2 Establishing the basics: significance, method and definition

It is scarcely necessary to emphasise the proliferation of mass graves across the globe. Indeed, as pointed out by the then Special Rapporteur on extrajudicial, summary or arbitrary executions, ‘[t]here is not one region in the world, not one historical period, that has not seen mass graves’ (UNGA 2020, para. 6). Moreover, news from Sudan (UNOHCHR 2023), Ukraine (Harding, for The Guardian, Reference Harding2022) and Gaza (Massud and Victor Reference Massud and Victor2023) amply underscore the topicality of the subject.

Over the past decades, strides have been made in the way that mass graves, together with their ramifications for victims, families and communities, are approached and understood. Forensic organisations have been instrumental in shaping scientific approaches taken to mass graves. Civil society organisations across the globe (e.g. Journalists for Democracy in Sri Lanka (JDS), the International Truth and Justice Project (ITJP), Center for Human Rights and Development (CHRD) and the Families of the Disappeared (FoD) 2023) engage with mass graves; international organisations advocate for and facilitate the resolution of missing persons cases (such as the International Commission on Missing Persons (ICMP) and the International Committee of the Red Cross (ICRC)); and legally informed, standard-setting scholarship to protect and investigate mass graves is readily available (Klinkner and Smith Reference Klinkner and Smith2020).

It is against this background that we sought to understand the benefits and risks of mass grave mapping, since mapping has the propensity to capture the magnitude and ubiquity of the existence of mass graves across the world. In other words: under what circumstances can mass grave mapping contribute to much needed sustainable engagement with the myriad of mass graves embedded in the earth’s surface? Such an enquiry will also lay the foundation for a fully reasoned response and approach to the Special Rapporteur on extrajudicial, summary or arbitrary executions’ call for global regularised mapping (UNGA 2020). Producing rights-informed mapping guidance will assist in realising this ambitious goal for a global mass grave map, designed to capture and recognise the magnitude and significance of mass graves internationally.

The project findings presented here stem from an investigation of real-world practice and experience that is grounded in turn in academic literature. At a foundational level, the research team conducted a detailed review of existing mapping projects, to identify the human rights agendas and assumptions that underpin them (see Section 3 for a full discussion). The review also encompassed the growing body of literature on the use and role of mapping, digital evidence and documentation for accountability and justice purposes (for example Freeman Reference Freeman2018 and Gillett and Fan Reference Gillett and Fan2023), including in relation to their purported benefits to the solving of missing persons cases (Sarkin Reference Sarkin2021).

This desk-based literature and legal material review was complemented by and tested through country case-studies. The chosen countries were Bosnia, Cambodia, Rwanda, Guatemala and Iraq. They reflect geographical diversity as well as different time periods, enabling analysis of shifting sociocultural, political and economic contexts and varying destruction experiences and risks. In all countries, mass graves are testimony to the unimaginable crimes committed there, with Bosnia, Cambodia and Guatemala having bespoke mass grave mapping projects (BIRN 2020; DC-Cam n.d.; and FAFG 2013), whereas Iraq has made some efforts (Yazda’s mass grave reports (Yazda 2018)) and Rwanda sees a focus on memorial sites that often are mass grave sites (ARCIGS n.d.). The selection of country case-studies also reflects differing levels and forms of mass grave destruction, expert and non-expert excavations, identification success rates and accountability mechanisms as well as local and international investigation efforts, thus offering a rich substrate to analyse the rights of all actors and affected communities. In all country settings, multiple missing persons cases remain unresolved, with many victims believed to rest in as yet unknown or uninvestigated mass graves.

For the purposes of the study and this paper, we operate with the following definitions:

Mass grave is as yet undefined in international law. However, the definition proffered in the Bournemouth Protocol on Mass Grave Protection and Investigation (Klinkner and Smith Reference Klinkner and Smith2020) has since been adopted in a number of UN documents (UNGA 2020; UNHRC 2022). For this study, we operate with the same definition:

By victim, we mean:

In the context of mass graves, having a definition that recognises secondary victims (i.e. families of the dead) is crucial; furthermore, the longevity of harm suffered by secondary victims, often across generations, deserves explicit recognition.

Open-source information is used here to mean ‘publicly available information that anyone can obtain by request, purchase, or observation without requiring special legal status or unauthorized access’ (Dubberley et al. Reference Dubberley, Koenig, Murray, Dubberley, Koenig and Murray2020, 9. See also Dubberley and Ivens Reference Dubberley and Ivens2022). In the context of mapping, we can distinguish the open-source materials or data (i.e. the information input) that are in the public domain but also then the actual form mapping takes (i.e. the research output), which could also be open-source.

Digital open-source information is used here to mean ‘open-source information on the internet, which can be accessed, for example, on public websites, internet databases, or social media platforms’ (Human Rights Center UC Berkeley School of Law and UNOHCHR 2022, 6).

Mapping is routinely geomatics driven but is used here to mean the diagrammatic recording of geographic features together with any associated and proxy variables (Congram et al. Reference Congram, Kenyhercz and Green2017).

The following section offers an analysis of the goals and benefits of mapping exercises in a human rights context.

3 Why mass grave mapping? An appraisal of goals associated with mapping projects

Mapping can have many different purposes and objectives; therefore an explicit statement of goals helps in understanding and clarifying the boundaries of the mapping exercise itself. In fact, the mapping of mass graves has been undertaken, sometimes as a bespoke mapping endeavour (BIRN and DC-Cam are prime examples); in other initiatives it is rather a by-product of past or ongoing investigations of mass graves (FAFG 2013); yet other examples list mass graves as part of sites where human rights violations were performed (such as Syria Justice and Accountability Centre 2024) and as part of memorialisation efforts (as in Rwanda). When reviewing literature as well as relevant mass grave mapping and human rights-related documentation project goals, a number of identifiable and evidence-based benefits can be discerned.

Generally, by making violations visible, including the scale and patterns of the human rights violation in question, a key driver is awareness raising. The following analysis is grouped around three sets of benefits (espoused, if not always strictly evidenced): the advancement of justice goals; the capturing of testimony and offering commemorative opportunities; and the pursuit of protection and monitoring.

4 An overview of applicable mapping steps

Goals dictate the direction and mandate of a mapping endeavour. But what does mapping entail? Based on our research and analysis, the following section offers an overarching understanding of the processes involved in mapping. Broadly, three phases can be distinguished: (1) planning; (2) data collection; and (3) presentation. These three phases are depicted in different colour blocks in the flowchart (Figure 1).

The flowchart, in its full design, is interactive and wherever a ‘hand icon’ can be seen, users can click to access a pop-up window (Klinkner and Smith Reference Klinkner and Smith2024). In the planning phase of a prospective mapping project, the starting point is to clarify rights (and corresponding duties) that arise in the context of mass graves (discussed under Section 5) to ensure a rights-informed approach to all other mapping activities. The contextual elements, including the ways in which the law in a country or region protects or engages with mass graves and its victims, are important. Similarly, the scope of the mapping exercise numerically, geographically and temporally will have to be clarified, including the actors involved in shaping the actual mapping project.

In many regards, the context and goals will be in dialogue and will shape what the project is seeking to achieve. From the scope of the project and its aim, the choice of methods employed, and risks follow accordingly, and with them a considered plan for data collation, storage and protection, usage, analysis and verification as well as a final presentation strategy have to be devised.

Only then can the actual data collation commence, depicted in red in the flowchart. For this stage, it is helpful to distinguish what type of information is collated. We distinguish between open-source, closed-source and mixed-source (i.e. a combination of open- and closed-source). Naturally, depending on what information is sourced, different training needs arise from which data collation, data entry and analysis follow.

In green then, is the presentation stage. Here we interject a ‘risk-check stage’ whereby risks to people, sites, evidence and society are considered before proceeding to the presentation as an open-source map, a map with restricted access area or a completely non-public-facing map that is behind protection with password access or through multi-factor authorisation. Common safety features should be put in place for the public, or part public-facing, maps to ensure errors can be rectified and edits made. Significantly, the process flowchart highlights the difference between consulting open-source material in the data collation phase and making the outcome of mapping available open-source. Indeed, the safety features available to mappers enable control over what is made publicly available, including in how much detail.

From the outset, for any mass grave mapping, it is worth highlighting operating principles that will govern any engagement with mass graves. While it is beyond the scope of this paper to elaborate fully on each of the principles (for this, see Klinkner and Smith Reference Klinkner and Smith2020, 6–7), the following operating principles will apply:

1. (1) The ‘do no harm’ principle, which for mapping has the potential to be far-reaching as it may include individuals, communities and even large populations. Any mapping attempt should thus actively seek to avoid undermining existing community cohesion, creating inequalities, instilling perceptions of bias or entrenching existing inequalities. It will be most likely rooted in, and adhere to, the cultural sensitivities, beliefs and norms of victims and/or their families.

2. (2) Physical and emotional safety of all those involved in the processes of mapping, from reporting to presentation of findings, are paramount and include considerations (without distinction) of dignity, privacy and well-being of the victims and their families.

3. (3) Independence and impartiality: without a non-discriminatory and impartial approach to the mapping process in its entirety, the legitimacy of the work may be questioned by the affected community. This is put to the test by the Balkan Investigative Reporting Network team based in Bosnia whose approach to mapping extends beyond Bosnia to include grave sites in Croatia, Kosovo and Serbia (BIRN/BalkanInsight 2020).

4. (4) Confidentiality: any data and information processing must be consistent with national legislation and data protection regulations. Organisations such as the International Commission on Missing Persons and the BIRN project mentioned above, as well as the Srebrenica Memorial Center, offer such reporting structures, but the perception of anonymity and confidentiality may depend on the mandate of an organisation as well as trust (which may differ from individual to individual).

5. (5) Transparency of process will help instil confidence in the rationale, methodology and methods employed in the entire mapping process.

6. (6) Accuracy and reliability of the data sources, analysed and presented as part of a mapping project are key. Pointing to the BIRN initiative again, it is worth noting that only fully investigated mass graves beyond a threshold of victim numbers are displayed on the map.

At a fundamental level these operating principles are linked to rights. They are based on the rights of all those potentially impacted by the indication of a mass grave on a map, and are instrumental to the protection of those rights, as well as to the safe production of the map. Respect for these principles will also support mappers as they seek to build relationships of trust with affected communities, and better ensure the validity of the map with the communities most directly affected.

5 Untangling the web of rights

Regardless of who undertakes mass grave mapping, whether at grass-roots level or as part of a top-down investigative process, a number of rights and rightsholders (not least the mappers themselves) will have to be considered. This is reflected in ‘Documenting international crimes and human rights violations for accountability purposes: Guidelines for civil society organisations’ (ICC/Eurojust 2022) which inter alia contains a section on burial sites and mass graves (ICC/Eurojust 2022, 36) as well as a section on documents and digital materials (ICC/Eurojust 2022, 30). Our research demonstrates that the rights potentially engaged by the mapping of mass graves are manifold and varied between the stakeholder groups.

6 Towards a data-gathering strategy: Ukraine pilot

Cognisant of what rights-informed mapping would require and mindful of the many goals that may be attached to mapping exercises, our next project step was to examine the methods that mass grave mapping might employ in relation to the data points, that is, the specific pieces of information and their corresponding value, gathered as part of the dataset. This is where the project conducted its pilot mapping exercise based on real-time emerging open-source information: when in February 2022 the full-scale invasion of Ukraine began, and open-source documenting and mapping casualties became a core activity, through the conduct of a pilot, close attention was paid to what surfaced in open-source media on the subject of mass graves. This was to get a better understanding of what setting up a mapping project during a time of war would entail and what issues this would raise for data-gathering strategies, specifically data points.

In the context of the ongoing armed conflict between Russia and Ukraine, mass graves in and around Mariupol made news headlines early in the conflict, with those found in Bucha perhaps the most prominent to date (Cotovio et al. Reference Cotovio, Pleitgen, Blunt and Markina2022). They have become widely associated with the possible commission of war crimes as investigators and forensic experts have started probing alleged crime scenes (Aljazeera 2022), including excavations of mass graves (France24 2022), and amidst many open-source investigations.

The aim of the pilot was to observe, collate and categorise information as reported on media outlets (including social media) on the subject of mass graves, to confirm and refine the data-entry points and identify associated challenges with data collation. Notably, there was no aspiration to deliver on any of the goals found in the literature or human rights projects as such, though, of course, the pilot was a systematic enquiry – but crucially, to inform the project itself. As a pilot, it was framed as an exercise of information collation. In addition, in line with the rights-informed approach designed to ensure the safety of stakeholders, it was clear that our pilot would not result in an open-source output (though relying on open-source input). The tapestry of rights engaged, as outlined above, is complicated, but particularly so during an ongoing armed conflict, hence this clear practical measure to not take any such risks and conduct open-source research remotely. That said, the well-being of the researchers who may be exposed to distressing images, narratives and information was carefully considered (and indeed was a concern that we had to mitigate against for ethical reasons at the outset of the entire MaGMap project). We therefore had a designated resource page with training materials on how to avoid any form of psychological harm.

Due to extensive engagement with the Bournemouth Protocol (Klinkner and Smith Reference Klinkner and Smith2020),Footnote ¹ the importance of the full life-cycle of a mass grave to collate information was clear: from the reporting, to the protection stage, to more detailed information as to who may be in the grave and how they died, to very important matters such as positive identification efforts but also justice – be that reparative, commemorative or indeed retributive.

This life-cycle must be understood, for it dictates what data categorisation might be utilised. With that in mind, our data-entry points were to capture:

• the emergence of mass graves;

• the recording of mass graves;

• the reporting on mass graves;

• information on victim numbers in mass graves;

• the protection of mass graves;

• the investigation of mass graves;

• post-investigative activities (including return of human remains; commemoration and justice efforts).

These data-entry points were captured in the pilot protocol to ensure uniformity in the approach to data collation and entry. In addition to this, there was a free-text data-entry point where the researchers could add any additional notes. In terms of keyword searches, we used: ‘mass grave’/‘massgrave’, ‘mass burial’, but also ‘missing persons’ and ‘enforced disappearance’ as proxies and for verification purposes.Footnote ² Core sources were Twitter/X (with #massgrave(s)) and Google News. We then identified several international organisations and official channels that might report on mass graves potentially confirming (or discrediting or omitting) prior reports.Footnote ³ Many of these entities’ updates and reports, as well as those from grassroots NGOs and organisations active in Ukraine in conducting OSINT investigations, were first captured on Twitter/X. The logic here was to provide cross-references with international bodies (UN entities and international organisations), followed by official domestic channels, international non-governmental organisations with relevant mandates and research bodies, including those operating in situ.

The amount of data in the main mass grave log (collated from English language sources) between April and June 2022 revealed thirty-seven alleged mass grave reports amounting to a total of 1406 data entries. This provided us with a useful substrate to triangulate with literature and rights-based considerations before drawing the following inferences:

1. (1) Due to the nature of the conflict, including loss, gain and/or control over territory, the speed of mass grave reporting and investigation was astonishing, especially in the early days when Russian forces were pushed back in and around Bucha. It is particularly noteworthy how quickly information was moving from the ‘discovered a mass grave’ stage to reportedly being investigated. And further, to a reburial (albeit without identification), resulting in graves marked with numbers as opposed to names, as no identification of the human remains had taken place (e.g. Arhirova Reference Arhirova2022). This has a significant impact on what data could be captured to ensure nuanced, accurate record keeping in relation to investigation, excavation and affiliated dates.

2. (2) The paucity of information from East Ukraine and Russia-controlled regions that was readily available was in stark contrast to information flows in territories controlled by the Ukrainian government. Indeed, it was when Russian troops left the territories that information emerged (Reuters 2022). Accessibility of information and accessibility to sites must thus be considered, which will be particularly pronounced during an armed conflict (and occupied territories). Naturally, this limitation also arises due to, and is exasperated by, the choice of media outlets and organisations that are scrutinised. The point here is to stress generally the issue of accessibility whether built into the method of data collation or due to physical constraints.

3. (3) A consequence of a lack of accessibility (both informative and physical) is an increased verification challenge, especially where information and evidence have potential utility as propaganda.

4. (4) Conversely, where there is an abundance of media reporting and a high information flow, this presents difficulties in making sense of a multitude of reporting of sites to ensure that they relate to one and the same site.

5. (5) Indeed, the accuracy of (and in) open-source reporting renders verification extremely difficult; it introduces elements of uncertainty into analysing, categorising and recording information.

6. (6) Allied to the issue of accuracy in reporting is the use of a clear definition: what is a mass grave? What is a mass burial site? This may not be a concern for a reporting journalist, but it will cause issues for a lawyer or mapper keen to detect the illegal activity connected to the burial structure which gives rise to the recording of a possible human rights violation. Of course, this will also have societal relevance (including political, social, community and cultural).

7. (7) This leads to a related issue: that of applying the definition to reports (i.e. is it a mass grave, a mass burial, a crime site, a killing site or a historic site?). Here we also noted a cautious approach by organisations such as the OSCE (Benedek et al. Reference Benedek, Bílková and Sassòli2022; Human Rights Watch 2022) and the UN, following the UN Secretary General’s visit, in using the term ‘mass grave’ (UN News 2022). This is in part due to the fact that there is no unified definition, but also because it indicates a gravity and severity of events that, in addition to attracting protection and the need to investigate, draws condemnation on behalf of survivors, families, affected societies and the international community.

8. (8) From the outset as signposted above and through our rights-informed approach, we conducted mapping in the safe data-protected space of our own organisation. However, the well-being of the researcher remained a concern and priority. This was mitigated through training materials, regular debriefings, and an approach where images were only viewed when necessary to avoid unnecessary distress.