Ethics of archival research on political violence

Mapping the field of ethics in archival research

Much of the guidance scholars get regarding ethical conduct of research is based on the standardized 1979 ‘Belmont criteria’, which state that human subjects must provide full consent to their participation in the research and understand fully any risks and benefits they may gain from it (Belmont Report, 1979). The Belmont criteria are often summarized as resting on three pillars: respect (informed consent and treating people with dignity), beneficence (conducting research to maximize benefit and minimize harm), and justice or fairness (avoiding prejudicial treatment of subjects).

As a baseline ethical guide, the Belmont criteria have already been shown to be inadequate for fully ethical conduct of research, especially research on conflict or on extremely vulnerable populations. For example, Belmont principles can often be meaningless in the field, if human subjects feel their consent is coerced or not fully understood, and if principles of confidentiality are perceived as disrespectful to informants who made the choice to publicly tell their story (Fujii, 2012). Further, the Belmont framework provides ethical protection only for subjects of research, but not for many others involved in the research process – interpreters, assistants, and researchers themselves. It also has no guidance for resolving multiple conflicting ethical dilemmas in the field, such as the tension between maintaining research integrity and answering to broader ethical challenges researchers may face in the community they work in.

Belmont principles seem even more inadequate for archival research, which often involves investigating the lives and activities of human subjects who are no longer alive. The dead cannot give consent, but even the other two pillars – beneficence and justice/fairness – become problematic. How do we judge the benefit of our research to a dead person and how do we measure harm done to the dead, especially those who led private lives and did not anticipate becoming a focus of anyone’s future research? How do we know what part of their story the dead would like to see told, in what context, and by whom? And why does this matter for the ethical integrity of scholarly inquiry?

These questions have been debated in academic disciplines that routinely rely on archival research. Historians have long discussed how to protect individual privacy when researchers discover and publish materials on someone’s life without their permission, consent or desire, especially on deeply personal matters such as, for example, sexual orientation (Schwarz, 1992). A lively debate was generated among historians over the question of ethics of using colonial archives for research on the colonized (Stoler, 2002). More recently, as archives and what we consider ‘archives’ are increasingly becoming digital, historians have grappled with the ethical use of materials that will inevitably make their way on to the Internet, potentially harming the subjects of study in unanticipated ways (Agarwal, 2016; Carusi & Jirotka, 2009). Scholars asked about the ethics of using personal archives and finding intimate and revelatory material, ‘as they touch not only the delicate papers of the dead, but the stories that might unsettle the sleep of the living’ (Cameron, 2001: 39).

These disciplines have also developed more elaborate guidelines about ethical conduct of archival research. For example, Anthropology instructs scholars to carefully investigate and report on the politics of archives, sources of funding, and ownership, as well as problematize the concept of data and records anonymization (Zeitlyn, 2012). History has moved to analyze archive as an object itself and has guided scholars toward deconstructing the authority of the archives, instead analyzing them as sites of power and narrative creation, and being self-reflexive about the researcher’s own status or other identity features (race, gender, class) that may allow some researchers access to some archives and not to others (King, 2016). Some historians have suggested that scholars need to include their ‘arrival stories’ in the archives in a similar fashion to how anthropologists report their ‘arrival’ to the field (Dirks, 2002, in King, 2016: 27).

Political science as a discipline that, fundamentally, is concerned with power is immediately well positioned to scrutinize archives as locations of power. Its methodologies and scope of inquiry, however, benefit from a more discipline-specific treatment. Some of the more obvious differences are in, for example, political science’s comfort with generalizability, with using local data for more general and abstracted conclusions. A more practical concern is the issue of political science training, which lacks a deeper engagement with issues of research ethics. The concerns and suggestions I develop below are therefore derived from a conversation with anthropological and historical approaches to archival ethics, but are also keenly aware of specific needs and challenges political scientists may encounter when doing archival research.

It is quite clear that political science has remained largely silent on this issue. In various methods guides for political science, there is barely a mention of ethics of archival research, other than the ethics of conducting research in the archive itself, which boils down to something like, ‘researchers must always treat archives with care: they must be left in the state we would like them to be when we find them’ (McNabb, 2010: 309). One of the main volumes on archival research in political science is virtually silent on the subject of ethics (Frisch et al., 2012). Political scientists who did think deeply about archival research still thought about it from the perspective of the archival research method. They asked how to do archival research to get the best data, how to compensate for incomplete records, how to triangulate with additional sources when archives are incomplete (Tansey, 2007). The 2018 special issue on ‘New Findings in Conflict Archives’ in this very journal includes a number of excellent contributions that illustrate how much archival research can advance our understanding of political violence (Balcells & Sullivan, 2018; also Leiby, 2009) – but it does not reflect, even in passing, on issues of archival research ethics.

The questions that have been raised are important questions about interpretation – how much value to place on archival documents that look like they were a product of consensus, when in fact they could have been a product of disagreement or conflict? What to do about missing documents, never produced and stored in an archive because the actors did not think they were important at the time? Most archival documents political scientists use would contain official documents which could have been results of intense negotiations, controversies or dissent – which the final, archived official document would not disclose (George & Bennett, 2005).

When discussed at all, archives are mentioned mostly as sources, with a warning to the researcher to be careful to avoid their ‘potential motivated and informational biases’ (Bennett & Elman, 2007: 183). Discussing methods of qualitative historical analysis, political scientists have warned against relying on secondary sources and taking the historians’ word for what is in the archives, and instead conducted our own independent archival research with different questions in mind (Larson, 2001; Thies, 2002).

All of these concerns, however, had to do with the validity of archival research. None of them with its ethics.

Developing quite independently from these methodological concerns, a growing body of work in political science has explored the ethics of conducting fieldwork during conflict. Especially relevant to the core concern of my article, conflict researchers have established the inadequacy of the Belmont principles for research during conflict (Campbell, 2017). Carefully crafted research questions and human subjects protocols may need to be completely revamped in a conflict situation and adjusted to the new circumstances (Mazurana, Jacobsen & Gale, 2013). Researchers' preconceived notions of risk may be completely inappropriate and risk to human subjects on the ground may be much higher than anticipated (Romano, 2006; Wood, 2006). Almost no guidance exists for maintaining the physical and emotional well-being of researchers working in extremely dangerous or traumatic environments (Sriram et al., 2009) or for protecting local assistants (Cronin-Furman & Lake, 2018).

The principal motivation behind this article, then, is to move the discussion of research ethics in political science forward by introducing the complexity of archival research ethics.

Archives as sites of research

Historians have grappled for a while with the question of what constitutes a ‘legitimate’ archive or source. While traditionally, ‘legitimate’ archives were official, public places where people of privilege (white, male, wealthy) deposited their documents, or the archives considered only those kinds of people worthy of archival files, History has begun to embrace other kinds of archives and give other kinds of unofficial voices (women, minorities, the poor) the right of narration. This has allowed historians to ask new questions about what kinds of actors drive historical events. By expanding our inquiry into the world of non-official, personal archives (including quotidian sources such as diaries or letters), political scientists can also begin to ask more nuanced questions about the roles of various actors in politics and political implications of their everyday practices.

In the course of research for my book on Holocaust remembrance in post-communist Europe (Subotić, 2019), I have come across – and purposefully looked for – archival material, some previously published, some not, from victims, perpetrators, and bystanders during the Holocaust in Eastern Europe. I have used diaries, letters from camps and ghettos, postwar trial proceedings, oral testimonies, and interviews (with descendants).

My fieldwork involved multiple and extended visits to three countries that make the core of the study – Serbia, Croatia, and Lithuania – over the course of two years (2016–18). I also visited many other sites, museums, and memorials across Europe as well as in the United States. I used multiple archives and multiple types of archives. Since I speak Serbo-Croatian, I used primary sources for these two country case studies, including material from the Archives of Yugoslavia, Archives of Belgrade, the Jewish Historical Museum, Institute for Recent History of Serbia, Croatian History Museum, Documenta Center, Jewish Community of Zagreb, and the National Museum for Contemporary History of Slovenia. Since I don’t speak Lithuanian, this case study relied partly on secondary sources and existing historiography of the Holocaust and its remembrance practices in Lithuania. However, I also made extensive use of primary sources such as video oral testimonies from Lithuanian survivors, bystanders, and perpetrators available in the video archives of the US Holocaust Memorial Museum in Washington, DC, and the Florida Holocaust Museum in St Petersburg, FL, which were translated into English.

Since I relied on first person accounts, diaries, letters, and testimonies, I used public archives where many of these materials were stored, but I also had access to a number of personal archives which contained sometimes exculpatory but sometimes implicatory documents about various family members’ conduct during the Holocaust. Ethical issues in those cases were much different and even more fraught. Information I discovered would be published for the first time and descendants may have given me access to their private archives without the expectation that material damaging to their family members’ reputation would catch my eye. Sometimes, they knew what documents were included in the private archive but they did not know or fully understand the exact historical and political context which would give these documents a much different flavor.

Ethics: Two examples from research on the Holocaust

Interpretation

The questions of ethical interpretation of archival documents are acute because the archival researcher encounters documents and objects that are often completely decontextualized from the social or biographical context in which they were once embedded. The researcher then has to interpret the material without that particular historical context in mind. What can one piece of paper tell us without a story around it? What do we do with someone’s party membership card? We may know this person was a member of a party (communist, Nazi, etc.) and this may be incriminating, but we don’t know much else. Was party membership coerced, was it voluntary, was this truly a choice? What inference can we make out of this one document? What other corroborating evidence do we need? Is this one document enough to make an argument about this person in our writing? And how do we, or should we, avoid judgment? This issue is particularly severe in research on horrific crimes, such as crimes of the Holocaust, when victims themselves have not always acted honorably, trying to survive (Brown, 2010).

The role of the researcher in interpreting archival documents is to begin by determining the historical context in which this document was produced (as well as transferred and disseminated) and explaining it in detail in the written research. If these are, for example, court proceedings minutes – what was the context of this trial? Was this a show trial, was it a trial with due process, or something in between? The researcher then needs to decide how much stock she should put in a document like that, or in witness testimony from such a trial. We need to embed these archival materials within the historical and political context of the time and evaluate the documents against other factors such as, for example, the severity of violence and repression in society. An archival document alone is not only meaningless, it can be highly deceptive and its interpretation ethically challenging.

Here it is useful to include an intervention by Lee Ann Fujii, who has argued that an integral part of any analysis of testimonies is not only an evaluation of their veracity, but also the ‘meta-data’ – ‘spoken and unspoken expressions about people’s interior thoughts and feelings’ (Fujii, 2010: 232). This intervention itself builds on the classic work on ‘hidden transcripts’ by James C Scott, who early on alerted us to the value of studying different expressions of resistance to power, the infrapolitics of the dominated that goes undetected by those in power, but should be the focus of any research on political violence (Scott, 1990). Fujii referred to live interviews, not archival material, but these are the same issues of tone, choice of words, unspoken subtext. This archival ‘meta-data’ can also include information about the provenance of archival documents, ownership, storage, presentation, condition, and all additional information that contextualize the piece of archival document or artifact and more comprehensively paint its biography.

The issue, then, becomes one of veracity of sources. Are all narrators equally reliable, and at all times? In using personal archives, there is a danger in non-reflexively relying on intimate documents, such as diaries, which may seem like the most authentic sources, but themselves are also part of interpretation. People are complicated; their narratives about others and about themselves are also complicated and not always truthful. The fact that the narrator is no longer alive makes inferring the meaning of her words that much more difficult, but critical to the enterprise. A more reflexive approach to archival research would include a consistent scholarly re-evaluation of our own process of archival selection and curation – an awareness of the politics of archives, the provenance of materials, and the ethical dilemmas of which discovered documents to publish and how best to justify and contextualize that curatorial choice.

Even more important is the veracity of the researcher’s interpretation and the ethics of this interpretation. In researching extreme violence, there is an incentive to document the most gruesome episodes, the most descriptive suffering, the most painful stories. The incentives for researchers to search for the most horrific detail may be scholarly (finding ‘the smoking gun’, the most evocative document) but they can also be professional (the payoffs in scholarly reputation of studying extremely difficult subjects and enduring emotionally difficult research). But scholars must avoid visualizing experiences of our subjects by making interpretive leaps and engaging in evocative imagined scenarios that the personal testimonies only hint at but do not fully convey. There is a temptation to describe events based on personal testimonies in a way to attract the reader’s attention in the dark, gruesome, and grisly world of extreme violence. The researcher becomes a curator of pain – we search for the most horrific detail, the most traumatic story, the most devastating letter, the most touching material object.

These are complex ethical issues Holocaust scholars have already explored in great detail. For example, how to contextualize and report on narratives of sexual violence in the ghettos and camps, events that survivors and witnesses described very differently as the years went on (Hájková, 2013). Other ethical challenges Holocaust scholars encountered are the posthumous classification, numbering, and anonymization of large databases of Holocaust victims – a process that disturbingly mirrors the processes of technocratic dehumanization done to these victims by the Nazi regime (Einwohner, 2011). Crane has argued that researchers analyzing atrocity photographs, especially photographs taken by the perpetrator of atrocity, are inevitably looking at it from the perspective of the perpetrator and so themselves participate in the atrocity after the fact (Crane, 2008). Jacobs worried about ethical implications of sexualized images of dead Jewish women she collected in her research on gender and Holocaust memorials (Jacobs, 2004). Beyond the Holocaust, scholars considered the problems of reproducing graphic photographs of lynching in the US South (Reinhardt, 2012).

What these examples demonstrate is that scholarly interpretation of archival material – especially material related to violence – is not just a matter of research validity; it is a matter of research ethics. It is only with comprehensive contextualization of these archival materials that we can not only accomplish scientifically valid inference, but also begin to fulfill an ethical duty to the subjects of our study. Finally, another reason to take seriously the ethical dilemmas of archival research is that unethical practices may ‘spoil the field’ in the same way unethical field research may ruin the field site for future research in the same location by other scholars.

Harms and benefits

Research on human suffering is so difficult that scholars writing on a much different topic – refugees – have already argued that it should only be justified if it contributes to the end of that suffering (Jacobsen & Landau, 2003). But this advice should still hold when our research is on past suffering. The harm to our dead subjects can continue and we need to think through ways to avoid or alleviate it.

The existing ethical guidance, as institutionalized in the Belmont Principles, is concerned with rights of the living. So what rights do the dead have? What is our ethical obligation to them?

Legally – at least in the United States – the dead do not have many rights when it comes to archival materials (McKee & Porter, 2012). They have even fewer rights if some of their personal archives – or parts of them – have already been made public (e.g. personal letters). A further question is what is public and what is private in an archive? If the material in the archive was placed there not by the author but by someone else after their death, who determines if these are publicly available and could be published (McKee & Porter, 2012: 69)?

But even if the dead do not have legal rights, we should still have ethical concerns about their dignity – avoiding, for example, the documented cases of researcher disrespect of indigenous peoples’ spiritual and cultural practices (Louis, 2007; also Christen, 2011). In the context of archives, the dead can no longer provide context, they cannot assert their wish that the documents not be made public, they may have changed their mind about something they wrote, and we would be publishing their thoughts that may not be representative of what they actually believed. In the context of the Holocaust, archival materials may uncover written testimonies produced under unimaginable distress (in camps, in prison, at shooting sites). These may be historically important, but the circumstances of their production color the text. These people wrote in extreme conditions, and they would have almost certainly produced different text, different emotion, and different message under different circumstances. Not all text should be evaluated the same. Not all text is usable text. Not all text is for the researcher’s gaze.

Archival material can cause significant reputational damage when unflattering documents, pieces of writing, or testimonies are uncovered. To what extent should we treat differently documents about public figures – and how public – vs. ‘regular’ people? In the context of conflict, however, everybody is implicated – everybody is a victim, a perpetrator, a helper, or a bystander, many are some combination of these roles. How do we sort out the reputational damage done to people by uncovering their role in the conflict – even if the role is a passive one, as a bystander? Is documenting bystanding of significant enough public value to accept reputational damage? Further, reputation may be damaged even if we do not publish the document – just seeing it ourselves may forever change the perception we have of someone. A further consideration is for scholars to think through the balance of harms and benefits not only in terms of current harms/benefits, but also in the light of possible future harms, such as when, for example, a changing political situation deems information seemingly innocuous at the time much more dangerous or damaging to the subjects later on. There is a cost in getting it wrong. But there is also a cost in telling the truth.

Further, in the context of archival research on political violence, should the same ethical standards apply to archival material on victims and perpetrators? It seems sensible to reserve the highest ethical standards for victims of violence, as the stripping of agency violence has produced is not something we should replicate in our research – we should be especially concerned with maintaining the dignity, humanity, and voice of those whom violence has deemed voiceless. The ethical calculation may be different in the case of perpetrators, not least because their actions almost always included a degree of choice (even if sometimes that choice was extremely limited). It is this element of choice that burdens perpetrators with responsibility, and it is this responsibility that ethically allows scholars to treat perpetrators in the archives with a different degree of protection than the one granted to victims. However, most conflicts we study are not so neat – they implicate victims who may have also at times acted as perpetrators, and often perpetrators who at some point in the conflict become victims as well. It is only through very careful curation of available documents that we can piece these stories together and make ethical judgments about how to properly contextualize the material evidence in front of us so that we maintain both the integrity of the scholarly project and the research ethics itself.

And here it is possible that the two standards will clash – what maintains the integrity and validity of the research may be in conflict with our ethical concern about the archival evidence we use as evidence in this research. While ethical concerns should in general be of higher order than the success of a scholarly project – there is always another article or book to write – in the context of research on political violence where the stakes are so high and the events we analyze have already caused so much damage the ethical concerns about harm to subjects of our research should be that much more of a priority.

These examples point to the inadequacy of treating all archival research the same, and also assuming, as much of political science does, that archival research is done in official, public archives. As a discipline, we should think more creatively about what kind of archival sources we could use in addition to official state archives, but also what ethical problems we might encounter as we move from the public to the private archival space.

When using private archives, consulting with the descendants about the disclosure of unflattering materials may be wise (and sometimes legally required), but this brings about its own set of problems. Descendants may not want to be associated with the beliefs or acts of their ancestors, or they may want to obscure what their ancestors did or said. Are we ethically obligated to accept their wishes? This question is particularly pertinent in research on political violence where bringing those previously undisclosed acts to light may itself be a meaningful process for victims or their own descendants. Further, analyzing beliefs and meaning people gave to events around them – however disturbing they may sound to the descendants (like Jovanović’s self-identification as a ‘pure Aryan’ felt to his daughter) is of key significance to our research.

The dead can neither consent nor reap benefits from our research, but there are ways in which we can conceptualize groups that could be owed benefits. In my research on Holocaust remembrance, I interviewed many children and grandchildren of Holocaust survivors, as well as many staff members of various Jewish museums, memorial centers, and community organizations. Without exception, they were grateful for the interest in their relatives and the history of the Holocaust in their own communities and understood my research to fulfill the Holocaust remembrance mandate to ‘never forget’. This communal benefit of remembering, and remembering with dignity, is a way to reconceptualize both the benefits and harms of historical research as being broader than those of the individual dead subjects and instead affecting a community as a whole. Here scholars should try to identify and engage with the continuous community of victims of violence (through victims’ groups, or civil society activists, or human rights organizations) to share the research and discuss sticky ethical dilemmas that have arisen during the research process. With its narrow focus on specific and individual human subject criteria, existing Institutional Review Board (IRB) guidelines are missing larger societal benefits and harms that are hugely ethically consequential.

The paradox, of course, is that in its exclusive focus on specific individual harm, IRB guidelines can also lead researchers into avoiding publishing conclusions that would be upsetting for participants – an outcome that is clearly problematic from the perspective of scholarly inquiry, especially on issues where more, not less social confrontation may be needed (Marzano, 2012). For example, Mila Dragojević’s study of ‘amoral communities’ at times of civil war presents an ethically challenging issue of how to report on archival findings that may implicate the descendants’ parents/relatives – or even themselves – in violence (Dragojević, 2019).

The politics of archives and politicization of research

In conducting research for my book, the access to archives was mostly without much difficulty. Some archives were extremely well organized, some smaller archives were open and accessible but poorly organized and difficult to use and cross-index. I was unable to get access to materials from two institutions I approached – the Institute for Contemporary History in Belgrade and the International Commission for the Evaluation of the Crimes of the Nazi and Soviet Occupation Regimes in Vilnius, Lithuania.

On my visit to the Institute for Contemporary History in Belgrade in July 2016, upon hearing that I was looking for the Institute’s own old (1991) publications on Yugoslav citizens deported to Auschwitz, one of the researchers rejected the premise of my research and claimed that ‘there were no Yugoslavs at Auschwitz’. ¹ While I was shocked at the bluntness of this denial, it did not come as a complete surprise. This Institute has become a hotbed of historical revisionism in Serbia and its most prominent historians have brought a series of lawsuits attempting to overturn convictions for collaboration of various Serbian World War II quislings (Dragojlo, 2016).

In Vilnius, I never got to visit the International Commission for the Evaluation of the Crimes of the Nazi and Soviet Occupation Regimes, as my multiple requests for a visit (and an interview) remained unanswered. While I was slightly put off by the unprofessionalism of ignoring scholarly inquires on the exact topic that is the Commission’s mandate, as in Serbia, I was not completely surprised by this. The Commission was initially established by the Lithuanian president in 1998, but has been embroiled in controversy since 2006, when one of its high-profile international members Yitzhak Arad, an Israeli Holocaust survivor and historian and a former director of Yad Vashem, was investigated by the Lithuanian prosecutor for alleged ‘crimes against humanity’. ²

The tendentious and highly political investigation of Arad was dropped in 2008, but the Commission never quite recovered from the scandal and the subsequent mass resignations of its many international members (Karn, 2015). The Commission was renewed in 2012 and has restarted some of its activities, but it has kept a much lower public profile amidst increasing negative international press that criticized its framework of studying both Nazi and Stalinist terror under the same rubric and downplaying extensive Lithuanian participation in the Holocaust.

The context of these two institutions is important because it raises the issue of the politics of archives and how we as researchers adjust to them. Archives are political institutions – the decisions about what to store and what to discard, what to classify and how, what to make accessible and to whom, are political (Tesar, 2015). These are not ‘neutral’ spaces – instead, they are gatekeepers of memory (Jimerson, 2003). By making decisions about who is important and worthy of an archival folder, and who is not, who is worthy of remembrance, and who is not, they themselves are ‘sites of power’ (Jimerson, 2009). Archives – in conflict but also non-conflict spaces – can take on a role of a censor, making some sensitive information accessible, and not other (Moore, 2010). While political science was late to the discussion of the politics of archives compared to other disciplines, the newer research using archival material in conflict studies has begun to address this issue more directly (Balcells, 2017; Balcells & Sullivan, 2018).

It was very clear to me that the two institutions that did not welcome me did so for political reasons. The integrity of my research project, however, did not much suffer from these absences, as I was able to find necessary material elsewhere, with the help of archivists at other institutions. I will never know, however, what else I could have discovered in the two archives had I been given access. In fact, the lack of access itself became a ‘data point’ in my research, as I described my encounter with these institutions in the book and contextualized their political positions.

I further encountered politicization of archives in a different way. While working on the Lithuanian case, I needed to look for archival information about a well admired national hero of Lithuania, who was revered for his fight against the Soviet Union. Recently, however, documents were uncovered that implicated him in the Holocaust of Lithuanian Jews (Balčiūnas, 2014). Lithuanian institutions have firmly defended his innocence and rejected out of hand the documents provided, even though they were complemented by similar implicatory material from the archives of the US Holocaust Memorial Museum. In the extremely anti-Soviet and now anti-Russian political environment in Lithuania, these archival collections were not considered usable because they were based on documents produced (or collected or classified) by the Soviet authorities in the aftermath of the war in 1945. The division among historians of the Holocaust on this issue was stark – non-Lithuanian experts routinely used these archives, including many documents about local perpetrators of the Holocaust who were tried and, in many cases, executed for collaboration by the Soviets after the war. Many Lithuanian scholars reject these documents and also reject most scholarship based on them.

Again, the issue with archival research here is political. And as I did in the case of closed archives, I discussed this controversy in the book and placed it within the context of contemporary Lithuanian politics and attitudes about the past. Of course, this is not to say that these problems are particularly unique to either Lithuanian or Serbian archival politics. The contemporary memory war in the United States over the legacy of the Confederacy and the US Civil War is an obvious example of politicization of both history and commemorative practices, as well as archival and museum access, especially in contested spaces in the US South.

Recommendations for ways forward

Archival research – like all research – requires reflexivity, but this concern for ethics in archives needs to be more central to the enterprise than it currently is. At all stages of the project – from identifying archives and conducting research, to writing and publication, scholars need to make ethical choices at multiple ‘ethically important moments’ that arise during research (Guillemin & Gillam, 2004). As Fujii so passionately advocated, ‘ethics is an ongoing responsibility’ (Fujii, 2012: 717; also Schaffer, 2015, ch. 5). Our ethical obligations do not end when the research is over. What we publish is the key question of our scholarly ethics (Knott, 2018).

Ethics of research are so important that they should not be an afterthought, nor should they be left to only some researchers more interested in or comfortable with ethical reflexivity. Ethics of archival research could and should be taught to our students. We help our students navigate the labyrinths of the IRB process, but we almost never discuss the ethical choices they have to make when they do archival work – choices that go far beyond the IRB. There is no reason why a serious engagement with archival ethics should not be a routine addition to syllabi on research methods.

In my own research on Holocaust remembrance, I have encountered ethical challenges that are immense and diverse – from the importance of getting it right, to concerns about exploitation (Jacobs, 2004), to the intricate issues of further dehumanizing and depersonalizing Holocaust victims by anonymizing, numbering, and classifying their stories (Einwohner, 2011). These standard practices that university IRBs require are an extremely problematic fit for Holocaust research, as they turn ‘the subjects into objects’ (Jacobs, 2004: 228). On the other hand, many political scientists will object to this critique, arguing – also persuasively – that producing research based on a de-identified database of victims may allow for powerful pattern identification and further contribute to our understanding of causes, methods, or consequences of violence.

Research on the Holocaust has brought in sharp relief the conflict between the ‘do no harm’ mandate of the Belmont Principles and the ‘never forget’ mandate of Holocaust scholarship and memory (Einwohner, 2011: 423). One way of correcting for this is to consult with descendants of those we study, but also consult with their ‘continuous community’ (McKee & Porter, 2012: 74) – which is why consulting with Jewish communities in all the countries I did research in was essential to my own scholarly process, while being quite cognizant of the fact that communities are never homogenous and do not speak with one voice.

Ethical concerns, however, do not end at the research stage. They continue throughout the project cycle, and can be especially acute at dissemination and publication stage (Sriram, 2009). Different audiences will use the published product for their own political purposes, and the author has very little, if any, control at all once the publication is out (Fujii, 2012: 722). Since my current project is on changing Holocaust remembrance, as well as increasingly frequent efforts at Holocaust revisionism across Eastern Europe, my research is already fully politicized, even before the book hits the presses. My task then, in writing, is to anticipate this politicization, present the arguments against it, and be prepared to continue engaging in a public dialogue long after the publication is out.

Finally, it is worth briefly addressing the ethical openness in the era of increasing calls for data transparency and replicability, such as the DA-RT movement. ³ DA-RT proponents are interested in the transparency of qualitative research, which they argue can be bolstered by possibility of replication, including replication of archival research. Even they, however, admit that interpretation of archival documents is going to be different among different scholars (Elman & Kapiszewski, 2014), a key point my article emphasized.

As many DA-RT skeptics have already indicated, calls for replicability of archival research are meeting strong resistance on both methodological and practical, feasibility grounds. ⁴ But some archival researchers have also pointed out important ethical issues of handling private archives. The DA-RT requirement to digitize and upload archival material would lead to a huge breach of trust with the people who have entrusted their or their family members’ very private documents to a particular researcher (either because of a personal relationship or trust in the research project’s integrity or significance). But they did not consent and very likely would not consent to this material being available for all the other researchers to look at, analyze, share and replicate, without the personal context that initiated the access in the first place (Gaikwad, Herrera & Mickey, 2017: 12).

Then there is the issue of context – as one of the qualitative researchers argued in a DA-RT deliberations report, when we work in archives we often read hundreds and hundreds of documents for context, tone, identification of various characters, and their importance, but only end up using one or two for publication. Replicating all the material we use would be excessive and almost impossible, but replicating only the two specific documents would be meaningless to another researcher without the context that guided this document choice (Gaikwad, Herrera & Mickey, 2017: 13).

DA-RT guidelines, however, are still interested primarily in the integrity of the research and only tangentially – and only in critical evaluation of its criteria – have issues of ethics arisen. Ethics of archival research remain without proper guidance.

In conclusion, in choosing archives to access, we should investigate the politics of the archives selected, the provenance of their collections, and disclose these findings in our research. In interpreting archival material, we should make extra effort to contextualize the documents and investigate and report on the historical and political context in which this document was created, who created it, and for what purpose. In making decisions about document selection, we should ask if the most extreme document is in fact the most representative, or can another document, less graphic, be used to make the same point. We should work hard to preserve the dignity of dead subjects – to avoid essentializing and reducing them to one dimension. We should provide as much complexity as possible to the subject who can no longer speak for herself.

My final appeal is to treat archival documents not only as ‘evidence’ or ‘data’ that we use to write about, but as information about something that really happened to people. When we do research on violence, these documents are not about an abstract ‘conflict’, but are evidence of a living, multigenerational trauma that we as researchers, if we don’t pay attention to deep ethical questions, can further contribute to and deepen.