Ethical concerns before, during and after ethnographic research

As I will soon be teaching a module on research ethics for PhD students in translation and interpreting, this post a bit more of a literature review. Social research is, at its core, a relational activity. We interact with colleagues, participants and audiences/readers.

So, how can we make sure these interactions remain ethical?

1. Ethics of anticipation: ethical review and informed consent

Anthropologists have long argued that the requirements of institutional ethical review (common in the UK and the US, not at all in francophone countries!) and informed consent do not fit the specificities of ethnographic field research, where ethics is a “recognition of the problem”, rather than a solution (Lambek, 2012, p. 142) and an on-going decision-making process that cannot be reduced simply to obtaining informed consent:

“A wider conception of ethics recognizes that moral judgements are made at every juncture of ’scientific’ practice, and these judgements are made together with other parties involved in the research. Despite wider pressures to do so, it is vital to resist discussion of ethics as a methodological and institutional ’tool’, or as a mode of legitimization and authorization, and to continue to argue that ethics involves a broader field of negotiations with varied moralities, philosophies and politics.” (Mookherjee, 2012, p. 133)

Ethical review and informed consent are problematic in ethnographic research, as it supposes that the researcher is able to anticipate “with whom, for how long, to what end, and where” she will work (Simpson, 2011, p. 380), which runs counter to the inductive, iterative and open-ended nature of ethnographic inquiry. Furthermore, it is not easy to define who is a participant, i.e. who is affected directly or indirectly by the researcher’s presence in the field. Therefore, “[a]pplying the logics of autonomous individuals and signed consent forms to the ethics of participant observation necessarily makes ethnographic practice appear non-consensual” (Mookherjee, 2012, p. 133). Secondly, obtaining informed consent mainly in order to allow the researcher to protect herself and avoid liability (cf. the “audit culture” in contemporary anthropology (Lambek, 2012, p. 141)) is an unethical use of research ethics.

In light of the specificities of ethnographic research, the American Anthropological Association has adopted a broader definition of informed consent:

“Anthropological researchers should obtain in advance the informed consent of persons being studied, providing information, owning or controlling access to material being studied, or otherwise identified as having interests which might be impacted by the research. It is understood that the degree and breadth of informed consent required will depend on the nature of the project and may be affected by requirements of other codes, laws, and ethics of the country or community in which the research is pursued. Further, it is understood that the informed consent process is dynamic and continuous; the process should be initiated in the project design and continue through implementation by way of dialogue and negotiation with those studied. Researchers are responsible for identifying and complying with the various informed consent codes, laws and regulations affecting their projects. Informed consent, for the purposes of this code, does not necessarily imply or require a particular written or signed form. It is the quality of the consent, not the format, that is relevant.” (American Anthropological Association, 2009, Art. 4)

Ethical review requirements vary considerably between countries, we can argue that there are “national cultures of ethics” in the social sciences (Fassin, 2006), linked closely to national traditions of higher education. The ethnographic method is particularly difficult to fit into ethical review requirements (“serious tension” (Simpson, 2011, p. 379); “useless restrictions” (Fassin, 2006, p. 524)). More importantly, ethical review and informed consent fail to adequately address the ethical concerns that arise during fieldwork:

“[The ethical review model] leaves aside most of the important ethical issues raised by fieldwork and gives a false guarantee that ethics is respected through purely formal operations. In this sense, as well, it threatens ethnography by proposing lazy responses to real needs and, thus, avoiding the more painful moral and political questions sociologists and anthropologists should deal with.” (Fassin, 2006, p. 524)

And create the illusion that ethical dilemmas have right or wrong answers:

“Sometimes the course of action is obvious but more often it is ambiguous, as one is confronted with novel circumstances and criteria, or even the ostensible absence of criteria, and pulled between different sets and kinds of obligations and commitments. It is a mistake —an ethical blind spot —to expect there is always an obvious, single, right, virtuous judgement to make or a correct path to follow in every situation” (Lambek, 2012, p. 143).

In the context of research involving participants from developing countries, the considerable power gap between researchers and participants, and the often vulnerable status of aid beneficiaries begs the question of their general ability to provide informed consent (Simpson, 2011, p. 379)! Relationships with research participants in the field are particularly challenging from an ethical perspective and differ considerably from the types of relationships built in an interview setting. One of the negative consequences of ethical review is a “conflation of interviews and ethnography”, where one becomes a proxy for the other and ethnography is reconfigured as a series of discrete, recordable interactions (Simpson, 2011, p. 381).

2. Ethics of interaction: relationships with participants

In development projects and humanitarian aid, researchers might easily be perceived as being part of the NGO/international organisation delivering aid (Batianga-Kinzi, 2014), and sometimes they are (Moussa, 2014). The power asymmetry and agency gap between researcher and aid beneficiaries creates a potential for dependency, including the solicitation of money (Molony & Hammett, 2007), and close emotional involvement (Ouedraogo, 2014; Sundberg, 2014). The “participant” in ethnographic field research comes in many forms, “informants, interlocutors, consociates, collaborators, consultants, and friends” (Simpson, 2011, p. 384), and exists as part of a wider network of relationships. This relationality of participants makes it difficult to manage informed consent in practice (e.g. A driver provides useful information about how an NGO is viewed by the local population. This information is crucial to understand our data, yet no informed consent has been obtained from the driver.).

The anthropologist as participant-observer becomes part of the field and of the network of relationships between actors. As interlocutors in the field are far from homogenous, conflicts might sometimes be unavoidable. Furthermore, researchers will likely encounter attitudes that go against their own values. The ethical challenge here consist of engaging with these attitudes, even though this might sometimes result in advocacy or taking sides. Far from a “rule-based notion of ethics”, researchers are confronted with building complex relationships that are “founded on trust, respect, and an avoidance of delimiting the subject” (Simpson, 2011, p. 385).

An example from my PhD fieldwork in Kenya:

A Kenyan informant, who had provided me with very useful insights about Kenya throughout my field work, got angry when, during my third visit to Nairobi I went to UNHCR to train refugee interpreters. Why was I spending time training Somalis when there were many Kenyan’s not able to get education? Why was I helping the refugees who were already getting everything for free and just benefitting from handouts? My liberal, generally open attitude had allowed me to build a good relationship with this Kenyan, yet in applying the very same attitude to Somali refugees, I was compromising this relationship. In the interest of my relationship with my Kenyan informant, and in the interest of an ethical relationship with my interlocutors in Nairobi in general, I sat down with him for a long conversation. My aim was to better understand his attitude, but also to humanise Somali refugees. We discussed ethics, in the form of Christian values that I knew he adhered to, and the implications of these values for interactions with others. In an environment where the risk of ethnic and inter-religious violence remains high, neutrality or pretending to agree with his attitude for the sake of a peaceful relationship did not seem like the most ethical way forward.

3. Ethics of textualisation: anonymity and dissemination of results

Confidentiality refers to not discussing research-related information with anyone outside the research team, while anonymity specifically refers to protecting the identity of research participants (Wiles et al., 2008; Saunders et al., 2015). Anonymity can be preserved by using pseudonyms for participants and research locations (Zolesio, 2011; Wiles et al., 2008; Saunders et al., 2015), or by completely omitting data segments from a publication (Thomson et al., 2005) (which makes it difficult to assess the rigour of a study! (Wiles et al., 2008)). In some cases, authors have opted to transform their qualitative data into a fictionalised account (for an example involving a development project see Rottenburg, 2009). Anonymising research locations allows for greater decontextualisation of findings, making it easier to draw comparisons across locations, however, it also creates the risk of over-generalisation (Nespor, 2000).

Two recent trends have imposed practical limits on anonymity: technology and the participatory turn. Technology, in particular social media, makes it difficult to fully anonymise research locations (researchers presence in a city or their participation in a public event will leave an on-line trace, as people might take photographs and upload these to social media) and research participants (who might be part of the researcher’s network of friends) (Nespor, 2000; Saunders et al., 2015). Additionally, research participants might want to be mentioned by name and credited as participants in a particular project (Saunders et al., 2015; Wiles et al., 2012).

Ethical decision-making at the writing stage involves a threefold relationship between informants, truths and publics, and a selection process whose “politics” are rarely made explicit (Simpson, 2011, p. 389).

4. Further reading…

For an entertaining yet very realistic insight into the ethical dilemmas researchers are faced with as participant-observers: Bogerhoff Mulder & Logson (1996); Barley (2000).

General introduction to different forms of ethnographic writing and the textualisation of culture: Van Maanen (1988); Clifford & Marcus (1986).

Ethical challenges of ethnographic research in development/humanitarian aid contexts: Aympam et al. (2014).


American Anthropological Association (2009). Code of ethics of the American Anthropological Association .

Aympam, S., Chelpi-den Hamer, M. & Bouju, J. (2014). Défis éthiques et risques pratiques du terrain en situation de développement ou d’urgence humantaire. Anthropologie & Développement 40-41, 21–41.

Barley, N. (2000). The Innocent Anthropologist: Notes from a Mud Hut. Waveland Press.

Batianga-Kinzi, S. (2014). L’ethnographie au risque de l’agression : expérience de terrain à risque. Anthropologie & Développement 40-41, 87–97.

Bogerhoff Mulder, M. & Logson, W. (1996). I’ve Been Gone Far Too Long: Field Study Fiascoes and Expedition Disasters. RDR Books: Oakland California.

Clifford, J. & Marcus, G. E. (Eds.) (1986). Writing Culture: The Poetics and Politics of Ethnography. Berkeley.

Fassin, D. (2006). The end of ethnography as collateral damage of ethical regulation? American Ethnologist 33 (4), 522–524.

Lambek, M. (2012). Ethics out of the ordinary. In R. Fardon, O. Harris, T. Marchand, C. Shore, V. Strang, R. Wilson & M. Nuttall (Eds.), The SAGE Handbook of Social Anthropology, Volume 2. Thousand Oaks, CA: SAGE Publications, 140–152.

Molony, T. & Hammett, D. (2007). The friendly financier: Talking money with the silenced assistant. Human Organization 66 (3), 292–300.

Mookherjee, N. (2012). Twenty-first century ethics for audited anthropologists. In R. Fardon, O. Harris, T. Marchand, C. Shore, V. Strang, R. Wilson & M. Nuttall (Eds.), The SAGE Handbook of Social Anthropology, Volume 2. Thousand Oaks, CA: SAGE Publications, 130–140.

Moussa, H. (2014). L’anthropologue entre les tyrannies des terrains et le choix d’une éthique. Anthropologie & Développement 40-41, 99–121.

Nespor, J. (2000). Anonymity and place in qualitative inquiry. Qualitative Inquiry 6 (4), 546–569.

Ouedraogo, R. (2014). Face à l’avortement : exigences éthiques et dilemme moral à Ouagadougou (Burkina Faso). Anthropologie & Développement 40-41, 123–141.

Rottenburg, R. (2009). Far-Fetched Facts: A Parable of Development Aid. Massachusetts: The MIT Press.

Saunders, B., Kitzinger, J. & Kitzinger, C. (2015). Anonymising interview data: challenges and compromise in practice. Qualitative Research 15 (5), 616–632.

Simpson, B. (2011). Ethical moments: future directions for ethical review and ethnography. Journal of the Royal Anthropological Institute 17, 377–393.

Sundberg, M. (2014). Ethnographic Challenges Encountered in Rwanda’s Social Topography. Anthropologie & Développement 40-41, 71–86.

Thomson, D., Bzdel, L., Golden-Biddle, K., Reay, T. & Estabrooks, C. A. (2005). Central questions of anonymization: A case study of secondary use of qualitative data. Forum: Qualitative Social Research 6 (1), no pag.

Van Maanen, J. (1988). Tales of the Field: On Writing Ethnography. Chicago, IL: University of Chicago Press.

Wiles, R., Coffey, A., Robinson, J. & Heath, S. (2012). Anonymisation and visual images: issues of respect, ’voice’ and protection. International Journal of Social Research Methodology 15 (1), 41–53.

Wiles, R., Crow, G., Heath, S. & Charles, V. (2008). The Management of Confidentiality and Anonymity in Social Research. International Journal of Social Research Methodology 11 (5), 417–428.

Zolesio, E. (2011). Anonymiser les enquêtés. Revue ¿ Interrogations ? 12 (Quoi de neuf dans le salariat ?), no pag.

Can interpreters be impartial in a war?

Interpreters mediate between speakers of different languages in a variety of settings. Consequently, different professions have emerged within the field, namely conference interpreting, court interpreting and community interpreting. Each with their respective professional codes of ethics. A principle that can be found in almost all of these codes is neutrality or impartiality. An interesting variation on the theme is “multipartiality”, a term I found in the code of ethics for community interpreters in Switzerland. While each of these three terms has a slightly different meaningmultipartiality” for instance, implies that the interpreter is simultaneously on everyone’s side, rather than on no one’s side (impartiality)that distinction is not important for the purpose of this article. Consequently, I shall opt for “impartiality” as my generic term and use this to refer to a positioning of the interpreter in a somewhat imaginary No Man’s Land situated between the different parties to the interpreter-mediated interaction or “trialog”.

If we want to understand the possibility and nature of interpreter impartiality in different settings, we need to take a closer look at the relationship between the two parties and the way in which information flows between them. If we look at a communicative event as a transaction between two parties, then information is the currency they use. As in any context “Follow the money!” is sound advice to understand what is really going on.

Based on my experience as an interpreter, interpreter trainer and (very occasional) user of interpreting services, I would argue that there are three types of transaction: extraction of information, dissemination of information and exchange of information. Each of these presents different challenges and opportunities as far as interpreter impartiality is concerned. Such a typology depending on information flow has, as far as I know, never been presented in the interpreting literature.

Scenario 1: Extraction of information

In this scenario, there is often no mutual agreement to enter into communication. Rather, one party tries to extract information from the other party. The purposes of information extraction can range from noble (a doctor attempting to better understand the patient’s symptoms, a social scientist trying to find out more about the situation of child nutrition in a remote Andean village, a social worker interviewing parents about their child’s special needs, etc.), to controversial (an asylum officer trying to obtain detailed information about the applicant’s past in order to assess the credibility of their story, a police officer interrogating a suspect, etc.), to outright ugly (a member of an armed group interrogating a prisoner of war under torture, a CIA agent using “enhanced interrogation techniques” to extract intelligence, etc.). All of these scenarios have involved interpreters, yes, even the one involving torture:

“Afghan officials are seeking the arrest of an Afghan-American interpreter on charges of murder and torture. The interpreter was reportedly identified in a video showing Mohammad’s abuse.” (Source: Democracy Now!, 22 May 2013)


Although both speakers might take turns in this scenario, sometimes even using similar amounts of speaking time, the information flow is largely unidirectional: A aims to obtain information from B. To achieve this aim, A might ask direct or indirect questions, or use more subtle approaches, such as providing some information of their own in order to encourage B to share their knowledge. Granted, B might sometimes be able to use the information they are holding as leverage (think for instance of a small-time crook who negotiates a mitigation of their sentence in exchange for important information about a larger network of organized crime) but this rarely gives them full control over the interaction.

Scenario 2: Dissemination of information

While this scenario might seem like “inverted extraction of information”, I would argue that what matters is who is in control. In the extraction scenario above, the party extracting the information is generally in control of the communication situation: they select the other party, they define where and when the interaction takes place, and they are the ones paying the interpreter.

In the dissemination scenario, control resides with the holder of the information. Once again, possible scenarios range from the benign (a community health centre organizing an information session for young mothers, a teacher informing parents about their child’s progress in school, an induction seminar for new members of staff), to the ambiguous (an advertising agency presenting the merits of their client’s latest products, a political party on the campaign trail), to the outright detrimental (a cult leader convincing teenagers to cut all contact with their families, a political representative convincing his constituents to take up arms against a minority group).


As in the extraction of information scenario presented above, speakers might take turns, although A will try to provide information to B and use B’s statements mainly to check whether information has been received and understood. As such, information flow remains unidirectional, A is the holder of knowledge that B is supposed to have acquired at the end of the encounter.

Scenario 3: Exchange of information

In this scenario we genuinely have two parties to the interaction. Their intention and willingness to communicate is mutual, and while the amount and quality of information each party is holding might vary, both are actively contributing to the conversation. What distinguishes this scenario from the two scenarios above is that both parties have the power to influence the course of the conversation and change the topic, be it by presenting information in a certain way or asking questions. Both parties are there to learn from and about each other. This type of scenario is quite common in conference interpreting settings (health experts from different countries coming together to discuss a global progress report) or business interpreting settings (a delegation from China comes to visit a car manufacturing plant in Germany). A variation of this scenario are “negotiation settings”, where the clear objective of the two parties is to resolve a dispute and/or reach an agreement about the best way forward. Such negotiations can of course be tense, despite a shared willingness of actors to engage in communication. However, both parties generally have an interest in the interpreter’s impartiality or might sometimes even use two interpreters (each side employing their own, as is often the case in bilateral meetings between heads of government for instance).


In this scenario, information genuinely flows both ways. Ideally, speakers take turns and use similar amounts of speaking time. However, even in cases where A is more knowledgeable about a topic, B will remain actively involved in the conversation, for example by asking questions, making suggestions and comments, thereby pushing A to clarify their explanations or change their point of view.

Implications for interpreter impartiality

When a communication scenario is inherently asymmetrical, for example aiming towards extraction or dissemination of information, impartiality is difficult for the interpreter to maintain. Indeed, there is sometimes no imaginary neutral territory between the parties that the interpreter can occupy, no shared space or contact zone that is created by the parties’ mutual agreement to enter into communication. In extraction and dissemination scenarios, the Gricean maxims of communication are often not respected. Instead, one party aims to influence the other party, and often to convince them to do something they would not otherwise have done (reveal strategically important information or quite simply change the way they manage their children’s nutrition). In these scenarios, the ethical space the interpreter needs to consider is not limited to her own code of ethics but to the nature of the communicative event and the ethical implications of the change that the party in control of the situation is trying to create.

Many will agree that teaching young girls about HIV/Aids aims to bring about a positive change, while convincing young men to kill their neighbors in protest against election results makes everyone worse off.

Extraction or dissemination of information are not inherently good or bad. But they make for communication settings with inherently unequal parties, which is important for the interpreter to keep in mind. The interpreter provides a service that helps A achieve their objective. Granted, B might be able to subvert the situation by refusing to supply/receive information or providing inaccurate or false information, but the bigger the power asymmetry between A and B, the smaller is B’s ability to interfere with A’s agenda. I would argue that, in this constellation, the interpreter is not impartial. Rather, in aligning her services with the objective of the communication scenario, she will quite naturally side with A, who is also generally the stakeholder employing her.

Information extraction and dissemination are also the scenarios that are most likely to “turn ugly”, with torture and indoctrination as extreme examples. In these cases interpreter impartiality is a moot point: Who would argue for impartial and accurate interpretation of statements extracted from a person being tortured?

We can, of course, take the easy way out and simply argue that the people translating during torture, enhanced interrogation or cultic indoctrination are not interpreters, or that codes of ethics only apply to trained professionals working in specific contexts. Nevertheless, professional interpreters have in the past expressed their solidarity and identification with colleagues (“colleagues”?) working for the military in active conflict zonesmaybe without fully understanding the limitations of interpreter impartiality in these complex, volatile and often dangerous environments. Furthermore, there are now numerous initiatives geared towards training and professionalization of interpreters working in conflict zones, as such, we cannot simply exclude these individuals from the scope of interpreting ethics.

Codes of ethics for conference interpreters tend to limit the question of impartiality to the attitude of the interpreter during the interpreted event. As such, these codes of ethics seem to presuppose a mutual willingness of parties to communicate, which is often not the case in other contexts. The relevance of these codes in situations where there is a strong power-asymmetry between parties or where interpreters are necessarily also stakeholders is therefore limited. Codes of ethics for community interpreters, on the other hand, have gone beyond the interpreting scenario to address the social responsibility of language intermediaries: part of the community interpreter’s mandate is to help migrants or minority speakers gain access to public services and institutions.

Indeed, sometimes, the decision to interpret in a certain context is already a decision to take sides in a broader ethical space.

Conceptualising interpreter impartiality in such contexts requires an engagement with broader ethical questions. That is something I did not find in the “Conflict Zones Field Guide for Civilian Translators/Interpreters and Users of Their Services” published by AIIC, FIT and Red-Tmaybe because of the authors’ proximity to conference interpreting. Impartiality is listed in the Field Guide as a principle and described as follows:

“Regardless of who engages you, serve all parties equally, without expressing your opinions or sympathies. You cannot be an advocate for any cause and must declare any conflict of interest.”

However, under the heading “Protection”, the same document informs interpreters that they “are not required to wear a uniform, unless [they] consent to do so“. Note that a uniform is not the same as unmarked protective gear, such as the bullet proof vests often worn by civilian journalists working in conflict zones.

I cannot help but wonder:

If wearing a US Army uniform in Afghanistan does not make you the advocate for a cause, then what does?

If the act of wearing a uniform cannot be interpreted by a civilian interlocutor as an act of taking sides and expressing an opinion, then what can?

Information, as interpreters are well positioned to know, flows not only through words. Our presence, our demeanor and our physical appearance also communicate a message. Let us not forget the fact that the military uniform is designed for the specific purpose of distinguishing military personnel from civilians and that, from the perspective of the law of war, there a reasons to doubt that a military uniform provides “protection” to an interpreter:

“The military uniform distinguishes the members of armed forces from the rest of the population. In international armed conflicts, members of the armed forces can lawfully take part in combat on the battlefield. Inversely, they can be lawfully attacked. The absence of a military uniform usually indicates that a person is a civilian, is therefore not allowed to perform military functions and must not be attacked.” (Source: Pfanner, 2004: 94

Furthermore, “civilian” interpreters working for the US military in Afghanistan for instance have gone far beyond wearing a uniform:

“Being a translator in a place like Kandahar conveys a distinct isolation; the high salary is coveted, but many feel that such work with the foreign soldiers is tainted. Working with the Special Forces is doubly so. Interpreters are not supposed to be armed, but the U.S. Special Forces have largely ignored those regulations.” (Source: Rolling Stone, 6 November 2013)

The quote above refers specifically to locally recruited interpreters, i. e. persons who are neither members of the armed forces nor US citizens but Afghans carrying weapons while also working as language intermediaries between the US Army and the local population. 

In sum, many difficult and painful ethical questions remain to be addressed with regards to civilian interpreters working in conflict zones. Despite their problematic status, we should of course not forget that some interpreters have been killed when left behind in their home country by nations who benefited greatly from their services, yet then refused to grant them asylum. I am not defending this decision nor minimising the human suffering that has resulted from it. However, my argument is that the status of these individuals from the point of view of professional interpreting ethics is problematic, in particular regarding the practical possibility of impartiality.

Interpreters are ambiguous actors, who, depending on the nature of the interaction they find themselves in, have the ability to either reinforce or subvert established power relationships. More often than not, the decision to act impartially during the interpreted event only reinforces the status quo.

The term “interpreter” has been and will continue to be used to refer to individuals with very different backgrounds, levels of training, job descriptions and roles. Professional codes of ethics are useful for specific sub-groups of interpreters (conference, community, court, humanitarian, military) and their strength resides precisely in their specificity. However, placing the question of interpreter impartiality in a wider context spanning different groups of interpreters and types of scenarios helps us analyse the wider implications of impartiality.

In particular, we need to take into account that interpreters might work in an institutional setting that is in itself highly partial and that impartiality at the micro-level of an individual interaction might not mean much when communication primarily serves the purpose of furthering an agenda.

In short: Interpreter impartiality is about much more than what the interpreter says or does not say. It depends on who they are, who their employer is, what kind of communication scenario they are working in, what they wear, and why they are there to begin with.