About the author(s):
Christiane Wilke is a professor in the Department of Law and Legal Studies at Carleton University, Ottawa. Her collaborative research on documentation and investigation of civilian casualties has been supported by the Social Sciences and Humanities Research Council (Canada) and has been published in Millenium, Humanity, and the Leiden Journal of International Law, among others.
This post forms part of phase two of the Beyond Compliance Symposium: How to Prevent Harm and Need in Conflict, hosted by the Armed Groups and International Law blog. The introductory post can be found here. The symposium invites reflection on the conceptualisation of negative everyday lived experiences of armed conflict, and legal and extra-legal strategies that can effectively address both civilian harm and humanitarian need.
Civilian harm is frequently underreported. Gaps in reporting and recognising civilian harm can occur for a number of reasons. An analysis of the US military’s practices of assessing civilian casualty allegations suggests that the persistent refusal to take civilian-generated evidence of civilian harm seriously is a major contributor to the US military’s under-recognition of civilian harm. When the US military assesses claims about civilian casualties, it routinely relies on aerial surveillance videos alone. In contrast, testimony by survivors, community members and local journalists are often subordinated or disbelieved. Civilian testimony is believed only if it is “corroborated” by aerial video footage or validated by Western journalists or humanitarian workers.
The failure to take civilian voices seriously in assessments of conflict violence is a distinct form of harm: it is an “epistemic injustice” that reveals that the knowledge of those with direct experience of harm is discredited. The refusal to accord civilian testimonies credibility also leads to a drastic undercounting of civilian deaths and injuries as well as to a truncated understanding of civilian harm.
© U.S. Air Force Photo / Lt. Col. Leslie Pratt (29/11/2008). An MQ-1 Predator, armed with AGM-114 Hellfire missiles, piloted by Lt. Col. Scott Miller on a combat mission over southern Afghanistan.
This analysis draws on a broader research project that examines US military bureaucratic records to understand how civilian harm is understood and investigated (see here, here and here) or are part of research work in progress. As a result, the examples focus on the US military. Although the US military is a distinctive military and bureaucratic institution, its practices likely reflect broader patterns through which non-Western forms of knowledge are devalued and harms inflicted upon non-Western societies are under-recognised. The discussion proceeds in three parts. First, it documents the systematic undercounting of civilian casualties by the US military. Second, it examines the hierarchies of credibility that shape knowledge and under-recognise the harm that Western countries have inflicted on non-Western societies. This blog post does three things: first it documents the systematic undercounting of civilian casulaties by the US military. Second, it examines the hierarchies of credibility that shape whose accounts of violence are treated as authoritative. Finally, it considers the implications of these practices for understanding civilian harm and accountability in armed conflict.
Undercounted: Civilian Casualty Counts by the US Military
In armed conflicts that involve Western militaries, media reports often heavily rely on the accounts of the alleged perpetrators—among other sources—that issue press releases containing numbers of civilian and combatant deaths and injuries. Yet US-led military coalitions have been notorious for under-recognising and under-reporting the civilian harm they have caused. For example, the US-led Coalition fighting ISIS in Iraq and Syria has acknowledged 1,457 civilian deaths and 411 civilian injuries for a time period in which independent researchers at Airwars have identified between 8,263 and 13,359 civilian deaths and about 5,933 likely civilian injuries. (While these organisations might not always agree on who counts as a civilian, the gap in counting seems largely due to the differences in the evaluation of evidence.) This is the problem of systematic undercounting of civilian harm. In addition, the scope of the harm is often misunderstood. As researchers have shown (see also here), civilian harm ranges far beyond deaths and injuries immediately caused by airstrikes: in the aftermath of attacks, many civilians experience displacement, loss of home, work, income, and community, as well as health effects due to injuries, the loss of water and electrical infrastructure, and food insecurity. The stark difference between externally validated numbers of direct deaths and injuries and the US-led Coalition’s own data is only the tip of the iceberg when it comes to under-recognising civilian harm. The under-recognition is not only reflected in the numbers of deaths and injuries, but also about the forms of harm that remain unrecognised.
Local and International: Hierarchies of Credibility
If civilian harm is underreported, it is not for lack of effort by survivors, witnesses, and advocates. An ongoing systematic analysis of US military records (most of which are available on the New York Times website thanks to the advocacy and award-winning journalism of Azmat Khan) reveals that the US military officers in charge of assessing allegations of civilian harm operate with clear hierarchies of credibility that devalue the perspectives of ordinary civilians. About 75% of the allegations that the US military found “credible” were “self-reports”: someone within the military flagged an airstrike because the aerial surveillance video suggested evidence of civilian casualties. The incident would then be evaluated on the basis of the aerial video footage. In the US military’s hierarchy of credibility of sources, their own institutional records and video footage rank the highest.
The files also disclose other choices about whom and what to believe. This author has observed that, while many allegations of civilian harm were made on the basis of local media and social media reports, those sources were usually dismissed.
For example, in one file, the US military assessors cite a “Facebook tribute page” to a family that stated:
“This is a picture of the martyr Riad Ahmed Rzeyk, who was forced out of his house in the Railway Quarter by Daash, and so moved to the Aminah neighborhood. There his house was bombed by mistake which led to his martyrdom and that of his four children and his wife.”
The US military closed the allegation because it:
“stems from a single, low-quality source” that provided insufficient information to make an assessment of credibility.”
The community members’ knowledge of the family, their journey, and their deaths is dismissed.
In contrast, when Western journalists and humanitarian workers reported on civilian victims of airstrikes, their accounts were given a much higher level of credibility. For example, on 22 October 2016, the US military struck a target in Fasitiyah, Iraq. The records show that before the strike, the crew “did not observe any civilian pattern of life” (or evidence of civilian presence) near the target. After the strike, the crew reported “the functional destruction of the building with no indication of civilian casualties,” including no efforts to aid or rescue victims. Eight days later, a person only identified as “a U.S. civilian aid worker in Iraq” reported that eight Iraqi civilians had been killed in an airstrike about a week prior. The assessors reviewed the evidence and conceded that it was more likely than not that eight civilians had been killed in this strike. The testimony of the humanitarian worker who is identified by US citizenship but not by name was deemed sufficient to “prove” the deaths. The incident also shows that civilians who had been sheltering, sleeping, or simply living in buildings are likely to remain unseen both before and after a strike. In other cases the same pattern appears, as the US military has belatedly conceded civilian casualties reported on by Human Rights Watch as well as journalists working for the New York Times, LA Times, and Buzzfeed.
Journalists and humanitarian workers are civilians, but they often occupy a special place in the hierarchies among civilians (see for example here for humanitarian actors) due to their citizenship and the identity of their employer. A review of bureaucratic records of US civilian harm assessments since 2007 that I conducted with co-authors has revealed steep hierarchies of security, credibility, and compensation (in case of harm) among civilians. Civilians with international (especially Western) passports enjoyed the highest level of security, credibility, and compensation. Local civilians without ties to Western institutions were the least likely to be believed – including to be believed to be civilians. Civilians who were local but had connections to Western institutions occupied a middle ground. When Western media or humanitarian organisations had local employees who were harmed, they were able to use their own credibility to advocate for their employees. For example, after the US military bombed an MSF clinic in Kunduz, Afghanistan after mistaking it for a different building, MSF vigorously asserted that all employees, patients, and visitors who had been injured or killed in the attack were civilians. The US military did not argue with this proposition.
Epistemic Injustices
In philosophy and the social sciences, researchers have used concepts such as “epistemic injustice” to grapple with the fact that in social life, some people are accorded less credibility than others. Miranda Fricker defines “epistemic injustice” as a situation “in which someone is wronged specifically in her capacity as a knower.” Kristie Dotson and Jose Medina connect epistemic injustices to broader social injustices: people who are deemed less credible are often socially marginalised due to their gender, race, citizenship, sexuality, disability, or other social markers of difference.
These injustices, failures to listen and failure to believe are not universal, but rather context specific. In the cases I have analysed, the US military was the institution that was asked to listen to, believe, and give credibility to the voices of civilians on the ground. Nevertheless, the US military valued their own aerial surveillance footage higher than most visual or testimonial evidence from the ground. They tended to believe local civilians solely in two types of situations: first, civilian testimony was considered credible when it coincided with what the military thought it already knew. For example, external allegations were cross-checked against internal institutional databases of potentially “corroborating strikes.” If the details about time and place of a strike coincided with a strike the military had a record on, they were willing to use their records to “corroborate” the allegation. In the absence of such institutional records, the testimony was simply “not credible.” Second, local civilians were believed when their voices were relayed (and, implicitly “verified”) by Western journalists or humanitarian workers.
In a few cases, Iraqi civilians were able to get their story across to US military authorities. They relied on forms of evidence that not only show how much they were disbelieved, but also that the focus on counting civilian deaths and injuries obscures other forms of harm. For example, on 15 May 2017, a US strike in Mosul damaged a building adjacent to the target. Five civilians were trapped in the basement. They had a cell phone and were in contact with Iraqi authorities. Due to the heavy fighting, neither the Iraqi authorities nor the US-led Coalition attempted a rescue. As the file reports, “Some neighbors, however, heard their cries and on or about 18 May were able to rescue two (2) of the civilians but unfortunately three (3) civilians perished shortly after rescue,” most likely due to dehydration. Here, the civilians’ ability to call authorities provided evidence that they had been trapped. The three deaths are noted, but there is no discussion of possible harm to the two survivors, including possible illness or the mental trauma of being trapped in a basement and witnessing their neighbours struggle and die without being able to help. In another case, an allegation of civilian casualties was deemed “credible” after a relative provided documentary evidence of the deaths and his relationship to the house: “The residence was registered to [redacted] a family member of the alleged CIVCAS [civilian casualties]. [redacted] provided a copy of the original deed to the residence, as well as the four death certificates of the four alleged civilian casualties.”
Conclusion
When we think about what it takes for Iraqi and Syrian civilians to be heard and believed, we can begin to consider how many civilians in similar situations might have been killed, injured, or displaced without any bureaucratic record of their suffering. We will never understand the scope and depth of civilian harm if we don’t listen to and believe those who have been harmed, as the excellent work by PAX and Intimacies of Remote Warfare illustrates. In order to challenge the dehumanisation of civilians, it is important to recognise and modify the “hierarchies of credibility” that render reports of civilian harm “non-credible” unless they are corroborated by aerial video or Western voices, and further, to be guided by civilians’ lived experiences of conflict in any responses to suffering in war.