The Boston College subpoenas were a legal battle over whether interviews gathered for the Belfast Project could be compelled for use in a criminal investigation—and whether an archive’s promise of confidentiality could survive a court order. The case exposed a sharp divide between what oral histories might reveal about the past and what courts could demand, identify and admit as evidence. It also showed that the archive’s protections were less secure than its interviewees had been led to believe.

An archive built on secrecy

The subpoenas arose from a project devised by Ed Moloney, who proposed that BC collect oral histories from people who had taken part in the Troubles. Admitting membership in the Irish Republican Army could expose former militants to prosecution, while speaking about paramilitary activity could bring the stigma and danger attached to being called a tout. Moloney’s solution was to interview participants while they were alive but seal their testimony until after their deaths, when they would no longer face prosecution or retaliation from former comrades. The project’s method was designed to reach people who would be unlikely to speak to conventional academic researchers. Anthony McIntyre, a former IRA member and prisoner known as Mackers, approached republicans who had become estranged from Gerry Adams. He used encrypted email, carried a digital MiniDisc recorder, and met participants for repeated sessions that could total ten hours or more. Recordings and transcripts went to the Burns Library’s Treasure Room. In the archive’s working materials, participants were identified by letters rather than names; only separate contracts linked those codes to individuals. That arrangement depended not only on secure storage but also on confidence in the confidentiality promise. When Ed Moloney prepared the participant contracts in early 2001, he suggested that BC’s lawyers review the wording. Bob O’Neill said he would send it to the university’s lawyers, but the agreements were never reviewed by them. They also lacked a qualification explaining that a confidentiality promise could not necessarily withstand a court order. The omissions became consequential when investigators sought access to the archive.

The subpoenas and the first disclosure

The demands arose in connection with the Police Service of Northern Ireland’s investigation of the abduction and murder of Jean McConville. In May 2011, BC agreed to turn over recordings of Brendan Hughes, who had died and whose testimony had already been published, in edited form, in Voices from the Grave. The material surrendered to the authorities was not similarly redacted: Hughes’s transcripts and recordings were unredacted. Ed Moloney objected that investigators now possessed information on which they could act and feared that the release would lead to demands for the identities of other interviewees. The difference between a published book and a raw archive mattered. Voices from the Grave had been edited to shield identities for legal and security reasons; the surrendered recordings had not. The disclosure therefore did more than confirm that a deceased participant’s account was accessible. It demonstrated how much more an unredacted interview could reveal, and how little the project’s confidentiality arrangements could control what happened once a court compelled the archive to cooperate. BC agreed to contest a later subpoena concerning interviews relevant to McConville’s disappearance. In court, the project’s defenders argued that its purpose was academic and that its commitments to participants should be respected. But the underlying promise had never been made legally secure in the contracts. The case thus challenged the archive on two fronts: the authorities sought evidence for an investigation, while the project’s own practices left doubt about what protections it had actually established.

Identifying the interviews

A judge asked McIntyre to help determine which coded interviews discussed McConville. McIntyre refused, arguing that doing so would cross the boundary between academic research and police investigation. O’Neill also declined, saying he had not read the interviews. Over several days during Christmas, Judge William Young read the republican transcripts himself. He found six participants who mentioned McConville, although one had referred to her only in passing, and authorized the release of recordings associated with five participants. The dispute over Dolours Price’s testimony revealed a consequential distinction within the archive. McIntyre had interviewed Price for roughly fifteen recorded hours, and those conversations did not mention McConville. Ed Moloney and McIntyre had therefore maintained that her Belfast Project interviews contained no material that could justify the subpoena. But Price had also given Moloney a separate, detailed account of McConville’s disappearance. Those recordings were not part of McIntyre’s interview with her. The claim about what was absent from one set of tapes was technically accurate, but it did not describe everything Price had recorded for the archive. The distinction illustrates why an oral-history archive cannot be treated as a single, uniform body of testimony. Who conducted an interview, what the participant discussed, and which recording the court examined all shaped what could be said about its contents. A denial about one interview did not settle the question of what existed elsewhere in the archive.

Appeals, codes and the limits of proof

Moloney and McIntyre appealed the disclosure decision to the United States Court of Appeals for the First Circuit and lost. A stay temporarily barred BC from handing over the interviews while they sought further review, but in spring 2013 the Supreme Court of the United States refused to hear their case. The legal contest had failed to establish the broad protection the project’s interviewees had expected. The archive’s coded identities created a separate evidentiary problem in the prosecution of Ivor Bell. One Belfast Project participant, identified as Z, was said by prosecutors to have acknowledged a role in the McConville murder. Bell’s lawyer argued that the tapes were unreliable and inadmissible, and that Bell was not Z. BC had lost the separate form linking Z’s code name to the interviewee’s real identity, leaving prosecutors without that document to prove the identification. A voice analyst later testified that Z was likely Bell; a police detective who had heard the interview interpreted its account differently from the defence. The recording could be historically suggestive while its identity and meaning remained disputed in court. The contrast sharpened a central difficulty in using oral history as criminal evidence: a participant’s account, its identification, its reliability and its admissibility are separate questions. The subpoenas compelled access to material that the project had gathered for posterity, but they did not make every allegation in it legally provable. The case left the Belfast Project’s promise of silence compromised, while the testimony’s capacity to settle responsibility remained limited.