Preliminary telephone communications

To collect qualitative and quantitative evidence on the actual situation on the ground in Tigray, we conducted numerous interviews with people from different woredas (districts) in Ethiopia. This was possible given our long-term involvement in Tigray and elsewhere in the country as researchers, through family links, as previous staff of the Bureau of Agriculture, as lecturers… Over time, a large network of friends and colleagues has been established.

In the context of these interviews, it should be noted that not only travel to, but also communication lines in Tigray were blocked on 4 November 2020, at the start of the war. In the month of November, there has been continuous black-outs for telephone lines as well as for internet connections. At that time, we started sending out email messages broadly to all our friends in Tigray, hoping that such “messages in a bottle” would be picked up by some (e.g. through satellite telephones) and could be distributed somehow. This also was the time at which PM Abiy Ahmed promised to conquer the town of Mekelle and the Tigray Region “at any cost” - which was later reworded in terms of “conducting a final offensive”.

From the end of November onwards, we managed to have some rare contacts through people finding telephone signals on mountain tops near the borders of Tigray. After December 2020, telephone communication had slowly become possible with Mekelle and a few other major towns in Tigray (with intermittences). It was at that time that the first narrations of large-scale massacres reached us and the outside world. Until then, only the May Kadra massacre was known, which was largely used by the Ethiopian government as a justification for warfare.

Collection of reports of massacres and civilian casualties

By January 2021, email connection, at least with Mekelle, became possible and the first lists of victims and rare videos of massacres started to circulate. These were than often further distributed through Twitter and other social media.

Our casualties database has been populated from a mix of sources, ranging from social media posts, media reports, advocacy groups listings (for instance Irob Advocacy) and direct reports (as posted for instance on the Tghat news site). The social media posts are mostly from family members and friends abroad who mourn the death of their loved ones, which they learnt about by telephone. For each victim, through our network, we have tried to contact one relative or friend to learn more about the circumstances in which the victim died. For this verification, some families also have provided a photo of the deceased person.

Among the incidents, there are many where the names of the victims are not known; the reporting of such incidents is based on media reports, reports of different NGOs and humanitarian actors and press releases. As lack of transportation and communication blockage in the region make reporting very difficult, the number of verifiable sources is minimal (the 5026 reports have been filtered down to 590 ‘events’) and the number of casualties is most likely an extreme underrepresentation. Significant cross-checking was done, starvation cases removed, killing sprees over several consecutive days considered as one ‘event’. A few duplicates were also removed.

Detailed verifications

By the end of April 2021, many of our interviews could only take place when people came to Mekelle or other major towns to get news from their relatives, or for cash withdrawals from the bank, which was then possible at a limited number of locations.

During May-June, when telephone connections were restored in many parts of Tigray, we managed to get in contact with our broader network again and could conduct many more interviews, in which we systematically verified circulating lists with other people from the same area. This database is largely founded on those telephone calls, for which we estimate to have included over 2000 telephone interviews. All interviewees accepted only to give information on the condition of strict anonymity.

At the start of our interviews in early January, people were afraid to speak, and used very indirect terminology. Starting May, most people talked openly on the situation and the numerous killings and other war crimes. One of the main reasons for this evolution was that there were hundreds of thousands of telephone calls narrating the plight to relatives. This makes that our correspondents assumed that when they transmitted information, it would not be singled out by security services. Another important reason was that the accumulation of negative war impacts had made people to abandon their traditional precautions when communicating plight – which traditionally cannot be done over the phone.

Our interviews have allowed a sampling of the casualties, but large parts of the full picture still have to emerge. Important to note also is that the magnitude of the humanitarian crisis that we contribute to unearthing through our interviews has been confirmed (for the worse) once media have had access to some parts of Tigray. And yet, in March-June these accessible parts were still small, with most media operating from Mekelle and reporting on Eastern and Central Tigray. After July 2021, journalists have not been allowed to access Tigray, and had to establish their private communication with contact persons.

Representativeness of the casualties and massacres dataset, and interpretation

We have noted that there are many families who, for various reasons, do not report the loss of relatives on social media, which limits our sample of fully documented casualties. According to tradition, in Tigray people do not tell the death of a person to relatives living abroad, but wait until meeting in person. This might have contributed to underreporting of casualties. Additionally, there are also massacres and casualties in remote areas and victims that may not be locals of the area (e.g. internally displaced people or Eritrean refugees), which makes it difficult to fully document these deaths.

The communication blackout and lockdown of the region make it very hard to get verified information, so the actual number of deaths is likely much higher than the sample that we have collected so far. An exhaustive inventory of all massacres would require visiting all villages and towns of Tigray and talk with the inhabitants. And then, the forensic investigation, with excavation of mass graves, still has to start.

For a cartography and interpretation of victims, authors, timelines of the massacres, we refer to the Tigray: Atlas of the Humanitarian Situation.