Pressing Evidence: Activating Khmer Rouge Archives

Abstract

Across the world, non-state actors are documenting international crimes and creating archives for accountability purposes. In this article, we consider how archives and their records are ‘pressed into’ legal service. At a time of wider archive creations, we suggest the archives pertaining to the Khmer Rouge regime (1975-1979) provide insights as a compelling ‘post-accountability’ case of the continuum of archival processes. By examining four Khmer Rouge archives, we demonstrate how records are activated in legal processes across different spacetimes, and how the records themselves ‘(im)press upon’ on the legal process. In these processes, different actors seek to control the narrative of the past through archival holdings. We find that entrepreneurial justice, especially in the crucible of a legal process, can create fierce competition between actors over the economic and social capital inherent in record-keeping that is ultimately detrimental to understanding and pluralising the past.

Introduction

In September 2022, the European Union Agency for Criminal Justice Cooperation (Eurojust) and the Office of the Prosecutor of the International Criminal Court (ICC) published jointly written guidelines for civil society organisations on Documenting international crimes and human rights violations for accountability purposes. As ICC Prosecutor Karim Khan and Eurojust President Ladislav Hamran note in their Preface, civil society actors across the world are critical to the ‘common work’ of fighting impunity and are ‘increasingly active in documenting core international crimes and human rights violations’ (Khan and Hamran, 2022: 4). Often, they arrive before or work in the absence of a ‘competent authority’ willing or able to investigate. While Khan and Hamran emphasise that civil society actors’ documentation for accountability purposes is needed and valuable, they also note the risks involved: of over-documentation, of re-traumatisation, and of compromising the quality of the evidence (6). The Guidelines come at a time when civil society actors across the world are working under increasingly challenging circumstances to record and analyse information for the purpose of assisting accountability processes. The archives created by this work, encompassing records in manifold genres and im/materialities, collected because they are believed to document harm or abuse and may evidence crime in a future public accountability mechanism, have been understood as ‘transitional archives’ (Viebach, 2021) or ‘justice archives’ (Rangelov and Teitel, 2023).

This form of ‘entrepreneurial justice’, where non-state actors step in to record harm in the absence of ‘competent authorities’, blurs the spheres of private and public, as civil society organisations or private individuals conduct work ordinarily conceived of as proper to state responsibility and authority (Burgis-Kasthala, 2020). The archives provoke claims and attendant anxieties about custodianship and privacy. Who should have ongoing control over records that document (past) harms? Who should have access, and for what purposes? As the records are archived on the premise of use in a future accountability process, how should the archives be managed so as to safeguard potential evidentiary value? And what happens after the legal process for which records were originally archived, in their ‘afterlife’ (Viebach, 2021: 405; also see Biber, 2018)? These questions are important as they relate to broader questions about authority, the nature of justice sought (and possibly ‘delivered’), and the inter-relations between archival institutions, their records and legal proceedings.

In this article, we consider the ways in which ‘justice archives’, and the records they contain, become ‘pressed into’ legal service in different accountability processes across spacetimes. At a time when there is broad interest in the creation of new archives, we turn to archives of the Khmer Rouge regime (1975–1979) as a compelling, long-term, ‘post-accountability’ case. These archives were created by a range of state and non-state actors for purposes that included future legal accountability. Legal accountability mechanisms, including a near 16-year tribunal process, have now run their course in Cambodia. As such, Khmer Rouge regime archives provide insight into the continuum of archival processes and politics.

The Khmer Rouge regime, formally established as Democratic Kampuchea, under the Communist Party of Kampuchea (CPK), and colloquially known as ‘Angkar’ (the organisation), controlled Cambodia including its capital, Phnom Penh, between 17 April 1975 and 6 January 1979 and was characterised by famine, forced labour and persecution. Around one fourth of the population perished (Tabeau and Kheam, 2009). In early January 1979, a coalition of Vietnamese and Cambodian dissident forces ousted the Khmer Rouge leadership from Phnom Penh. Work to document Khmer Rouge crimes began immediately and continues to this day. Over the years, several archives have been established, and there have been two legal processes. In 1979, a People's Revolutionary Tribunal (PRT) held leaders Pol Pot and Ieng Sary in absentia responsible for genocide, and between 2006 and 2022, the Extraordinary Chambers in the Courts of Cambodia (ECCC), established through a 2003 agreement between the Cambodian government and the United Nations (UN), investigated the responsibility of 10 individuals, sent five persons to trial and convicted three persons for the most serious of crimes. During this period from 1979 until today, a range of actors – public and private, Cambodian and non-Cambodian – has created and maintained archives. Throughout this period, and more intensely in recent times, the role and status of the archives have been critically discussed.

In the following, we analyse four Khmer Rouge archives: the S-21/Tuol Sleng Genocide Museum (TSGM) Archives, the Documentation Center of Cambodia (DC-Cam), the ECCC and the Legal Documentation Centre related to the ECCC (LDC). Apart from the LDC, which is much younger, these are long-term archives relating to the Khmer Rouge regime, and each has attracted significant scholarly and activist attention over the years. We draw from extensive joint and individual research on legal and cultural responses to the Khmer Rouge regime (e.g., Elander 2018; 2022; 2023; Hughes and Elander 2022; Hughes 2003; 2015; 2020). We have followed the discussions on archives in and about Cambodia since 2003 and our disciplinary differences as a socio-legal scholar and a geographer have generated productive conversations about how records operate in relation to broader political and legal processes.

In critically examining the creation and maintenance of these archives, we demonstrate how the activation of records is affected by and relates to the legal form, how the records themselves ‘(im)press upon’ on the legal process and how previous activations ‘press on’ new ones. We find that entrepreneurial justice, especially in the crucible of a legal process, can create fierce competition between actors over the economic and social capital inherent in record-keeping that is ultimately detrimental to understanding and pluralising the past. We begin this article by describing our focus on and approach to archives and their relation to criminal law, before turning to the four archival institutions and the question of how records are activated by and yet ‘press upon’ law and legal process.

Archives and (International) Criminal law

‘Transitional archives’ or ‘justice archives’ are characterised less by the past than by a future orientation (see also Derrida, 1995). But the intensity, passion and desire to record harm through archiving is ‘a work of mourning’ in that it ‘produces memory, but produces forgetting at the same time’ (Derrida, 2002: 54). Despite the public good usually claimed, the work of recording harm and creating the archive both preserves evidence and endangers it. Archiving is always, as Derrida establishes, archiviolithic (Derrida, 1995: 10): it harbours records safely inside a realm of ready activation for a future process, while at the same time increasing their vulnerability as co-located, compartmentalised and contained objects. In our case, records of harm are, in being gathered in an archival space or institution, safeguarded for future reference and simultaneously made vulnerable by that spatial or political location, which may meet with public exposure, private enclosure or various other forms of neglect, secrecy, capitalisation or forgetting.

The dominant frame through which justice archives are produced – and their doubled remembering and forgetting of harm enabled – is that of (international) criminal law. While human rights organisations have long documented violations for purposes of condemnation, the turn to international criminal law (ICL) since the 1990s means that the legal framing is increasingly that of crime and criminal law, rather than violations and human rights law. This shift has affected the lens through which activists see and understand harm in a way that ‘reinforces an individualized and decontextualised understanding’ (Engle, 2015: 1071). The information needed to demonstrate individual criminal responsibility is not necessarily the same as that needed to show human rights violations by a state. Like all frames, it includes and excludes. If the work of recording harm is premised on it later becoming a criminal evidence investigation, what information is considered important and what is not? The question becomes: how does the frame of ICL affect archives and record-keeping activities, these stores being (after Derrida) always already ontologically vulnerable?

In this paper, we draw on the work of archival theorists who understand records and record keeping as part of a continuum, rather than as having distinct lifecycles, and as ‘activated’ across time and space rather than as atemporal or aspatial. The Records Continuum Model (RCM) was developed by archivists to conceive of records and recordkeeping activities as operating across multiple times and spaces (Upward, 1996). ¹ It identifies four dimensions of recordkeeping: creation, capturing, organisation and pluralisation. A significant difference of the RCM for our purposes here is the prominence given to pluralisation of a record whereby it is: ‘disembedd[ed] …from its original multiple organisational and/or personal contexts and carr[ied]… through spacetime’ (McKemmish, 2017: 139). Archivists working with this model also draw on Derrida to emphasise how ‘records are “always in a process of becoming”’, (McKemmish, 2001: 334) defined not so much by their ‘content or informational value’ as their ‘evidentiary, transactional and contextual nature’ (ibid: 335). As Upward put it, ‘disposition is cyclical and never final… Records can even have multiple lives in spacetime as the contexts that surround their use and control alter and open up new threads of action, involving re-shaping and renewing the cycles of creation and disposition’ (Upward in McKemmish, 2001: 336). It is in this sense that we examine the ‘afterlife’ of the archive. By recognising the open-ended nature of archives, how they are constantly ‘(re)assembled, (re)purposed and (re)contextualised’ (Viebach, 2021: 404), the afterlife is not about record disposition, but about pluralisation and (re)creation, even as this provokes social and political resistance and opposition.

Taking this approach also allows us to trace the way records are ‘activated’ at different times and across different archives; the same record may be called upon at the same time, but by different people and for different purposes. Each recontextualisation affects the record. As archival theorist Eric Ketelaar (2001: 137) explains:

Every interaction, intervention, interrogation, and interpretation by creator, user, and archivist is an activation of the record. The archive is an infinite activation of the record. Each activation leaves fingerprints which are attributes to the archive's infinite meaning.

This is important because ‘every activation of the archive not only adds a branch to what [Ketelaar] propose[s] to call the semantic genealogy of the record and the archive [but] every activation also changes the significance of earlier activations’ (Ketelaar, 2001: 138). Contexts affect a record, changing its value and significance. In our case, the contexts of ICL and its attention to evidence affects the record's value and significance. By tracing the activation of Khmer Rouge records across time and space, we demonstrate how the legal form of evidence – as imagined by civil society actors and as practiced by courts – presses on records and on archives. We suggest that the way the legal form of evidence affects records can be understood as a kind of pressure or pressing (leaving an imprint) that in turn affects the value or meaning of the record and does so in relation to both earlier and later activations.

Our use of the metaphors of pressure and imprint builds on Ketelaar's notion of ‘fingerprint’ and takes inspiration from two additional sources. In her work on the cultural politics of emotions, Sara Ahmed (2004/2014) points to how impressions work by pressing something on us – this can be an ‘impression of others’ or something or someone ‘leav[ing] me with an impression’ (6). Importantly, the ‘value’ of something does not (necessarily) reside independently in an object but instead in its capacity to affect, in its giving more or less of an impression. For Ahmed, emotions are, similarly, ‘produced as an effect of [their] circulation’ (45). In this vein, we suggest that records take on particular ‘values’ that are dependent on their activations and the impressions they are able to leave. The metaphor also has a legal dimension. When judges assess evidence, they are tasked to weigh the evidence in relation to claims. The evidence gives a claim weight, and in weighing different pieces of evidence against one another, judges make a decision. Evidence is weighed, has weight and does so differently depending on the legal issue. The concept of pressure denotes how a record can be affected by external contexts when being activated and weighed in court.

Scattered Records: S-21/Tuol Sleng

The work to document harm inflicted by the Khmer Rouge leadership began immediately after they were expelled from Phnom Penh. The archive of their most deadly security centre, known as S-21, is well known, in part through its deposition within what is now the TSGM. During his leadership of the security centre, chairman Kaing Guek Eav, alias Duch, instituted detailed administrative procedures and insisted on pain of death that his staff minutely record prisoners’ passage through the institution. A distinct documentation unit, including a photography subunit, ‘was responsible for transcribing tape-recorded confessions, typing handwritten ones, preparing summaries of confessions and maintaining the prison's voluminous files’ ² (Chandler 1999: 27). A paper-based file was created for each new adult prisoner arriving at the centre to which their ‘mug shot’ photographs (a frontal and a side-view) were attached (Song, 2021: 24). The main aim of the interrogations under torture and of the archivists attached to these practices was to document traitorous behaviour and identify prisoners’ ‘strings’ or ‘networks’ of association, so that these otherwise unknown traitors could be arrested. This deadly aim suffuses the archive they created over the years of 1976–1979 (see Caswell, 2014). When the Khmer Rouge fled Phnom Penh in January 1979, there was little time to destroy the records, and many papers, photographs and photographic negatives were simply left behind. The S-21 site was found only days later by a Vietnamese army unit, including a photographer and film cameraman, who alerted the new authorities. In film footage from soon after this discovery, documents and photographic records are shown scattered across the site. ³

The records found at S-21 were (re)assembled and (re)organised by newly appointed museum staff (see TSGM, 2019). As Song (2021) chronicles, one challenge of archiving the documents left behind at S-21 (an estimated 700,000 pages) was their disorganised and incomplete state. In 1979, with limited resources and personnel, staff needed to try to preserve the records and prevent them from being souvenired, as well as to create some order out of them. Their work had two immediate aims. The first of these aims was to assist requests by Cambodians who came to the site in search of information about their missing loved ones (Ledgerwood 1997). As Song (2021) details, records were organised to facilitate such requests. After initial steps to categorise the documents by type, the documents were arranged in alphabetical order. This was done ‘to make it easier to search for the prisoner names’ (Ung Pech, quoted by Song, 2021: 78).

The second aim was to identify records that could evidence crimes committed by the former regime. In March 1979, a special Crime Investigation Section was tasked with identifying incriminating documents. This had an important political purpose of legitimising the new government of Cambodia, the People's Republic of Kampuchea, and records were shown, for example, to international journalists, lawyers’ delegations and UNICEF and ICRC relief officials (Hersh, 1979; Hughes, 2020). In August 1979, the PRT was held. At the PRT, S-21 was the focus of some 11 of 200 documents presented (UNESCO Cambodia nomination, 2023; Jarvis, 2024). Moreover, PRT judicial officers visited the site to view materials there including structures and objects of detention and torture, the archives and mass graves. Documentary evidence presented at the PRT included a list of hundreds of individuals killed at S-21 as ‘Important Culprits (arrested from 1976 to April 9, 1978)’, and an excerpt from a Khmer Rouge-era ‘Conference of ‘S-21’ on Conduct of Interrogations, July 2023, 1977’ that openly refers to ‘methods of torture’ (see de Nike et al., 2000: 410–411). Moreover, PRT witnesses such as S-21 survivor Ung Pech and expert Professor Vandy Kaonn, who wrote a report on crimes against the population of Phnom Penh, referred to S-21 documents as supporting evidence (de Nike et al., 2000: 75–82; 287–304). Although the PRT lawyers included both Cambodian and non-Cambodian, and international experts testified to the applicability of the charge of genocide (see Quigley, 2000), the international community largely dismissed the PRT as a show trial. And while the PRT judgement and transcripts were made available in Khmer, French and English, they mostly languished after the trial until they were rediscovered in the early 1990s (see Jarvis, 2000). In this activation at the PRT, the records were caught in the tension between evidence and ‘show trial’: not only pressing the broader international community to recognise the crimes but also being imprinted with the motivated dismissals of the tribunal being a spectacle and ‘for show’.

Given the challenges of this early period, fundamental archival principles were not met when attending to the records. Perhaps most significantly, the archival principle of attempting to keep records within their original order was abandoned. Archivist Frings-Hessami (2019) argues that this amounts to a ‘neo-colonial exploitation’ and should be understood as an example of a new and fifth dimension of the RCM: ‘appropriation’. However, the reasons for not upholding the original order are likely to have been much more mundane, including the original S-21 arrangement of documents being unknown (Song, 2021). One witness attested:

It's being organized as best the archivists can do, at Tuol Sleng. When I was there in 1981 you had piles on the floor of file folders full of photographs, full of negatives with these confessions, and slowly they are trying to organize, make archives of it (Hawk interview, 1983).

As Song (2021) details on the basis of official reports from the period, staff tasked with organising the records were working under difficult conditions, emerging from a genocidal regime and in a society where almost everything was missing, including pens and paper, and museum staff were being paid in rice, wheat grain and corn (TSGM, 2019: 74). Those assigned the task were not trained archivists. With that in mind, the fact that the archives were attended to at all is significant.

While the archive was re-assembled, the site of S-21 was memorialised as Tuol Sleng Museum of Genocide Crimes (Song, 2021; Benzaquen, 2016; Ledgerwood, 1997). Although the archives themselves were not generally made accessible to visitors, images of individual documents were included in the permanent ground-floor exhibition of the Museum. Gradually, non-state archival assistance was sought, or offered and accepted. These developments saw the copying and sometimes the transferral of original records into overseas archives, bringing both benefits and losses. Some losses of primary records reportedly occurred during the early period of the museum (Maguire 2005), including at the time of the PRT (TSGM, 2019: 105), and later there was a de-contextualisation of the S-21 prisoner portrait photographs internationally (e.g., Hughes, 2003; Benzaquen, 2016). But it is not always clear when, which or how records left TSGM, given that some records were likely destroyed by Duch and his staff, the remainder were disorganised when they were found, others were taken for study or ‘protection’ by visitors ⁴ or staff, and yet others were shared by museum staff with foreign visitors for purposes of preservation and restoration. Early on, for example, S-21 photographs and negatives were sent to East Germany and Vietnam for chemical treatment and development, providing the prints that have been hanging at the Museum since 1980–1981 (TSGM, 2019: 120; Chandler, 1999: 27).

While Cambodians largely called upon TSGM for information about individuals, non-Cambodians began visiting the archives for wider research purposes in the early 1980s (Ledgerwood, 1997; Song, 2021). This group includes historian Ben Kiernan, human rights activist David Hawk, political scientist Craig Etcheson and legal scholar Greg Stanton. All four worked to document what had occurred in Cambodia, and eventually set up documentary projects: Stanton the Cambodian Genocide Project (now within Genocide Watch); Hawk the Cambodian Documentation Commission (CDC) under the auspices of the Center for the Study of Human Rights at Columbia University; and Kiernan, assisted by Etcheson, as director of the Cambodian Genocide Program (CGP) at Yale University. Based on Hawk's document store and further field research, and publicly spearheaded by Cambodian survivors in the diaspora, the CDC sought (unsuccessfully) to have a state bring an application to the International Court of Justice for genocide proceedings against Democratic Kampuchea. In addition to this use in an (abortive) legal case, the S-21 archive continued throughout the 1980s to attract scholarly attention and formed the foundation for much scholarly work on the regime (see Becker, 1986; Kiernan, 1985; Chandler, 1999). As Song (2021) argues, it was this early interest of scholars and activists and their work with primary sources at Tuol Sleng that led a larger number of scholars to studies of the Khmer Rouge leadership and life under its control. ⁵

We have here drawn attention to the ways in which the S-21 records, following their distinct and disturbing creation, have been continually implicated in practices of (re)capture, (re)organisation and pluralisation – in the first instance by state actors in legal service to the PRT, and later by non-state actors in relation to a proposed legal case. Desires around exhibition and preservation generated moments of non-state and state-sponsored pluralisation and de-contextualisation (of records and knowledge about these records). Private actors who participated in the copying of S-21 records in the 1980s did then and have since (re)organised their collections into new archives in Cambodia and in the North American universities that funded their original or subsequent research. Private actors claiming copyright and control of the photographs and reuse of records from the former Khmer Rouge regime was not limited to non-Cambodian groups. Come the mid-1990s, Cambodia's largest and most well-known private archive, the Documentation Centre of Cambodia (DC-Cam), was being instituted under the auspices of Yale University's CGP.

Re-Assembling Archives: DC-Cam

Long-time Cambodia scholar Ben Kiernan, with support from Craig Etcheson and Helen Jarvis, founded Documentation Center of Cambodia (DC-Cam) as a field office of the CGP in the wake of the US Congress’ passing of the Cambodian Genocide Justice Act (1994) and the US State Department funding Yale University to conduct the work. The CGP sought to ‘conduct research, training and documentation on the Khmer Rouge regime […] assemble evidence concerning the leadership [and] determine whether the […] regime violated ICLs’ (DC-Cam website: Our History). Youk Chhang, a Cambodian American survivor who had returned to Cambodia during the United Nations Transitional Authority period, was appointed as DC-Cam director.

The DC-Cam office and archive was originally located in a small modern building behind a large black security gate in the shadow of Phnom Penh's Independence Monument. David Hawk donated part of the collection he had amassed during the 1980s to DC-Cam, and other private archives and public collections held or discovered by Cambodian ministries were given to DC-Cam for indexing and safekeeping. By the late 1990s, DC-Cam was not only holding important stores of documentation produced by and about the KR regime but also producing new knowledge about the historical crimes of the Khmer Rouge by way of its dry season field trips, mapping expeditions and mapping reports. The Center was locating, researching (including semi-structured field interviews) and plotting (using new GPS technologies) scores of prison, burial and memorial sites across the country (see Jarvis, 2002). DC-Cam had good working relationships with several government ministries. The Center attracted visiting international researchers and large numbers of Cambodian students. Most of these young people worked in a voluntary capacity, but some stayed on to become paid staff members and were further supported by DC-Cam to study for tertiary degrees abroad. DC-Cam eventually branched into publishing, including research monographs written by its staff, and translated thousands of DK-era documents (Etcheson, 2005: 69).

While geopolitical interest in an internationally supported tribunal of the former Khmer Rouge waxed and waned over the late 1990s and early 2000s (Etcheson, 2019; Fawthrop and Jarvis 2005), DC-Cam's records seemed only to grow. Renovation works re-started endlessly at the DC-Cam office to add additional floors for ever more file cabinets, boxes, computers, photocopiers and desks. Caswell (2014: 84) describes the entrepreneurial dynamism of the organisation at the time:

DC-Cam staff possessed the political savvy and professional legitimacy to attract American and Western European funding. Certainly, the connection to Yale University […] added to the project's legitimacy, while Chhang's personal story of torture under the regime, his unwavering dedication to holding the perpetrators accountable, and his talent for navigating both the Cambodian political system and the world of international funders lent an unprecedented drive to the organization.

In 1997, DC-Cam re-launched as an ‘independent Cambodian research institute’ and it has since operated as a not-for-profit organisation funded by various government bodies including USAID.

DC-Cam's archive reportedly contains over a million documents (DC-Cam website: Documentation) and can be broadly divided into documents created during the Khmer Rouge regime and those created subsequently. Documents created during the regime include Minutes from the CPK Central and Standing Committee meetings, telegrams, daily reports, the magazines Revolutionary Flag and Revolutionary Youth, photographs and other records. Documents created after the regime include records created by DC-Cam, such as maps of mass graves and many interviews with survivors, alongside other later records such as documentary films, scholarly publications and reports.

During the years of activism and negotiation that preceded the establishment of the ECCC, DC-Cam's archive was presented as key to any accountability measure. The 1999 Report by the UN Group of Experts, established to assess the possibilities of legal accountability for Khmer Rouge atrocities, singled out DC-Cam as the ‘most impressive and organised’ of efforts to accumulate evidence on Khmer Rouge atrocities (UN Group of Experts, 1999: 46). While the report was careful to point out that DC-Cam has not been ‘oriented towards investigation in preparation for prosecution of particular individuals’ (UN, 1999: 49), it nevertheless concluded that the store held by DC-Cam ‘[made] clear the commission of serious crimes’ (UN Group of Experts, 1999: 58). And while the UN Report was careful to note that DC-Cam's archive evidenced crimes but not individual responsibility, Steve Heder, in his influential Seven Candidates for Prosecution (2001), analysed DC-Cam's records to identify seven individuals against whom there was a prima facie case of criminal responsibility. Finally, John Ciorciari, long time DC-Cam legal advisor, wrote with Youk Chhang a chapter published in 2005 on how documentary records, many of which were held by DC-Cam, could be used against the former CPK leaders as evidence of criminal responsibility. For an NGO that had set out to ‘compile and organize information that can serve as potential evidence in a legal accounting’ (DC-Cam: Our History), these publications were demonstrative of DC-Cam's success.

Evidencing Crime: DC-Cam at the ECCC

By the time that an Agreement to establish the ECCC was signed in 2003, DC-Cam enjoyed significant influence. As the ECCC was swearing in officials and opening investigations in 2006–2007, DC-Cam director Youk Chhang was featured in Time magazine as one of ‘60 Asian heroes’ and one of Time100's most influential persons. DC-Cam's engagements with the ECCC were manifold. Together with David Scheffer, a Professor at Northwestern University School of Law in the USA, DC-Cam founded the Cambodia Tribunal Monitor Consortium to ‘provid[e] public access to the [ECCC] and open, online discussion throughout the judicial process’ through trial monitoring, curated expert commentary and online access to some ECCC records and filings (Cambodia Tribunal Website). In addition, DC-Cam established a program of outreach activities, including a Study Tour program similar to the one later run by the ECCC Public Affairs Section. The Center continued to host academics and practitioners with interests in both its own work and that of the ECCC. Finally, DC-Cam provided the ECCC with records from its archive. According to DC-Cam (DC-Cam Website: Documentation), this amounted to over 500,000 pages of documents and included both Khmer Rouge regime records and DC-Cam created ones. These documents were used to evidence crimes committed by the accused persons.

It is worth pausing to consider how something becomes evidence at an internationalised criminal court. A record that has been created, captured and organised in one archive, is pluralised and re-organised into a legal archive, but here as potential evidence. This is not unusual. Aside from those records created during (judicial) investigations, records are rarely created for purposes of legal evidence but instead have previous lives. In being presented as evidence, a record is, as Clarke and Kendall (2019) point out, ‘brought into a relationship with legal categories and ways of knowing’ (2019, p. 93). This is not an automatic process. As evidence, records are far from self-evident or inherently truthful and instead ‘evidence must be spoken for’ (2019, p. 94). The activation of the records as evidence thus involves representation and mediation through the legal process. And as we will show, the activation of records held or created by DC Cam as evidence at the ECCC was not as smooth as predicted. This activation is illustrative of the potential difficulties of shifting authority across disciplines (from, say, history and archive studies to law). To better understand how this occurs, some background on ECCC legal procedure is necessary.

Among international(ised) criminal courts and tribunals, the ECCC stands out for its civil law procedure. The Internal Rules were drafted to be in accordance with Cambodian laws, which are the heritage of its French colonial past (see Low, 2023). Accordingly, investigations at the ECCC were conducted by impartial Co-Investigating Judges (CIJ) who, after receiving Introductory Submissions by the Co-Prosecutors, examined both inculpatory and exculpatory evidence. At the end of their investigations, the CIJ issued a Closing Order to either send a case to trial or dismiss it. Between 2006 and 2022, the CIJ opened four cases into the responsibility of ten people with two cases proceeding to trials: Case 001 against Duch, and Case 002 against Nuon Chea, Khieu Samphan, Ieng Sary and Ieng Thirith, although the latter two individuals died before the conclusion of trial. Cases 003 and 004 remained highly controversial and, despite extensive investigations, never proceeded to trial. Case 002 covered the entirety of the temporal and spatial jurisdiction of the Court – crimes committed within the borders of Democratic Kampuchea during its reign from 17 April 1975 to 6 January 1979 – and focused on the four most senior surviving leaders. Its Closing Order is a sizable document of some 740 pages with more than 5400 footnotes, referencing Khmer Rouge-era records, interviews, maps and other materials. Due to its size and complexity, the case was severed into two sub-trials: 002/01 and 002/02, and in each sub-trial, further records were presented. The number of documents considered in total in each case was thus substantial. For example, in Case 002/02, this amounted to some 10,800 documents (ECCC Website: Case 002).

The use of records from DC-Cam's archives as evidence was already questioned in the first trial at the ECCC, against S-21 Chairman, Duch, ⁶ but it was in the trial of Case 002/01 that these records came under sustained scrutiny. Early in this trial, the Trial Chamber Judges announced that documents and other sources referred to within Closing Order 002 would be considered ‘put’ before the Chamber and ‘entitled to a presumption of relevance and reliability (including authenticity)’ (ECCC TC 9 Apr 2012: 2). In other words, documents and other sources relied upon in the Closing Order would not be presented for questioning but instead presumed to have met the relatively low bar of prima facie relevance and authenticity. In common law jurisdictions, the laws of evidence are extensive and detailed, with considerable attention given to whether a particular piece of evidence is admissible. By contrast, civil law jurisdictions tend to take a broader approach in terms of admissibility with the key question instead being about ‘weight’, as in how much weight the judges accord a particular piece of evidence. At the ECCC, all evidence was admissible unless otherwise stated, which is associated with the onus to prove the guilt of an accused being on the Co-Prosecutors (ECCC Internal Rule 87.1). Decisions were based on evidence put before the Chamber and subjected to examination. For something to be ‘put’ before the Chamber, it must normally be ‘summarised, read out, or appropriately identified in court’ (ECCC Internal Rule 87.2–87.3). At the ECCC, this practice was meticulous and time consuming as the bench and parties sought to ensure, across three languages, that everyone had noted the correct file number and accessed the (same) file. While far from spectacular, the practice is highly performative. By reading out a file number and the document title, something that was for the Chamber hereto non-existing (even if perfectly visible to everyone present) was put before the Chamber and was thereafter considered part of the proceeding. This practice is, in the words of Jeffrey (2021: 904), a ‘conversion of matter and speech to evidence’, but one, as pointed out above, highly mediated. In deciding that all documents referred to in the Closing Order would be considered put before the Chamber, many unexamined documents were thus rapidly and as a group ‘converted’ to evidence.

The Defence counsel objected. Not only was this a very large number of documents but of specific concern were the records sourced from DC-Cam. Defence counsel for Nuon Chea argued that DC-Cam had been ‘working with an actual goal to have [Nuon Chea] and other … Democratic Kampuchea leaders prosecuted’ and was thereby ‘biased towards the Prosecution’ (ECCC Trial Transcript (henceforth, TT), 16 Jan 2012: 87–88). Similarly, Defence counsel for Ieng Sary argued the archival organisation was ‘not objective’ but a ‘party of interest … with inherent prejudice towards the accused’ (56). The Defence counsel were in this way questioning whether, in sourcing records from DC-Cam's archive, the CIJ had also incorporated DC-Cam's broader system of knowledge including their implicit assumptions about what was important to record, and what was not. The Chamber called a witness from DC-Cam to answer these questions of authenticity, reliability and archival method. DC-Cam first sent deputy director Vanthan Dara Peou, but when he was unable to answer all questions, the Chamber directly called DC-Cam director Youk Chhang.

The questions posed to Chhang (ECCC TT, 2 Feb 2012) illustrate how each activation of records affects all future activations, and how archival actors approach records and archiving practices differently. Often, the Defence counsel sought to devalue the records by putting pressure on the archivist, that is, on Youk Chhang as DC-Cam representative. Focusing on DC-Cam's use of terms that also have specific meanings in legal procedure, the counsel sought to discredit DC-Cam's archival methods. For example, why had Chhang described a document as ‘confidential’ if the ECCC had categorised it as ‘public’ (91–92)? If DC-Cam was ‘analysing’ records and organising them into folders named after accused persons (e.g., ‘Nuon Chea’), had they been assessed as to whether they were inculpatory or exculpatory (73–83)? And how and why were DC-Cam using the term ‘evidence’ in their strategic documents and mission statements? In his responses to these questions, Chhang mostly referred to what might be described as disciplinary difference. As he suggested, ‘law does not monopolise the word … “evidence”’ (85). ‘Evidence’ is used across academic disciplines to denote facts or information that demonstrate truth or validity. In describing their history online (DC-Cam Website: Histories), DC-Cam emphasises that it is not a judicial body with legal authority. But can a private, civil society actor that steps in to fill a perceived ‘gap’ and collects information on atrocities, overstep its authority? When and where does its information gathering encroach on activities that should be exclusive to ‘competent authorities’? And how might the legal evaluation of these records be affected?

A central concern for the Defence counsel was how DC-Cam's many statements on accountability and on the responsibility of particular individuals impacted their archival practices. For example, in their ‘Accountability Project’ (2000–2007), DC-Cam conducted over 10,000 interviews with Khmer Rouge survivors – cadres, victims and their families – to ‘shed light on the roles and activities of the lower-level personnel’ as ‘a way to illuminate chains of command, reporting practices, and other institutional features of the DK regime that can help to hold leaders accountable’ (DC-Cam Website: Promoting Accountability). For the Defence counsel, work such as this suggested DC-Cam was prejudiced towards the responsibility of the accused persons. Defence counsel implied that this prejudice affected their archival practices by directing who they interviewed and how, and in determining what was relevant to record and to archive. This points to the distinction Ketelaar (2001: 132–133) makes between archiving and archivalisation. Whereas archiving ‘is mostly understood to be the activity that follows upon the creating of a document’, archivalisation involves ‘the conscious or unconscious choice (determined by social and cultural factors) to consider something worth archiving. Archivalisation precedes archiving’. The larger question raised by Defence objections at the ECCC is whether a private, non-state organisation that, on the one hand, actively promotes legal accountability can, on the other hand, remain impartial as to individual accountability. In this instance, it was the breadth of DC-Cam's activities (beyond record-keeping) and the possibility that this breadth and framing had affected its archivalisation – choices made as to the content and the organisation of DC-Cam's archives – that promoted Defence claims of bias and prejudice and risked the evidentiary weight of records held by DC-Cam.

Alongside these questions relating to DC-Cam's broader activities and how those activities potentially affected archival practice, Defence counsel also asked about DC-Cam's adherence to archival principles. One of these is the principle of archival integrity, which holds that a body of records created or originally catalogued together, should remain together. Integrity, as noted above regarding the S-21/TSGM Archives, is a challenge when re-creating or re-assembling archives that have been scattered due to conflict, intervention or lack of maintenance. Locating records and re-assembling archives has been one of DC-Cam's core missions, and they have over many years collated documents that had been re-located across Cambodia and beyond. By taking on their custodianship, DC-Cam has preserved and ensured the safety of records that may otherwise have been at risk of neglect or even destruction.

But DC-Cam is not the only archive in Cambodia and, at times, Cambodia scholars have argued about where specific Khmer Rouge records belong. For example, there are photographs of S-21 staff that Hawk gathered in the 1980s that were part of his donation to DC-Cam, despite their originating from the S-21 archive at TSGM (ECCC TT, 1 Feb 2012: 64). In 2012, DC-Cam was gifted other prisoner portraits – seemingly original S-21 records – by an anonymous donor (Jenkins, 2012), which they have not, to our knowledge, returned. When Chhang was asked explicitly whether DC-Cam holds original records ‘originally located at the Tuol Sleng Genocide Museum’, he prevaricated before replying that they had not taken originals from the Museum, but that ‘there are documents which [they had] obtain[ed] from others… [that] are kept [at DC-Cam]’ (66). He appeared to be indicating that, while DC-Cam does not take originals from TSGM, if they locate a record elsewhere that originated from S-21/TSGM, it is not their practice to return that record. While this is not necessarily relevant to the determination of authenticity at the ECCC, it does raise broader questions about adherence to archival principles (such as integrity), and the pragmatic and political relations between different archival actors. In this circulation, records appear as precarious valuables not to be shared.

All records that DC-Cam provided the ECCC were copies. No original was ever transferred from DC-Cam to the ECCC. Early in his trial, accused person Nuon Chea rejected the authenticity of some documents – in particular magazines – on the basis that they were black and white copies, and he therefore couldn’t assess their authenticity. His counsel submitted that the Chamber should adopt a ‘best evidence rule’ according to which whenever original documents were available, these should be inspected rather than copies, and that copies were ‘unsuitable to prove the facts [the evidence] purports to prove’ (ECCC TT 16 Jan 2012: 32–45 and Internal Rule 87.3c). While this was not adopted by the Chamber, some concerns were alleviated by providing better quality colour copies. In the discussion of ‘originals’, however, it emerged that DC-Cam considered an ‘original’ to be the original in Cambodia, as the ‘primary copy’ of an original document held elsewhere. Thus, for example, if the original record is held in an archive abroad, the DC-Cam held copy is catalogued by the Center as the ‘original copy’ (ECCC TT, 1 Feb 2012: 41). This is an unusual interpretation of originals and copies and potentially significant as it risks obfuscating provenance – knowledge about the ‘origin or source of something’ (Caswell 2014: 18). Given the way records are circulated and activated across time and space, their authentication requires clarity on provenance.

While the Trial Chamber eventually found DC-Cam's methods to be reliable and dismissed all claims to the contrary (ECCC TC, 9 Apr 2012: 2, 24–28), we find that the Defence questions and Chhang's testimony shed light on the limits of entrepreneurial archival work and the care required when non-government entities archive for the purpose of later legal accountability. The interactions between DC-Cam and the ECCC regarding records illustrate differences in their respective concerns as archival entities, and how record pluralisation is not necessarily as straightforward as copying an original record or first copy. When activated in a legal process, records are mediated by – pressed upon – contextual concerns that may be different for different courts. Defence questions on DC-Cam's broader activities should also prompt reflection for ‘transitional archivists’ or ‘justice archivists’ on the appropriate breadth and framing of activities undertaken by an archival organisation; too great a breadth or advocacy may serve to undermine the credibility of a record taken from an archive into a criminal trial.

Legacy Archives: the ECCC Resource Center and the LDC-ECCC

The ECCC was never meant to be a permanent institution. It was an extraordinary institution, established to prosecute a limited number of people for specific crimes. Although there was disagreement between the national and international sides on exactly how limited the number of prosecutions should be, the temporary nature of the ECCC was never in question. Despite this, the nature and location of ECCC documents once the Court had ceased its operation was less clear, and some questions remain. In this final section, we examine the archival practices relating to records amassed for accountability purposes and then activated by the ECCC legal process as yet another ‘afterlife’ of Khmer Rouge regime records. Two archival institutions are in focus: the LDC relating to the ECCC and the ECCC Resource Centre. Here, judicial documents become publicly accessible as a ‘legacy’ of that Court and pluralisation is not so much a move from one archive to another as a change in the status of the records created and held by the ECCC. As in previous activations, tensions between different archival institutions are pressing on records and practices and affecting their value and meaning. In contrast to previous activations, these two archives have, publicly at least, the same raison d’etre and share a physical location.

As the ‘headwaters’ of both these legacy archives, the ECCC archive is made up of documents generated by the Court (comp. Redwood 2021) including judgements, decisions, interviews, transcripts, submissions and internal memos as well as copies of records held elsewhere, such as those discussed above and sourced from TSGM or DC-Cam. The ECCC Records Archives Unit is responsible for the management of all the Court's judicial records and exhibits, including their processing, cataloguing, distribution, security and access. Records are classified as public, confidential or strictly confidential (ECCC Practice Direction 2014: 2.d). Internally, the ECCC uses the ZyLab records management system, which provides authorised persons with access to case files, or sections thereof. The public has not had access to ZyLab but has instead relied upon the ECCC website to access public documents. Over the years, most public documents have eventually been uploaded on the website, but the poorly functioning search engine has made access difficult at times.

The fate of ECCC records beyond their judicial function was part of the early discussions of an internal Legacy Working Group initiated by ECCC international Co-Prosecutor Robert Petit in 2008. A Working Paper of the group shows their thinking across seven different areas of the court, including records, archives and the materials in the ECCC library. In 2008, the group invited prominent archive specialist Trudy Huskamp Peterson to advise on documentary legacy and the ECCC (ECCC Legacy Group 2008). In March 2010, the Legacy Working Group was transformed into two formal organs of the ECCC – the Legacy Advisory Group and the Legacy Secretariat (ECCC Administrative Circular, 2010). At a 2012 conference on ECCC legacies, jointly organised by the ECCC and non-government organisations, several participants raised the question of long-term management and archival access, focusing on digital possibilities (ECCC and CHRAC 2012). In the wake of the conference, a newly energised national side of the Court submitted a ‘Legacy Team’ budget line which included funding for legacy and ‘Virtual Tribunal’ work (ECCC Budget, 2012). This initiative was shut down when either the UN or donors objected to the proposed inclusion; publicly, an ECCC spokesperson reported that donor countries had requested the Court to ‘postpone the Legacy component’ (Neth in Wallace, 2014). While budget stress may also have played a role (Dittrich, 2016: 153), it seems that those publicly arguing for a continued focus on case work prevailed, successfully promoting the view that it was too soon to be funding work towards the Court's legacy. As a result, the proposed $492,500 was frozen during the 2012–2013 budget approval process (ECCC Revised Budget, 2012–2013). By the start of 2014, ‘legacy’ had disappeared from the ECCC annual budget altogether (ECCC Proposed Budget, 2014–2015) and with it the opportunity it provided for critical early thinking on what might happen to the ECCC's documents and records.

By 2014, Craig Etcheson (former CGP Project Manager and investigator for the ECCC Prosecution) was claiming in an Opinion piece in The New York Times that the UN and the Cambodian government had agreed that the ECCC archives would ‘remain in Cambodia and under the control of the Cambodian government’, something that he considered made them ‘vulnerable to tampering or being sealed’ (Etcheson, 2014). David Scheffer, who was by then the UN Secretary General's Special Expert on UN Assistance to the Khmer Rouge Trials (and in that capacity negotiating on ECCC completion strategy, including the future of the ECCC archives) rejected this (2014), stating that nothing had been decided. DC-Cam's Youk Chhang shortly thereafter (in Freeman, 2014) announced that DC-Cam would provide access to ECCC records. Exactly how this would be done, given that DC-Cam, as a private entity, only had access to the already-public documents, was not clear. Such speculation and contestation over ECCC documents and records can be understood as an index of their perceived importance.

In 2015, the Cambodian government formally established the LDC relating to the ECCC (LDC-ECCC, hereafter LDC) as a repository for and to facilitate public access to the public documents of the ECCC trials (LDC, 2017). The LDC has a status equivalent to a Cambodian government department and is founded on an exchange of letters between the two parties to the ECCC (the Cambodian government and the UN) that was followed by a Cambodian Sub-Decree in November 2015 and a Memorandum of Understanding between the ECCC and LDC, signed in October 2016. In 2018, the provision of access to public records at the LDC was recognised by the ECCC Trial Chamber as a collective and moral reparation in the Case 002/02 judgement against Nuon Chea and Khieu Samphan (ECCC TC, 2018: 4429; see also Sperfeldt 2022). The LDC was initially housed in a building shared with the Cambodian Bar Association in Phnom Penh's northwest suburbs, but later moved to a building adjacent to other government departments in the city centre. The LDC received its first set of records – public documents pertaining to Case 001 – in late 2016. Both digital (soft) copies and hard copies were provided, with hard copies transferred to a thermally controlled, state-of-the-art archives room, facilities funded by a donation from the Government of Japan.

The exact parameters of what will eventually be held at the LDC remain unclear. The MOU between the LDC and the ECCC is confidential, and therefore provides no clarity for external observers. While the full name of the archive – the LDC relating to the ECCC – appears inclusive of any document relating to the Court, the relevant sub-decree refers to ‘public documents’ of the ECCC. Much hinges on what the category of ‘public documents’ might eventually include (see ECCC Supreme Court 2023), and the work of reclassification is currently underway. It is also unclear whether the LDC will hold records that the ECCC only has copies of, such as those held by DC-Cam and Tuol Sleng, or copyrighted records, such as books and films that were part of the ECCC case files. The LDC's digital copies are organised into a ‘Virtual Tribunal’. While there is also a ‘Virtual Tribunal’ at the Center for Human Rights and International Justice at Stanford University, the LDC's Virtual Tribunal is a different database and system. The Center and the LDC had discussed possibilities for an ongoing collaboration, but the collaboration did not eventuate when the ECCC legacy line item in the 2012 budget was frozen. Although the records in the LDC Virtual Tribunal are digital, in order to access them, one must visit the LDC in person or request a staff member to email them, in which case they arrive with an ‘LDC’ watermark across each page. In this way, LDC records are visually ‘owned’ by the LDC in a way that is not common for legal archives or databases, for example, the International Residual Mechanism for Criminal Tribunals, an archive of the now ceased International Criminal Tribunals for the former Yugoslavia and Rwanda, or the ICC Legal Tools Database, which holds records from not just the ICC but also a range of national and international bodies.

Despite the existence of the LDC since 2015, DC-Cam has long purported to be the most appropriate institution to host the ECCC's documents and records. As noted earlier, DC-Cam has provided (via the Cambodia Tribunal website) digital access to some public ECCC records. In 2019, DC-Cam published findings from its own ‘Public Opinion Survey on the Disposition and Custody of the ECCC archives’, conducted with the support of USAID. It found that a majority of the 1500 people surveyed wanted the archives to be preserved in an ‘independent and neutral’ location. Four institutions were named: the National Library, DC-Cam, Hun Sen Library and the Library of the Ministry of Justice. It is unclear whether survey respondents chose these from a predetermined list, and if so, whether that list initially included the LDC or even a residual version of the ECCC itself. DC-Cam's self-presentation as the most suitable repository for ECCC documents and records continues its longer bid to become and remain Cambodia's premier archive related to the Khmer Rouge regime.

On 31 December 2022, the ECCC ceased its judicial functions, and on 1 January 2023 commenced a phase of residual functions. This followed a 2021 report by the UN Secretary General on the proposed residual functions of the ECCC (UNSG 2021), which, rather than closing the Court and creating a new entity for residual functions, recommended keeping the ECCC for an initial period of three years but reducing its scope to ‘residual functions’. Notably, the report presents the archives as a key component of the ECCC's legacy (UNSG 2021: 38). In a new Annex to the original 2003 Agreement that was presented with the Report, the two parties to the ECCC agree to preserve the Court's archives ‘in accordance with international standards and [ensure] that they are as broadly accessible as possible’. Each of the two parties will ‘have a complete set of the archives’, and once the residual functions phase ends, this will include all classified documents and materials (Annex, Article 3). It is also notable that the LDC is not mentioned in the Report or the Annex. This may be because the LDC operates as a Cambodian department and the UNSG considers its role and function to be a question for the Cambodian government. But given the recognition of the public access provided by the LDC as a reparation in Case 002/02, and its ongoing work to establish itself as a public repository, this absence in the Report and Annex is curious.

On becoming a residual mechanism, the ECCC moved from its original complex in the outskirts of Phnom Penh, to the same building as the LDC in Phnom Penh's city centre. A new thermally controlled repository for records was built on the ground floor, and in September 2023, a publicly accessible Resource Centre opened. At the Centre, anyone can access public records via the Centre's computers or by viewing copies of some public case files, as well as access other resources relating to the ECCC, ICL and the Khmer Rouge period. With large and small meeting rooms, a Moot Court and both soft and hard copies of public ECCC records, the Resource Centre aspires to become a research and information hub for the general public.

The relation between the ECCC Resource Center and the LDC is unclear, particularly as the two are now located in the same building. The LDC, having initially prepared the whole building for its work, was considerably reduced to one side of the building with the early 2023 arrival of the ECCC. The space it had prepared for the public with computers and ECCC-related materials was reassigned to the ECCC Resource Center, and the room it had prepared to store original documents was replaced by the new room on the ground floor prepared by the ECCC. The relation between the two institutions is, at the time of writing in early 2024, very much still in development.

Public access to ECCC records will shape and be shaped by the future role of these archives. This shaping includes decisions on (re)classifications – records being and becoming public, public but with redactions, or wholly confidential – and issues of practical management. In March 2023, the ECCC Supreme Court Chamber issued guidelines for the reclassification of documents pertaining to Case 002. The Chamber decided on a presumption of public records, except for reasons such as personal security and privacy, in particular relating to witnesses and civil parties (victims participating in the cases, see Elander 2018; Killean 2018), a sensitivity long sought by victims’ lawyers. Shortly thereafter, the Chamber adopted the same guidelines for Case File 004/2. While the guidelines provide some clarity, it remains unclear whether they will also apply to records for the more controversial Case 003, which never proceeded to trial. Another uncertainty is whether the public will have access to ECCC records that do not relate to any particular ECCC case, such as memos, memoranda of understanding and other working documents. It also remains unclear how access to redacted records will be managed practically, given that the reclassification for Case 002 alone involves some 24,000 documents comprising 260,000 pages, some of which will be made completely public but others in part redacted, and in three languages. Finally, the management of confidential records beyond the three-year residual phase remains unclear, as does the status of its Resource Center beyond this period (also see Tan 2022). Under the new Annex to the ECCC Agreement, the UN and the Cambodian government shall each have a full set of the archives (contra Etcheson 2014 above), which will include classified documents and materials once the residual functions are completed (Annex Art 3.3). Given that the residual phase is currently scheduled to end in December 2025, and two archives are purportedly pressing for a public legacy for the ECCC, just who will manage public and confidential records, and how this will be done, are urgent and critical questions.

The politics of archival location and access to potentially sensitive documents notwithstanding, the pluralisation and activation of Khmer Rouge records across these archives continues to be cyclical in nature. Ongoing contestations and creations of archives in Cambodia affect the understanding of ever-proliferating records: as accessible, as a judicial legacy, as a precarious historical record or as politically loaded spoils. The political and economic capital gained from controlling access to these records is attractive to both state and non-state actors, not least because of the leverage for additional project funding that they afford. ⁷ In this way, the imperative of ‘entrepreneurial justice’ can be seen to continue beyond amassing and delivering records into justice processes into the capitalisation of archival afterlives.

Conclusion

In November 2018, the LDC was a space of anticipation. Public access to ECCC trial records as provided by the LDC had just been recognised in the ECCC Trial Chamber Judgment (2018) as a collective and moral reparation, and the relief and excitement was palpable. On our visit, staff showed us the archive room with its thermal control, a Public Reading Room, public access computers and meeting rooms for civil parties. At this time, most of the shelves in the archive room were empty with only a few shelves holding printouts of public documents from the trial in Case 001. In late 2023, it was the ECCC Resource Center that was a space of anticipation. While the LDC had been confined to a smaller section of the building, the ECCC Resource Center buzzed with possibilities and promise for future uses. This sense of futurity is arguably characteristic of all archives and archiving (Derrida, 1995) but takes on a particular charge and poignancy in the case of post-conflict and post-accountability contexts. In Cambodia, a 16-year legal process has followed a near 30-year wait for justice.

Archives created for the purposes of evidencing international crime are future oriented; they are prepared so records can have future use. But as we have shown here, even as there is a strong future orientation, the activation of records is cyclical and pluralised across archival institutions. Records are created for particular purposes, yet often exist long beyond these purposes. For example, records created at S-21 to identify ‘traitorous elements’ have since been used to evidence persecution and torture. When extrapolated from their contexts, the records come to mean something else and can help enact profound change.

For ‘transitional archives’ or ‘justice archives’, the connotation between records and evidence is particularly fraught. In this article, we have demonstrated how expectations of an evidentiary form – as imagined by civil society actors and as practiced differently by different courts – presses on records. When pressed into service as evidence, previous activations and prior contexts can be determinative. For example, the breadth and framing of DC-Cam's work (the context in which records were created or collected) provided grounds for defence lawyers to question their records. While the ECCC Trial Chamber eventually found DC-Cam's records reliable, the questioning is a salutary caution for organisations that aim to provide legal evidence. The main reason for our attention to the operations of DC-Cam and to DC-Cam held records as evidence at the ECCC lies with DC-Cam's ascendence to the position of ‘an arbiter of memory in Cambodia’ (Manning, 2017: 97). DC-Cam is a powerful knowledge broker whose claims to records are rarely questioned (e.g., Caswell 2013, 2014).

As is characteristic of entrepreneurial justice (Burgis-Kasthala, 2020), DC-Cam is a private institution that stepped in to fulfill a role that state actors were unable or unwilling to play. It has since made strategic innovations that have kept funding and interest flowing. DC-Cam's engagements with the ECCC exemplify what difficulties might arise when civil society actors provide a court with archival material and are called to the stand – namely contestation over authority, archival organisation, the provenance of evidence and bias. This private actor's ongoing claim to be the legitimate repository for all records pertaining to the Khmer Rouge – even those generated by the ECCC – is suggestive of the competition even ‘successful’ justice archives may eventually face. It is in this context of entrepreneurial competition that the ECCC is using its residual functions phase in part to revisit the value of its legacy archives. The relations between the ECCC Resource Center, the LDC, DC-Cam and Tuol Sleng will be key here. In this pluralisation, ECCC records will take on yet new meanings, imprinted by the legal process through which they have passed. Being ‘always in a process of becoming’ (McKemmish, 2001: 334), the records have the potential to shape accounts of that process, and Cambodian historical dialogue and accountability in fields anew.