The US bombing of a hospital run by Medecins Sans Frontieres (MSF) this month in Kunduz Afghanistan, which killed 22 people including staff and patients, has variously been described as an ‘unfortunate accident’, ‘collateral damage’, and a ‘war crime’. The labelling of this event can be seen as a function of the subject position of the different agents articulating it: the US military, the White House, or MSF. The US military has changed its story about what happened several times now and has used different terms to describe it; President Barack Obama apologised for the bombing and described it as an accident; and MSF has consistently claimed that the bombing constituted a war crime and is in direct violation of international humanitarian law and the international law of armed conflict. The true nature, extent, and reality of this event may never be known even if a full independent investigation is carried out (something MSF is demanding) and the contested nature of labelling the event will continue.
These recent events in Afghanistan are not unique. Armenia (1915), Rwanda (1994) and Srebrenica (1995) have all been labelled differently by different protagonists and agents since they have occurred. For some, these are genocide, for others they are ‘politicide,’ and for others ‘massacres’. Such labelling, or ‘naming’ and ‘framing’ is critically important for those seeking justice for past wrongs; more specifically, it is a mechanism for retributive justice that seeks to establish culpability and legal responsibility for mass violations of human rights, extrajudicial killings, or genocide. The international law of armed conflict, international humanitarian law, and the international law of human rights have sought to provide more precise definitions of these terms in an attempt to establish categories and threshold conditions for determining whether an event in which mass killings have taken place constitutes genocide. Indeed, the terminology used to describe atrocities determines the nature of the international community’s response.
The contestability and legal uncertainty of such events have motivated my own research on human rights and transitional justice. For years, I have been studying and evaluating the different ways in which empirical social scientific and statistical approaches have been applied to human rights violations committed during periods of authoritarian rule, occupation, and conflict, and how such approaches provide valid and reliable forms of evidence that can be used across different transitional justice mechanisms. There have been significant developments in statistical methods in providing estimations of the total number of people killed within such periods, the disproportionate treatment of some groups in society, and the geo-spatial patterns in gross violations of human rights and extra-judicial killings.
This work raises significant questions about the legal mandates of official truth and justice mechanisms, the ways in which past wrongs can be documented and analysed, and how the results of such systematic analyses can be used for processes of retributive justice. Across these different dimensions of truth and justice processes, the ways in which events are named and framed has an impact on the ways in which perpetrators can be prosecuted.
In the spring of 2013, there was one such event that brought together all these different dimensions, and which provided the basis for my research experiment on naming and framing genocide: the domestic trial of General Efrain Ríos Montt for genocide in Guatemala. Between the late 1960s and early 1990s, Guatemala was embroiled in a violent conflict between various rebel groups and government forces, which led to a large number of killings estimated to range between 125,606 and 138,742 people. For the period between 1982 and 1983, when Ríos Montt was in power, the conflict saw its most violent period with a significant spike in killings. Many considered him responsible for genocide against the Mayan Ixil indigenous people, and human rights groups, lawyers, and families and friends of the victims mobilised to bring him to justice.
For many years, Ríos Montt enjoyed impunity, but with significant developments in universal jurisdiction (helped along by the 1998 Pinochet extradition case in the UK), the turn of the 21st Century opened the way for prosecution. While other countries have used international bodies for prosecuting former leaders, such as the International Criminal Tribunal for Rwanda (ICTR), the International Criminal Tribunal for the Former Yugoslavia (ICTY), and the International Criminal Court, Guatemala established a domestic special court to try Ríos Montt for crimes against humanity and genocide.
The trial was presented with evidence that included forensic records, details on mass graves and exhumations, statistical analysis on the disproportionality of killings meted out against the Mayan Ixil community, and testimonies from families and friends of the victims. On 10 May 2013, Ríos Montt was found guilty of genocide and sentenced to 50 years in prison. The result was appealed and subsequent legal developments since the verdict have included an annulment and delay in retrial proceedings, as well as medical testimony concerning Ríos Montt’s fitness to continue to stand trial.
My own research saw this trial (despite its unresolved status) as an opportunity to understand the different ways in which civil society names, frames and shames events such as those that featured in Ríos Montt trial. Professor Thomas Scotto and I began our research by developing an experimental research design that would test the degree to which different frames of the events in Guatemala were related to respondents’ propensity to hold Ríos Montt and/or the military responsible for committing genocide; we also examined respondents’ differential support for national or international legal processes for retributive justice.
Using the purpose-built EssexLab, which has individual workstations and supporting software for experiments, and a ready pool of respondents, we designed an experiment with a control group and two treatment groups, each with approximately 50 respondents drawn from the pool and randomly assigned to the different groups. All three groups were asked to watch a 12-minute video on the case of Guatemala produced by the Public Broadcasting Service (PBS) in the United States in the run-up to the trial. The video presented a strong frame of genocide. Respondents in the control group were shown the video and then answered a series of questions about Ríos Montt, the army, and the appropriate legal entity to provide justice mechanisms (i.e. should it be national or international).
Respondents in the first treatment group were first shown the video and were then exposed to a strong counter frame, which suggested that the rebel groups involved at the time were threatening the state, that the country was in a civil war, and that any response to rebel threat was a matter of national defence. They were then asked the same questions as the control group about Ríos Montt, the army, and the appropriate legal entity to administer justice. Respondents in the second treatment group were also shown first the video and were then shown a strong supporting frame that contained the legal definition of genocide as set out in the 1948 Convention on the Prevention and Punishment of the Crime of Genocide. They were then asked the same questions as the control group and the first treatment group about Ríos Montt, the army, and the appropriate legal entity. All respondents were also asked a basket of demographic questions, political knowledge questions, and foreign affairs questions to account for potential confounding effects.
Our analysis of the data yielded compelling results. First, there was support for the culpability of Ríos Montt across all three groups of respondents, which was impervious to any framing effects. We found that 25.5% of the control group, 31.9% of the first treatment group, and 32.7% of the second treatment group said ‘definitely yes’ to our question on whether Ríos Montt was guilty of genocide. But the differences between these percentage response rates are not statistically different, suggesting that neither the counter-frame nor the supporting frame had an effect on changing the views of respondents.
Second, in contrast to the results for Ríos Montt, group views on the culpability of the army were significantly subject to framing effects. On our question about the army, we found that 43.1% of the control group said ‘definitely yes’, while support for culpability in first treatment group decreased to 23.4%, suggesting a direct effect of the strong counter-frame of civil war and national defence. Support for culpability in the second treatment group increased to 63.6%, suggesting a direct positive effect of the strong international legal frame. Third, we found significant framing effects for the appropriateness of domestic versus international mechanisms of transitional justice. 35.3% of control group respondents said that the decision to prosecute Ríos Montt is a domestic matter, while only 21.3% of the counter-frame group and 10.9% of the supporting frame group thought it was for Guatemalans to prosecute. In other words, both the counter-frame and the legal frame had a direct impact on respondents’ views in that the matter was found to be for the international community to decide rather than by any domestic mechanism. These different findings were upheld through more advanced analysis that included consideration of additional confounding factors.
All in all, we argue that these results show how vulnerable attitudes to genocide are to different framing effects. A simple counter-frame or supporting frame led to significant changes in respondents’ views on the army’s culpability and the appropriate judicial processes. While respondents’ views on Ríos Montt were not susceptible to such framing effects, responses to the army’s role suggest that framing and the narratives used to convey genocide can greatly influence the ways in which people make judgments about culpability as well as the appropriateness of domestic or international justice mechanisms. Our research requires much further work, repeated experiments with different respondent pools, and further analysis that accounts? for additional confounding factors; however, even these preliminary results suggest that our response to genocide is subject to significant framing effects, and that advocacy groups, human rights NGOs and policy makers must be conscious of such effects when engaging in attempts to craft and pursue strategies for retributive justice.