At an unknown moment, to an unknown place, the witnesses began to disappear. For four years, the Office of the Prosecutor of the International Criminal Court (ICC), the patchwork global justice body, gathered a robust docket of anonymous eyewitnesses, each planning to testify against Uhuru Kenyatta, Kenya’s sitting president, and his fellow architects of violence. Kenyatta’s colleagues, in the days following Mwai Kibaki’s probably-stolen 2007 election, stoked the mass killing of hundreds of people. Some were civilians, and some not; others, somewhere uncomfortably in between. These were Kenyatta’s crimes, the witnesses would say.
Except now, they will not. Its political future secure–Kenyatta’s party out-performed its opposition in a 2013 contest both fairer and less violent than its prequel–Kenya’s current national government has waged clandestine sabotage against the witnesses and their testimony. The ICC witnesses are identified, harassed, intimidated; disappeared, in that ominous passive voice. Once, a witness testimony was the only feature of their unspoken name; now, absent audience–stripped of key evidence, the Office of the Prosecutor recently requested a procedural delay–the act of witness is fully anonymous.
During the last several months, Kenyan poets have adopted the many mantles of the court’s silenced witnesses. The ICC Witness Project, as the poetry collection is titled, is a testimony never voiced. The collection tells a story, of the resilient violence of Kenya’s politics, but it never becomes one; there are neither characters to admire or despise nor arcs to follow. Despite this, as Aaron Bady writes, the poets’ stanzas are a narrative of sorts, hastily compiled from the anonymous trauma of Kenya’s violence. From Witness 130, of 144:
“After the killing blow
comes the
next one
And the one after”
Post-facto memory is the subject of Kenya’s witness-poetry; what comes before violence is of little concern. Even so, the anticipation of violence–policymakers refer to this as “early warning”–occupies a similar mode. Like the poets’ memory-fragments, the events that precede violence–“indicators”–are fractal objects. If indicators imply eventual violence, their conclusion is haphazard and undetermined. Rwanda’s present is a useful example. As Jay Ulfelder notes, following a summary of multiple expert opinions, future mass civilian violence by government-sponsored forces is likely. When precisely this event will occur is unknown.
That we can know better is a frequent assertion among mass atrocity analysts, who now use various tools–quantitative, qualitative, geospatial–to anticipate mass violence. Each early warning system differs in its research design, its algorithmic model, and its criteria for “success”–that is, whether its indicators correctly predicted new mass violence. In several cases, these systems are buttressed by new technologies, which collect, manage, and process new types of atrocity information. A recent OpenCanada post by Robert Muggah, a researcher and advocate, applauded the “breathtaking potential” of these digital systems:
“A first generation of early-warning systems designed to protect civilians from extreme violence emerged after the failure to prevent the Rwandan genocide of 1994. Many early innovations were taken up by inter-governmental, multilateral and bilateral agencies. A second generation of crisis mapping and prevention tools was spawned a decade later owing to widespread improvements in digital connectivity, cloud computing, and the proliferation of ICTs. The most prominent of these are Ushahidi, Frontline SMS, and other crisis mapping platforms. Both grassroots organizations and individuals are deploying these new tools in conflicts and humanitarian crises around the world. There is also a third generation of emerging digital systems that are providing 24/7 surveillance in the world’s hotspots through a combination of earlier methods together with Big Data analysis and drone surveillance.” [Emphasis in original.]
The use of these technologies is not monolithic, and these systems often work in tandem. Indeed, beyond their common humanitarian concern, it makes little sense to discuss the U.S. Holocaust Memorial Museum’s public early warning system, which uses large data-processing software, among other sources, to develop event forecasts, alongside the efforts of the Humanitarian UAV Network, which gathers humanitarian information through remotely-controlled aerial vehicles (“drones,” as it were).
Whatever their different approaches, each early warning system shares a single feature: a common theory of change, that better public knowledge–and foreknowledge–can reshape the future of mass atrocities. Almost every project at the small intersection of technology and mass atrocity prevention aligns with a recent wave of enthusiasm for the open-source. At its best, the project design is transparent, and its output, accessible. This makes sense: if you intend to improve public action against mass atrocities, the public should be able to access your tools.
Early warning advocates rightly champion Kenyan organizations who, in the face of escalating violence in early 2008, transmitted atrocity reports from the field. Organizations like Ushahidi recorded violence where communication technology allowed access to SMS-based platforms. In support of domestic security forces–those protecting civilians, at least–and international organizations, volunteers used digital means to record new local outbreaks. In the end, it was the digital “public”–in Kenya and abroad–that bore witness; where the appropriate technology was inaccessible, or where civilians remained too vulnerable to transmit reports, that “public” was a partial fiction. The same lyric testimonies, transcribed by the ICC Witness Project, that expanded this public in the aftermath of Kenya’s violence revealed its initial limits.
As both Sean Langberg and Danny Hirschel-Burns observe, the distribution of atrocity information to vulnerable populations is the next great challenge of mass atrocity response. Over the past three decades, witness-bearing has amply advanced both the general human condition and its guiding norms. But as civilian violence continues, it may be time to look beyond its moral virtues. The global flow of information is never a neutral arbiter; during some atrocity events, the same public knowledge that seeks to protect vulnerable populations may only deepen their insecurity. In those circumstances, witness, in the globally public sense, causes little impact at best, and significant harm, at worst. Meanwhile, atrocity information gathered by actors beyond the conflict zone–from the field, or through various surveillance technologies–rarely returns to the civilians who need it most. As presently designed, public digital witness will not restore that flow of information, nor warn civilians of emerging threats vulnerabilities. The reconstruction of either lies first in the renewed strength of civilians’ analog public–the social bonds that distribute information, and those that restrict it–and only then in the novelty of their digital technology.
The need beyond witness is a strategic question: how humanitarian organizations think about their role in civilian protection, and how they identify the purpose of the information they gather. Its operational counterpart is trickier. In this vein, humanitarian groups may learn a thing or two from the task of intelligence, albeit with wildly divergent objectives. For intelligence organizations, the security of information–in many cases, dangerous information–is paramount. Atrocity information should be no different. If global public knowledge of impending violence can make civilians vulnerable, the information returned to its potential beneficiaries should be secure.
daniel solomon 