War-Crimes Mass-Grave Excavation: ICTY, EAAF and Rwanda

The Equipo Argentino de Antropologia Forense (Argentine Forensic Anthropology Team, EAAF) was founded in 1984, the year Argentina's military junta relinquished power. Between 1976 and 1983, the junta's security services had abducted, tortured, and killed an estimated 30,000 people, the Desaparecidos (the Disappeared), whose fates were denied by the state. Many bodies had been dumped in the River Plate or buried in clandestine graves in municipal cemeteries under false names, or thrown from aircraft over the ocean during the so-called vuelos de la muerte (death flights).

The newly-elected government of Raul Alfonsin established the Comision Nacional sobre la Desaparicion de Personas (CONADEP) in 1984, which documented 8,961 verified cases and called for forensic investigation of clandestine graves. The Argentine judiciary invited Clyde Snow, then with the FAA's Civil Aeromedical Institute in Oklahoma City, to Buenos Aires in 1984 to advise. Snow, who had spent two decades doing skeletal trauma analysis on aircraft accident victims, looked at the problem and realized that the standard approach of simply removing bones from graves and describing them was scientifically useless without rigorous archaeological excavation context: who was adjacent to whom, what body position suggested about how the person had died, what artifacts were associated with which individual.

Snow trained a group of Argentine medical and anthropology students in forensic archaeological excavation and osteological analysis. That group became the EAAF. Mercedes Doretti, then a social anthropology student at the University of Buenos Aires, was among the founding members. The EAAF's first major excavation was at the Grand Bourg Municipal Cemetery, where an anonymous collective grave yielded remains that could be matched against CONADEP case files. The osteological evidence (gunshot wounds to the skull, wrist positions consistent with having been bound, age-at-death consistent with victim profiles) provided the first scientific documentation of systematic execution.

The EAAF's most historically significant early case was the identification of Liliana Carmen Pereyra, a young woman abducted in October 1976, one of the first Disappeared to be scientifically identified from skeletal remains. The identification used odontological comparison, osteological age and sex estimation, and the matching of a healed wrist fracture documented in antemortem medical records. The Argentine Federal Court accepted the identification, and the evidence was used in the 1985 Juicio a las Juntas (Trial of the Juntas), the first time any Latin American government put former military leaders on trial for crimes against humanity. Clyde Snow's expert testimony in that trial established the forensic anthropologist as a legitimate courtroom witness in human rights cases, a role that had not previously existed.

The EAAF's international expansion followed directly. By 1992, EAAF teams were working in El Salvador (the Mozote massacre site, investigated jointly with the Truth Commission), in Guatemala (sites from the 1960-1996 civil war), and in Chile (identifying victims of the Pinochet regime). In the decades since, the EAAF has worked in Peru, Mexico, Indonesia (investigating 1965-66 killings under Suharto), the Philippines, Bolivia, Uruguay, and Colombia. The EAAF's methodology, rigorous stratigraphic excavation followed by individual skeletal inventory and biological profiling, DNA sampling, and careful reconciliation against antemortem records, became the template that the ICTY forensic team and later the ICMP adopted.

Mercedes Doretti, who became co-director of the EAAF and its international operations coordinator, brought the EAAF model to Bosnia. She served as a forensic anthropology consultant to the ICTY during the first wave of Srebrenica excavations in 1996 and helped establish the standard operating procedures that the ICTY Office of the Prosecutor's forensic team used through 2001.

The International Criminal Tribunal for the former Yugoslavia (ICTY) was established by UN Security Council Resolution 827 in May 1993. By 1995, the ICTY had jurisdiction over war crimes, crimes against humanity, and genocide committed in the former Yugoslavia since 1991. The Srebrenica massacre of July 1995, in which approximately 8,000 Bosnian Muslim men and boys were killed in the Srebrenica enclave and surrounding areas, was from the outset the ICTY's most important investigation.

The ICTY forensic investigations at Srebrenica began in 1996 under the direction of William Haglund, who had worked as a forensic anthropologist for the King County Medical Examiner in Seattle and who had recent field experience in Rwanda. The excavations were conducted under difficult conditions: the sites were in an active post-conflict zone, access was politically contested, and the Bosnian Serb authorities initially denied that large graves existed. Aerial photographs taken by US reconnaissance satellites and leaked to the press in 1995 had already identified several grave locations, and the ICTY used these to direct the first excavations.

The primary graves at sites including Branjevo Military Farm, Cerska, Nova Kasaba, and Lazete were excavated between 1996 and 2001. The forensic findings established the pattern: large numbers of adult male skeletons (consistent with the demographic profile of the separated population), perimortem gunshot wounds (fresh bone fractures consistent with bullet trajectories, copper bullet jacket fragments, lead core fragments), ligature marks on wrists (cord or wire), and blindfolds of cloth or plastic wrap. The biological profiles, the age estimates (concentrated in the 16 to 60 male age range), and the forensic pathology findings established that the deaths were executions, not combat casualties.

The secondary and tertiary grave problem was the most complex osteological challenge the investigation faced. When the primary graves were disturbed in September 1995, the disturbance did not produce clean transfers: individual skeletons were fractured and partially moved, meaning that one individual's remains could be split across a primary grave and one or more secondary graves kilometres away. The ICMP's approach (the ICMP took over identification work from the ICTY in 2001 and has continued since) uses DNA as the primary tool for connecting body parts across sites. When a DNA profile from a femur in a secondary grave matches a DNA profile from a skull in a primary grave, the connection is made: both fragments belonged to the same individual, and the geographic connection between the two graves proves that disturbance occurred.

The legal impact of the forensic evidence has been substantial. In the Krstic trial (2001), the ICTY Trial Chamber found that Radoslav Krstic, as commander of the Drina Corps during the Srebrenica operation, bore responsibility for the killings. The forensic evidence, including the osteological MNI calculations, the MLNI analysis that estimated the true individual count was higher than the element-by-element MNI, and the DNA identifications, was central to the Chamber's finding that the killings constituted genocide. The MNI and MLNI methods used at Srebrenica are explained in full in DVI from the osteology angle: MNI and commingled remains. Krstic was convicted; the Appeals Chamber in 2004 upheld the genocide finding. In the Mladic trial (verdict 2017), the same forensic record, supplemented by more than 20 additional years of ICMP identifications, supported convictions for genocide and crimes against humanity.

The Rwandan genocide of April to July 1994 killed an estimated 500,000 to 800,000 people, overwhelmingly Tutsi civilians, in approximately 100 days. The International Criminal Tribunal for Rwanda (ICTR) was established by the UN Security Council in November 1994 to prosecute those responsible. Unlike the ICTY, which operated alongside an active forensic team in the field, the ICTR initially struggled to mount systematic forensic investigations of the massacre sites.

The Kibuye Catholic Church massacre, in which several thousand Tutsi civilians who had sought refuge in the church and surrounding sports complex in Kibuye prefecture were killed on 17 April 1994, became the focus of the ICTR's first significant forensic investigation. Clyde Snow, by then in his seventies and already the acknowledged founder of war-crimes forensic anthropology through his Argentina work, led the initial survey team to Kibuye in 1995. Snow's team documented skeletal remains still present in the church and on the surrounding grounds, conducting an initial skeletal inventory and estimating MNI from the element counts.

Physicians for Human Rights (PHR) conducted the subsequent excavation at Kibuye under contract to the ICTR Prosecutor's office. PHR's team, which included multiple EAAF-trained anthropologists and pathologists, excavated mass graves in the churchyard and conducted a full mortuary analysis. The findings included skeletal evidence of perimortem blunt-force trauma (cranial depressed fractures and radiating fracture patterns consistent with machete blows and club impacts), panga (machete) cut marks on long bones and skulls, and gunshot wounds. The demographic profile of the recovered remains was consistent with the civilian refugee population: adults of both sexes and sub-adults of all ages, with no combat-age male concentration (unlike Srebrenica, the Kibuye victims were not sex-selected before killing).

The ICTR proceedings established several significant legal principles for war-crimes forensic anthropology. In Prosecutor v. Kayishema and Ruzindana (1999), the ICTR Trial Chamber accepted osteological evidence (skeletal MNI counts, trauma patterns, demographic profiles) as proof that mass killings had occurred at specific sites. The Chamber also accepted that the demographic profile of victims (civilian men, women, and children) was inconsistent with the claimed defence that the killings were combat-related. The osteological evidence thus served both to establish the fact of killing and to characterise the nature of the killing as genocide rather than armed conflict.

The Khmer Rouge regime of Democratic Kampuchea, which governed Cambodia from April 1975 to January 1979, killed an estimated 1.5 to 2 million people, approximately 25 per cent of Cambodia's population, through execution, starvation, forced labour, and disease. Execution sites, mass graves, and the bones of victims were widely visible and extensively photographed from the early 1980s onwards, when Vietnamese forces and international journalists accessed the country after the Vietnamese invasion ended Khmer Rouge rule.

The primary forensic challenge in Cambodia was not finding the mass graves but preserving them from inadvertent destruction. The Choeung Ek site outside Phnom Penh, now the most-visited Killing Fields memorial, had been partially excavated by Vietnamese-backed Cambodian authorities in 1979-1980, with remains exhumed and placed in a memorial stupa. This excavation, conducted without systematic archaeological protocol, destroyed stratigraphic context that would have been essential to scientific analysis. Many other provincial sites were similarly disturbed.

The Extraordinary Chambers in the Courts of Cambodia (ECCC), the hybrid tribunal established in 2006 jointly by the Cambodian government and the United Nations, relied on documentary evidence, survivor testimony, and photographic documentation more heavily than on osteological analysis, in part because of the disturbed state of many sites and in part because the passage of nearly three decades since the killings made contextual osteological evidence difficult to interpret. International forensic anthropology advisors, including consultants from the EAAF and from the Dutch NFI, advised the ECCC on the interpretation of skeletal evidence recovered at specific sites.

The DC-Cam (Documentation Centre of Cambodia), established in 1995 with funding from the US Department of State, has conducted extensive documentary mapping of Khmer Rouge mass grave sites across Cambodia, identifying over 20,000 mass grave sites. This mapping project, though not primarily an osteological exercise, has informed the design of future forensic investigations if the Cambodian political situation permits.

The lesson from Cambodia for the broader field of war-crimes forensic anthropology is methodological: the absence of rigorous archaeological excavation protocol, even by well-intentioned authorities in the immediate post-atrocity period, can permanently compromise the forensic evidence base. The recommendation arising from the Cambodia experience, formalised in the 2001 Stover-Ryan framework for war-zone forensic anthropology ethics (discussed in Section 6), is that early intervention with trained forensic archaeologists must be a priority in the immediate post-conflict phase, before political authorities or grieving communities disturb sites with their own excavations.

The Anfal campaign, conducted by the Iraqi government of Saddam Hussein against the Kurdish population of northern Iraq between February and September 1988, killed an estimated 50,000 to 180,000 people and involved the use of chemical weapons, including the Halabja attack of 16 March 1988 in which mustard gas and nerve agents including tabun and sarin were used against a civilian town. Most of the Anfal victims were men and boys taken from their villages in military sweeps and killed in mass executions, with bodies buried in remote desert sites in southern Iraq or dumped in the Hamrin Mountains.

Post-2003 access to Iraq allowed the EAAF, Physicians for Human Rights, and the UK Forensic Archaeology Group (part of the UK Crimes Against Humanity team) to begin systematic investigation of Anfal mass grave sites. The scale was unprecedented: an estimated 271 mass grave sites had been identified across Iraq as of 2006, of which the Anfal sites in Kurdistan and the execution sites from southern Iraq (victims of the 1991 Shia uprising suppression) were the most numerous.

The osteological evidence from the Anfal sites documented the sex-selection pattern consistent with other genocide-execution contexts: adult male skeletons predominated, with some sites containing almost exclusively males and others containing family groups. Ligatures, blindfolds, and perimortem gunshot wound patterns were consistent across sites. The EAAF's molecular identification work, supported by mitochondrial DNA analysis and kinship comparison against reference samples from Kurdish families, has produced identifications at several sites.

The evidence was used in the Iraqi High Tribunal's proceedings against Saddam Hussein (executed December 2006), against Ali Hassan al-Majid (Chemical Ali, executed January 2010, convicted of genocide for the Anfal campaign), and in subsequent proceedings against other Anfal commanders. The IHT accepted forensic anthropology testimony as expert evidence under Iraqi law. The Anfal campaign was ruled to constitute genocide against the Kurdish people.

In parallel, the post-Saddam period saw investigation of mass graves from the 1991 suppression of the Shia uprising in southern Iraq, particularly in the Najaf and Karbala regions. These graves contained different demographic profiles (more family groups, including women and children) than the Anfal sites, consistent with the nature of the 1991 suppression as population-wide rather than sex-selected.

Eric Stover, the director of the Human Rights Center at the University of California Berkeley, and Rachel Shigekane published a framework for the ethics of war-zone forensic anthropology in 2002, building on Stover's earlier work in 1999 with the International Committee of the Red Cross. The Stover-Ryan framework addressed four ethical tensions that are specific to forensic anthropology in conflict and post-conflict settings, tensions that do not arise or arise differently in domestic casework.

The first tension is between investigative thoroughness and the needs of survivors. Families of the missing want exhumation to proceed quickly, so that they can have bodies returned for burial and begin mourning. Rigorous archaeological excavation is slow: a properly excavated mass grave site may take weeks or months to fully document. The pressure from families and from governments eager to demonstrate accountability can lead to rushed excavations that destroy contextual evidence essential for prosecution. The Stover-Ryan framework recommended that forensic teams establish a formal family liaison function, communicate regularly about the pace and purpose of the investigation, and resist pressure to accelerate in ways that compromise evidence integrity.

The second tension is between the scientific independence of the forensic team and the political interests of the bodies that fund the investigation. ICTY investigations were funded by the UN; EAAF investigations are funded by a combination of foundations, governments, and the families themselves. The funding source creates potential pressure: a government funding an investigation into atrocities committed by a rival may be tempted to request selective disclosure of findings. The Stover-Ryan framework required forensic teams to publish their findings in full and to resist selective disclosure, noting that partial disclosure is itself a form of evidence manipulation.

The third tension is the survivor identification versus criminal evidence tension. Human remains serve two functions in war-crimes contexts: they are evidence for prosecution, and they are the family's loved one. The decision of when to return remains to families (which can be done relatively quickly after osteological and DNA analysis) versus retaining them as physical exhibits in an extended prosecution process creates genuine ethical difficulty. The EAAF developed the practice of making full scientific documentation (photographs, measurements, biological profiles, DNA profiles) before returning original bones to families, so that physical return does not compromise prosecution evidence.

The fourth tension is the safety of the forensic team itself. The admissibility standards and qualified-opinion framework that govern expert reports emerging from war-crimes casework are mapped in the ABFA, SWGANTH, ENFSI FAWG quality frame. War-crimes forensic anthropologists routinely work in areas where the perpetrators of the crimes under investigation retain political or military power. EAAF team members have received death threats in Guatemala, Mexico, and the Philippines. The ICTY forensic team worked in a demilitarised zone under NATO protection. The PHR team in Rwanda worked under UN security protocols. The Stover-Ryan framework noted that physical safety of team members must be a primary operational planning consideration, not a secondary concern subordinated to the investigation schedule.

1. Site security and perimeter establishment
   Establish a controlled perimeter around the grave site in coordination with security forces (UN, NATO, or local law enforcement as appropriate). No unauthorised entry. Site coordinator controls access log.
2. Surface documentation
   Total-station survey of the site perimeter and all visible surface features. Aerial photography by drone or aircraft if available. GPS coordinates of all significant features. All documentation timestamped.
3. Grid establishment
   Establish a 1 by 1 metre (or 2 by 2 metre for large sites) excavation grid from a fixed datum point. Grid squares labelled alphanumerically (A1, A2, B1...). The datum point is referenced against national grid coordinates for the jurisdiction.
4. Stratigraphic excavation
   Excavate by trowel within each grid square, working in natural strata or arbitrary 5 cm horizontal datums. Soil matrix from each stratum sampled for pollen, entomological evidence, and soil chemistry. No mechanical excavation until the element zone is reached.
5. In situ documentation of elements
   Each element photographed in situ before lifting. Position plotted on the grid plan. Depth from surface recorded. Associations (clothing, ligatures, blindfolds, artifacts) documented photographically and in the excavation log before disturbance.
6. Bagging, labelling, and chain of custody
   Each element or associated artifact placed in a labelled evidence bag with: site code, grid square, stratum number, element description, date, and excavator initials. Chain of custody form begun. Bags sealed and signed by excavator and scene supervisor.
7. Mortuary analysis and DNA sampling
   Transported under seal to the mortuary facility. Skeletal inventory (Buikstra-Ubelaker 1994). Biological profile. Trauma documentation. DNA sample collection (petrous bone preferred for degraded remains). DNA reference samples from families coordinated through the family liaison function.