways in which the practice of human rights have elicited two registers of anthropological engagement, which resonate with anthropological engagement with medical humanitarianism. One register focuses on "co-construction" or "giving voice", in which 'anthropologists set out to express both moral and analytical empathy with subaltern subjects in the field of study' (Jean-Klein and Riles 2005, p. 176). Engagement with medical humanitarianism in West Africa has revealed anthropological solidarity with diseases-affected communities through the process of 'giving voice' to these communities, especially around their suffering. As Jean- Klein and Riles (2005) noted in the field of human rights administration, the narration of individual and collective experiences of the epidemic in itself 'became its own form of relief ' (p. 177).
Thus, ethnographic approaches can be seen as a therapeutic resource through which disease-affected populations can articulate their suffering as narrative (see Jean-Klein and Riles 2005). The other register focuses on 'denunciation' , which 'entails condemning the proliferation of technocratic regimes and the injustices inherent in organized and official relief efforts' (Jean-Klein and Riles 2005, p. 174), such as the critique by anthropologists of the dehumanising approaches to burials and quarantine (Fairhead 2014). Thus, as Jean-Klein and Riles (2005) have argued for those in human rights administration, the anthropologist working in medical humanitarianism becomes an instrument for ethical practice through active listening, providing anthropological knowledge and challenging reified views of culture. As highlighted by the anthropologists deployed in the West African Ebola epidemic, the anthropologist is uniquely placed within the field of medical humanitarianism to blend co-construction and denunciation.
However, whilst uniquely placed to co-construct and denunciate, anthropology's engagement within broader epidemic response structures has revealed similar discomfort that anthropologists encounter when engaging more broadly with global health initiatives that 'stresses populations over people and indicators over values and practices' (Reynolds and Lange 2019, p.3). In particular, challenges have emerged for anthropologists when they are positioned as experts in the field, rather than the contexts that they work in, which are similarly experienced by anthropologists engaging with biomedicine and global health that do not always embrace ethnographic data as legitimate (Reynolds and Lange 2019). This challenge is exacerbated by the widespread shift in global health research from the application of anthropological methods to qualitative methods, usually due to funding constraints or other exigencies that do not accommodate in-depth anthropological research (Reynolds and Lange 2019). As within global health research, the anthropologists who worked in medical humanitarian emergencies were expected to 'ignore many of the theoretical and interactive techniques bequeathed by anthropology's long history to accommodate the expectations of colleagues, fit within budgets and timetables, and adhere to other apparently necessary constraint' (Reynolds and Lange 2019, p. 10). We have seen, though, how anthropologists involved in the West African Ebola epidemic accommodated such constraints by deploying evidence in two ways: as synopsis of longterm, contextualised ethnographies, and through rapid analyses of field notes from engagement with communities. Whilst not conventional practices by anthropologists, these approaches satisfied tensions alluded to by colleagues in the field about the 'slowness' of the process ethnographic evidence-making. However, as Jean- Klein and Riles (2005) have argued these approaches can become 'detachable from the unique purpose for which they were designed, that is to make knowledge of a particular kind -ethnographic knowledge' (p. 182). Thus, they argue, whilst such approaches maintain anthropological expertise, they focus the role of anthropology on providing facts and in this way devalue ethnography, which sets anthropologists apart from other disciplines involved in medical humanitarianism. Reynolds and Lange (2019) argue for the importance of anthropological engagement in global health because it takes such multiple perspectives and positions, and thus in agreement with Jean-Klein and Riles (2005) we argue that anthropologists engaging with medical humanitarianism must maintain a commitment to ethnography as a 'form of care for the discipline itself ' (p. 174)

Conclusion
The West African Ebola outbreak has provided a lens to examine the methodological and moral liminalities of anthropological research in the context of a medical humanitarian emergency. Henry and Shepler (2015) argued that 'We need a humanitarian anthropology that is embedded in that response, yet is able to be critical of it' (p. 21). Our paper shows that anthropologists deployed during the West African epidemic were able to straddle both positions, through con-construction and denunciation. However, questions of legitimacy arose in their deployment, as responders on one hand, and academics on other. This process of gaining legitimacy from all these different constituencies is a particular burden for anthropologists as opposed to other sorts of academic and humanitarian responders deployed during outbreaks. The experiences, and successes have, though, led to an increased involvement in subsequent outbreaks, including the Zika outbreak in Brazil. However, and the more recent Ebola outbreaks in the Equateur and North Kivu regions of the Democratic Republic of Congo have reminded the global health community of the challenges encountered by responders during the West African epidemic in terms of community reluctance and retaliation toward intervention, fears, rumours and raised shown that lessons learned about the involvement of anthropologists are not necessarily being applied to understand how to engage communities (Nguyen 2019).Thus, there remain questions about in what ways anthropologists' involvement in response on the ground, given that there remains a focus on risk communication and deployment only once an epidemic is protracted or a negative event occurs.
These questions continued to be asked through new social science positions, such as those on the rapid support teams, including the UK Public Health Rapid Support Team. Funding opportunities for social science research for epidemic preparedness, the GOARN social science group, the Epidemic Response Anthropology Platform (Institute of Development Studies and London School of Hygiene & Tropical Medicine 2019) also ensure that these debates continue. We conclude that whilst these anthropologists have strengthened the argument for the involvement of anthropologists in medical humanitarian response there is still a need for empirical knowledge to inform the ways in which anthropologists are deployed that ensures that in-depth and critical insights revealed by anthropology are taken seriously, and the practices of co-construction and denunciation remain central to anthropological engagement with medical humanitarianism.

Ethical approval
Ethical approval for this study was received from the London School of Hygiene and Tropical Medicine Ethics Committee.