Keywords

10.1 Introduction

This chapter offers a case study of the Belfast Project, set up by Boston College in the US in the early twenty-first century. The Project’s remit was to collect and store accounts of the late twentieth century conflict in Northern Ireland commonly known as ‘the Troubles’. People who provided accounts for the Project were given written guarantees that not only their own accounts, but also the Project itself, would be kept secret until after their deaths. However, some years later, information about the Project was made public by the Project’s own director while some of its participants were still alive, with complex and far-reaching consequences.

Researching violent conflicts inevitably raises difficult ethical issues (Brigden and Gohdes 2020). This chapter covers some of the ethical issues raised by the Belfast Project and, in so doing, explores the inherent tension between research ethics and legal ethics (Adams 2014).

10.2 Background

10.2.1 The Troubles

The Northern Ireland conflict of the late twentieth century, known as ‘the Troubles’, was a political and sectarian conflict over whether Northern Ireland should remain within the UK, or leave and form a united Ireland with the Irish Republic. The Troubles began in the late 1960s and ended around the late 1990s or early 2000s. It is difficult to date the conflict precisely, as there were several significant events and developments in the late 1960s, each of which could be held to be the start of the Troubles, and there is a similar picture in the late 1990s/early 2000s for the end of the Troubles (Fitzduff and O’Hagan 2009).

The Troubles is not an isolated conflict. In fact, as long ago as the early 1600 s, Scottish and English settlers colonised the north-east of Ireland, forcing the Irish people who lived there from their homelands (Fitzduff and O’Hagan 2009). This colonisation is known as the Ulster Plantation. The settlers were Protestant and had strong cultural ties with Scotland and England; the Irish people were Catholic and culturally Irish. The inequalities between these two sections of the population were never redressed and non-violent and violent conflicts have occurred regularly in Northern Ireland over the last four centuries (Fitzduff and O’Hagan 2009).

In 1801 the United Kingdom (UK) was formed. At that time the UK was made up of England, Scotland, Wales and the whole of Ireland. After the First World War there was an Irish war of independence which led to the formation of the Irish Free State, now known as the Republic of Ireland. Northern Ireland as it exists today was formed by legislation passed in 1921 partitioning the island of Ireland (Fitzduff and O’Hagan 2009). This placed the counties of Antrim, Armagh, Down, Fermanagh, Londonderry and Tyrone in Northern Ireland, and the other 26 counties in the South.

The name ‘Londonderry’ is contested, with most of those supporting the Union favouring the name Londonderry, while most Irish nationalists prefer Derry. This is just one example of the polarisation of society in Northern Ireland; at a more macro level, ‘peace walls’ were built during the Troubles along streets in several towns and cities to keep nationalists and unionists apart in an effort to reduce violence (McGrade 2017).

In the early 1960s a number of initiatives were developed in Northern Ireland to combat discrimination against Catholics. By 1967 these had cohered into a civil rights movement with the formation of the Northern Ireland Civil Rights Association (NICRA) which aimed to secure the rights of all citizens, regardless of their religious or political affiliations, through public protests on the streets (McKenna undated). However, unionists saw NICRA as being a republican, not an egalitarian, organisation, and accused it of working to undermine the state of Northern Ireland. Civil unrest reached a peak in summer 1969 and after the ‘battle of the Bogside’, the name given to three days of violent confrontation between Catholic residents and Northern Irish police in Derry/Londonderry, resulting in the British Government deploying British troops to keep the peace on the streets of Northern Ireland, where they remained until the early twenty-first century. In 1972 the civil rights movement ended, and NICRA was disbanded when London suspended the Northern Ireland parliament and took control of the region from Westminster.

Thousands of people were killed in the Troubles and tens of thousands were injured. Those who were killed included 1,785 civilians, more than half of whom were killed by loyalist paramilitaries, and over 1,100 British soldiers, most of whom were killed by republican paramilitaries (Sutton undated). Although peace then prevailed for the most part for over 20 years, the post-Brexit uprisings demonstrated that feelings on the issues could still be very strong.

10.2.2 Archives

An archive is a collection of documents or other records of historical interest, and/or the place where such a collection or collections are stored. This storage may be in digital or bricks-and-mortar spaces. Archives are usually seen as neutral, inactive resources which people can use at will for academic, cultural, or recreational purposes. However, archives are not value-free, and not all archives are accessible to everyone. Archives can also be seen as a form of power, an attempt to control the past by privileging some stories and marginalising others (Schwartz and Cook 2002), and/or by placing restrictions on who can use, and when and how they can have access to, the archival material. Archives ‘are not passive storehouses of old stuff, but active sites where social power is negotiated, contested, confirmed’ (Schwartz and Cook 2002, 1).

Archives are often associated with institutions. Institutional archives usually have policies to govern their operation. However, these policies vary a great deal between different institutions (Wood et al. 2014). There are no overarching rules or guidelines governing the policies of an institutional archive so, in essence, an institution can write its own policy.

Archivists know that in post-conflict situations, ‘documenting or disclosing the provenance of materials may put those who created, collected or provided those materials at considerable risk’ (Wood et al. 2014, 412–3). There is also risk that the materials, so carefully created, collected or provided, may be destroyed or relocated and so lost to history.

Collections of records that document any violent and systematic abuse of power may be known as ‘human rights archives’ (Caswell 2014, 208). These can include stories recorded by survivors of a human rights crisis. Creating, preserving, and using records documenting human rights crises, such as the Troubles, is a process fraught with political, ethical, legal and cultural challenges (Caswell 2014).

10.3 The Case Study

In 2001, the Burns Library of Boston College in the US set up an oral history archive focusing on the Troubles. The archive was known as the Belfast Project. The intention was to include perspectives of those on both sides of the conflict, to ensure accurate records that could not be lost or distorted by history. This was important because there was, and is, a high level of disagreement between loyalists and republicans about what constitutes ‘truth’ and ‘facts’ about the Troubles and related issues (Inckle 2015). The Project director was Ed Moloney, who had a background as an Irish journalist covering the Troubles, and in 1999 had fought and won a court case against a court order he had received to hand over some journalistic interview notes of interest to UK anti-terrorist authorities (Palys and Lowman 2012).

Moloney recruited insider researchers—two academics who were also convicted ex-paramilitaries, one Loyalist and one Republican—to conduct extended interviews with participants from their own side of the conflict (Inckle 2015). These were research interviews rather than journalistic interviews. The ‘Agreement between the Trustees of Boston College and Edward Moloney, Project Director, to Interview Members of Irish Republican Paramilitary Organizations and Provisional Sinn Fein Regarding their Role in the “Troubles”’ states that interviews are to be documented on audio or video tape and transcribed (Belfast Project Agreement 2001). Transcripts would be deposited in the archive, together with signed statements of authenticity, under an alphanumeric code for anonymity. The key providing the link between codes and names would be kept in Boston and could only be seen by the Project director and the librarian who managed the archive. These protections were put in place because the interviews would inevitably include secret information, such as accounts of criminal activities including bombings and murders, and that information would be dangerous for participants and others if it became public (Inckle 2015). Over the next nine years, the interviewers conducted over 200 interviews which were placed in the archive.

Boston College gave the researchers ‘Agreement for Donation’ forms for participants to sign, guaranteeing that the information they provided would be kept safely within the archive until after their death. Even so, pseudonyms were used and, as we have seen, careful processes to protect participants’ anonymity were put in place (Cardenas 2019). Furthermore, the whole Belfast Project was to be kept secret, with both interviewer and interviewee required to sign an agreement stating that neither would tell anyone else about the Project without permission from Boston College. The intention here was to ensure that participants felt able to give complete and truthful accounts of their experiences in the Troubles (Inckle 2015). The Belfast Project Agreement, signed by Ed Moloney in 2001, said ‘Each interviewee is to be given a contract guaranteeing to the extent that American law allows the conditions of the interview and the conditions of its deposit…’ but this potential limitation was not communicated to participants (Palys and Lowman 2012). In practice, nobody working on the Project, in the US or in Northern Ireland, had any idea that they could be forced to make interview transcripts or recordings available to a third party while interviewees were still alive (Breen-Smyth 2019).

In 2010 Ed Moloney published Voices From The Grave, a book based on interviews with two of the Project’s participants, loyalist David Ervine who had died in 2007 and republican Brendan Hughes who had died in 2008. The book was accompanied by a TV documentary with the same name. The Irish media became interested, which brought the Project to the attention of the UK legal authorities, who learned that the archive might contain evidence to help them clear up unsolved murders from the conflict. The UK Government asked the US Attorney General to subpoena Boston College to make them hand over all material related to two of the interviews, and the subpoenas were delivered in May 2011 (Palys and Lowman 2012). Delivery of one interview was straightforward because the participant had died. The other participant was still alive and, after much legal argument, Boston College handed over the material to the court for a judge to read and make the final decision. After due consideration the judge decided to release the material to the UK Government.

This action by Boston College did not contravene the Agreement the College signed with Ed Moloney, but did contravene the guarantee of protection given to participants on the ‘Agreement for Donation’ forms. Staff of Boston College argued that they had only guaranteed anonymity within the limits of the US legal system, not internationally, and so had to be bound by the judge’s decision. The action of Boston College had terrifying consequences for members of the Project team, some of whom feared for their lives while others were advised not to travel to Northern Ireland because the risk was too high (Inckle 2015).

Ed Moloney feared further subpoenas and suggested moving the archive to the Republic of Ireland, but Boston College disagreed. In August 2011 a second set of subpoenas were delivered to Boston College, this time asking for ‘any and all… information’ contained in the archive about an unsolved murder (Cullen 2011). More legal argument ensued, with Ed Moloney and one of the interviewers getting involved in filing motions and swearing affidavits to try to protect the participants. This time Boston College handed over half of the archive to the court for a judge to read and make the final decision. In February 2012 the American Sociological Association, and members of the Boston College Chapter of the American Association of University Professors, made public statements of support for the researchers. However, the judge gave the UK police access to material deemed to be relevant to criminal inquiries (Breen-Smyth 2019).

While it was good news for the UK criminal justice system, this had major consequences for several republicans and loyalists who had participated, or been named, in the interviews. In March 2014, the republican former IRA leader, Ivor Bell, was arrested and charged with soliciting the murder of Jean McConville in 1972. In April 2014, the republican Sinn Féin leader Gerry Adams was taken in for questioning and was released several days later without charge. Bobby Storey, another Sinn Féin leader, was arrested in December 2014 and later released without charge. In June 2016, the loyalist former Ulster Volunteer Force and Red Hand Commando member ‘Winkie’ Rea was charged with 19 offences including aiding and abetting murder and conspiracy to murder. All of these actions were based on evidence from the Belfast Project (Breen-Smyth 2019).

In May 2014, Boston College offered to return interviews to their originators on request. But this was too little, too late, and the legal processes continued. In July 2016 it was announced that Ivor Bell would stand trial, charged with ‘encouraging persons to murder Mrs McConville and endeavouring to persuade persons to murder her’ (BBC 2019). However, Bell had developed vascular dementia so was deemed unfit to stand trial. This led to a legal process known as ‘a trial of the facts’, in which the truth of the allegations against a defendant is examined rather than the defendant’s guilt or innocence of the crime with which they have been charged. As a result of this process, the judge, Mr Justice O’Hara, ruled that the evidence provided by the recordings was unreliable. In particular, the judge ruled that the researcher had asked leading questions, and that the promise of confidentiality, while designed to promote truth-telling, could equally have given the interviewee freedom to tell lies, distort the truth, or mislead the researcher (McKeown 2019). As a result, in October 2019 Ivor Bell was acquitted of involvement in the murder of Jean McConville.

10.4 Analysis

It is very difficult for researchers or participants to assess all the potential risks that may arise from doing or taking part in research. People are generally poor judges of risk, for a range of reasons such as having inaccurate or incomplete information, biases including optimism bias or availability bias, and the important role of context. Also, risk is particularly difficult to perceive when it is in the future because of the increased levels of uncertainty which mean the past is not always a trustworthy guide. Power dynamics, in particular, can change in unexpected ways, with potentially harmful implications for research participants and researchers (Parkinson and Wood 2015; Brigden and Gohdes 2020). Also, advances in technology can cause inaccessible or private information to become accessible (Bridgen and Gohdes 2020). As the case of the Belfast Project shows, there is even uncertainty about whether research institutions will keep their promises when facing external pressure (Thaler 2021). This level of uncertainty seriously compromises the principle of informed consent because it is not possible fully to inform someone about the risks they would be taking if they participate in research (Parkinson and Wood 2015). And it is not only researchers and participants who may be endangered by research, but also the ‘unintended research participants’ (Bridgen and Gohdes 2020, 256) or ‘non-consenting others’ (Mannay 2016, 123) who may be involved by being mentioned by participants without their knowledge or simply by having a stake in the research topic, such as by being on one side or the other in a sectarian conflict like the Troubles.

Some human rights archives ‘can play a key role in helping societies deal with painful pasts and build peaceful futures’ (Caswell 2014, 209). The Belfast Project had the opposite effect. The actions listed above, taken as a result of the Belfast Project, are only those we know about. The exposure of the Belfast Project will, if nothing else, have caused alarm and fear, stress and anxiety to surviving participants and their families and friends. Making any data from this kind of research available to people outside the research team can lead to retaliation and great damage to individuals and communities (Parkinson and Wood 2015). Also, this kind of case damages research as a whole, with reputational damage to researchers whether or not they were involved with the Belfast Project itself, and decreased willingness of potential participants to take part in future research. The twenty-first century has seen increasing calls for archives to take an active role in pursuing human rights and social justice (Wood et al. 2014). This case study functions as a cautionary tale within the ongoing conversations around the possibilities and challenges of working in this way (ibid).

Some might argue that the ethical problems in this case are linked to the specific research method used, i.e. interviewing, or—more broadly—that these kinds of ethical problems are inherent in qualitative research. However, it is clear from other research that the ‘politics of information’ affects research into political violence that uses a wide range of methods, from ethnographic participant observation to quantitative and digital research (Bridgen and Gohdes 2020). Also, the problems created by the Belfast Project are not specific to oral history research. Regardless of the research method used, doing research with people who have been involved in violence can create short-term and long-term risks for both participants and the ‘non-consenting others’ (Mannay 2016, 123) named by participants. Giving people access to a research dataset, such as that contained in the Belfast Project archive, ‘can bring to attention previously hidden connections, relationships, histories, and contexts that risk having harmful personal or political effects for research participants’ (Bridgen and Gohdes 2020, 257). Furthermore, ‘Entire communities might suffer “collateral damage” from research that makes sensitive information visible to a new audience’ (Brigden and Gohdes 2020, 256).

One of the reasons given by Mr Justice O’Hara, the judge in Ivor Bell’s ‘trial of the facts’, for ruling that the evidence was unreliable, was that the researcher had asked leading questions. Yet the Belfast Project researchers were recruited specifically from each side of the Troubles, to interview people from their own side. This makes sense because to gather information from people who are implicated in violent conflict, a researcher would need to demonstrate understanding of their participants’ perspective on that conflict (Thaler 2021) and to be trusted, at least to some extent, by those participants. In this context, in the course of a qualitative interview which is in fact a conversation and where both parties to the dialogue know they share a perspective, some leading questions seem almost inevitable. Even if there were no leading questions, it is virtually impossible to exclude the influence of a researcher’s own standpoint from data gathering (Thaler 2021), particularly when the researcher’s standpoint is explicitly aligned with participants’ standpoints. In this kind of research, data is not collected from participants by interviewers, it is constructed by participants and interviewers together. And other factors could render such ‘evidence’ as unreliable in legal terms, such as retrospective bias which inevitably affects the views and interpretations of participants and researchers when discussing events of the past (Thaler 2021). Yet for researchers, social commentators, historians and others, this kind of evidence is vital.

Research into violent conflict is suffused with power at macro and micro levels. Violent conflict itself is a display of power, often in response to other displays of power. Then there is the balance of power between researchers and participants, each of whom has the power to tell the truth or to lie, to keep or break promises. Wood et al. (2014, 401) assert that ‘When power is denied, overlooked, or unchallenged, it is misleading at best and dangerous at worst. Power recognized becomes power that can be questioned, made accountable, and opened to transparent dialogue and enriched understanding’. But when legal or state power comes into play, this is not necessarily the case. Though those powers may be contested, it may not be possible—at least in the short term—to question them or to demand accountability.

Archives in general have been described as sites of power contestation, though this is usually more covert than in the case of the Belfast Project. Expressions of political and social power within and through archives are usually held to relate to who makes and uses archival records, and why (O’Toole 2002, 45). In the case of the Belfast Project the power contestation is more overt, with confidentiality and secrecy promised by researchers being breached, leading to legal tussles over access to confidential archival materials, and the release of some of those materials with dramatic consequences for both researchers and participants.

10.5 Lessons Learned

Violent conflict is a pressing social problem which often leaves an unwelcome legacy (Crooke 2010). This means conflict can recur after prolonged periods of calm, as shown by the Brexit-related uprisings in Northern Ireland in the spring of 2021 after 20 years of relative peace in the region. Research into the causes and consequences of violent conflict is essential for understanding how such conflict can be prevented or resolved (Thaler 2021). That research needs to be conducted with extreme care, and no promises should be made that will not be kept. Yet researchers also need to recognise that they may not always be able to safeguard the data they gather, though they should always make every effort to do so.

Truth and facts are not singular and identifiable but multiple and contested. This is the basis on which Boston College set up its archive, and this also plays out in the tussle between research ethics and legal ethics in this case. The criminal justice system is not the only mechanism that can put pressure on researchers to release confidential data. Academic journal publishers, editors, and peer reviewers have also done so (Parkinson and Wood 2015). Unless or until there is legal recognition of researcher-participant privilege, it will not always be possible for research data to be kept secure both ethically and legally.

The Belfast Project worked, initially, because Boston College and the Project staff created conditions in which trust could be established and built, and participants trusted the Project and the College—or, at least, the researcher they spoke with who was explicitly on their side. It is a testament to the Project that this trust was maintained even though they were gathering data from both sides of the conflict, when so much mistrust has built up between loyalists and republicans over many decades. Then, when the Project’s existence was made public, contravening the conditions in which that trust existed, the trust broke down. The consequent reputational damage to Boston College, Belfast Project staff, and research, researchers and research institutions more generally, means the mistrust generated by the Project’s breach of confidentiality is much more widespread than the trust built up by the Project before that breach. Reputations are slow to establish and grow, and quick to damage or destroy.

Kahryn Hughes and Anna Tarrant offer a useful summary of the implications of this case:

At its simplest, the case of the Belfast Project established that, despite assurances to the contrary, the safeguarding and confidentiality of archived data is not necessarily always possible, or it might only be possible for certain sorts of data at given historical moments, regardless of the contractual agreements in place at the time consent is given or sought for archiving. Legal-political changes have the power to destabilise such agreements or contracts; and thus assurances given by organisations such as universities, or individuals such as researchers, cannot be understood as enduring for all time. The [Belfast Project] case is an extreme example that reflects the highly charged character of those particular data. Most interview data would not provoke international political interest and risk of this sort. Nevertheless, it is a useful example to underscore the changing and potentially fragile contexts through which data may pass, and the limits of researcher control and protection of them. (Hughes and Tarrant 2020, 45)

10.6 Implications and Recommendations for Efers

This case shows that, far from being inert resources, archives can ‘engage in powerful public policy debates’ (Schwartz and Cook 2002, 2). This is one reason why it is useful for policymakers to know about and understand archives. Another is that archives can provide a useful resource for policymakers. We know that policymakers use a variety of sources of evidence in their work, such as research evidence, theoretical evidence (ideas, concepts, models), expert advice, political and professional knowledge, and experiential evidence or testimony (Glasby 2011; Nutley et al. 2012; Sohn 2018; Bache 2019). There is also a role here for archival evidence. Someone making policy to help with conflict management might have found the Boston College archive to be a very useful resource.

In one sense, the case of the Belfast Project shows what can happen when policies come into conflict. In this case the Boston College policy on confidentiality for Belfast Project participants came into conflict with the criminal justice system policy of working to bring offenders to justice. Policy conflict is a complex arena with varying levels of intensity and action, affected by the different attributes and cognitive and behavioural characteristics of individual policy actors (Weible and Heikkila 2017). But put simply, when policies come into conflict, the people who operate those policies also come into conflict, each group trying to gain the upper hand. In the case of the Belfast Project, legal ethical policy won out over research ethical policy. However, there is no clear policy hierarchy at national or international level (though such a hierarchy may exist at institutional level), so it is possible that on another occasion, if operated differently, research ethical policy could prevail. And we know that policy conflicts can lead to new policies (Weible and Heikkila 2017). Perhaps this case will ultimately lead to more robust policies around researchers’ rights to confidentiality, such as those which exist for journalists (Adams 2014). Some will argue that the requirements of the criminal justice system should take precedence over the ethics of research. However, researchers need to do everything they can to ensure the welfare of their participants, even when those participants are implicated in violent conflict, and this requirement should be supported by well-made policies.

The workings of violent conflicts are invisible to most people, as are the workings of researchers, and the workings of the state including the making of policy (Bridgen and Gohdes 2020, 263). Yet one factor linking violent conflicts, research, and policymaking is trust. Trust is often seen as an attribute of individuals, but it may be more useful to consider it as a resource which is essential for co-operation in complex societies (Cairney and Wellstead 2019). Of course, individuals base actions on ‘trust calculations’, but these inevitably exist within, and are influenced by, a wider context (Cairney and Wellstead 2019, 5). In essence, trust helps us to reduce uncertainty and get things done. When trust breaks down, we often turn to the law. Clearly the law has a vital role to play in our societies, but it is not perfect, and it can be a very blunt instrument. Policy can help to sharpen its edge.

Policymakers need to consider cases such as the Belfast Project when making policy about the management. storage and sharing of sensitive information. Flexibility within such policy is essential because of the evident level of uncertainty involved in gathering and storing information. It is clear that compelling researchers to share information can cause real harm to individual participants and their communities, and to researchers themselves. Managing uncertainty requires continual negotiation, adaptation, and improvisation, and scope for these should be built into any relevant policy.

These ethical dangers are not unique to cases involving primary data. Even if information about violent conflict is publicly available, using it as secondary data for research may bring it to more people’s attention which could lead to ‘new political incentives for retaliation against participants’ (Bridgen and Gohdes 2020, 257). This can lead to reputational damage, loss of social status, ostracization and even more violence. Yet in order to make policy about violent conflict, policymakers need access to relevant findings from good quality ethical research. When legal or other requirements take precedence over research ethics, research is compromised, and findings which could be invaluable to the next generation of policymakers may not exist.