A Set of International Crimes without Coherence or a Proper Name: The Origins of “Atrocity Speech Law”

[Gregory Gordon is Associate Professor of Law, Associate Dean for Development and External Affairs and Director of the Research Postgraduates Programme at The Chinese University of Hong Kong Faculty of Law.  He was formerly a prosecutor with the International Criminal Tribunal for Rwanda and the U.S. Department of Justice, Office of Special Investigations.]

I have always felt that great scholarship is born of great frustration. And that’s what inspired me to write Atrocity Speech Law: Foundation, Fragmentation, Fruition (Oxford University Press 2017). Why was I so frustrated? The answer goes back to my salad days as a lawyer with the Office of the Prosecutor at the International Criminal Tribunal for Rwanda, when I was assigned to the “Media” team. We investigated, and eventually indicted, certain newspaper and radio executives/employees responsible for inflammatory rhetoric disseminated in the lead up to and execution of the Rwandan Genocide. But there were few legal precepts, and even less jurisprudence, available to guide us. What little there was emanated from Nuremberg, where rabid Jew-hating journalist Julius Streicher, Nazi Radio Division head Hans Fritzsche and Reich Press Chief Otto Dietrich had been prosecuted. So, from a legal perspective, we had to be resourceful as we constructed our ICTR media cases centering on charges of direct and public incitement to commit genocide and hate speech as the crime against humanity of persecution – only the latter having been charged against the just-mentioned Nazi propagandists (other possible speech-related charges available to us were instigation and ordering). But given that we were venturing onto what was largely legal terra nullius (especially with respect to incitement), we often had to grope in the doctrinal dark. So that was the first stage of frustration.

Still, from a law development perspective, I remained sanguine. The key “Media” defendants were ultimately convicted – Ferdinand Nahimana and Jean Bosco Barayagwiza (founders of the infamous Radio Télévision Libre des Milles Collines or RTLM, otherwise known as “Radio Machete”), RTLM announcer Georges Ruggiu, and extremist Hutu newspaper editor Hassan Ngeze. The judgments against them, along with that of Mayor Jean-Paul Akayesu for incitement to genocide (the first in history) – offered hope that a decent foundation had been laid for a law that could effectively govern the relationship between speech and atrocity. But that hope turned out to be misplaced.

Over time, for example, it became clear that there were problems with the formulation and application of the incitement crime, comprising the elements of “direct,” public,” “mens rea,” “incitement” and, possibly causation. Issues arose with respect to each of these. I will not deal with each of them here but will provide some brief examples for illustrative purposes. For instance, thanks in large part to the Akayesu judgment’s paying wishy-washy obeisance to both French- and English-language sources, treatment of the “direct” element was schizophrenically situated somewhere between Common Law and Civil Law conceptions. Unfortunately, the French word for “incitement” – inciter – was also the French word for “instigation” – one of whose elements is resultant violence. So that seemed to engender confusion with respect to causation. Incitement, as an inchoate crime, should not require causation. But Akayesu and its progeny were examining causation in the factual portions all the same and the Akayesu judges even went so far as to assert the need to prove “a possible causal link” between the relevant speech and subsequent violence in that case.

As for the “public” element, its inadequacies were exposed in the so-called “roadblock cases” at the ICTR. Even though inflammatory speech uttered at roadblocks was in a “public” place — because on public roads accessible to all citizens — the speech was held not to be “incitement” because “members of the public” were not present. But if enough persons were present at the roadblock, then “members of the public” could be considered in attendance and the speech could be considered “incitement.” This was a distinction without principle and did not seem to be justified from a policy perspective. Moreover, an advocate’s voice in closer proximity to a listener is arguably more compelling than one from a distance. In other words, private incitement can be just as lethal, if not more, than public.

There were problems with crimes against humanity(CAH)-persecution too. That crime consists of (1) knowingly uttering speech as part of a widespread or systematic attack against a civilian population; (2) resulting in a severe fundamental group-rights deprivation (motivated by group-status); and (3) considered as being of the same gravity level as the other CAH acts. Unfortunately, the ICTR and ICTY have adopted polar opposite positions regarding the issue of whether hate speech not calling for violence can serve as the actus reus for persecution as a crime against humanity. The ICTR concluded that such rhetoric blatantly deprives the target ethnic group of fundamental rights and thus could be the basis for charging persecution. But the ICTY, in the Kordić judgment, found that hate speech not directly calling for violence did not constitute persecution because it failed to rise to the same level of gravity as the other enumerated crimes against humanity acts (such as imprisonment or deportation, for example). And so doctrinal gridlock ensued.

Instigation, the prompting of another to commit an offense (resulting in commission of the offence) – with a connection between the prompting and the crime (i.e. a “contribution”) has also been plagued with problems. As we have seen, it has been consistently confused with incitement and this has exacerbated the muddled jurisprudence regarding causation in both bodies of law. Moreover, there is no consistent approach to the crime’s “contribution” requirement, leading to a series of disjointed pronouncements regarding the degree of contribution and, reading last year’s horrid Šešelj judgment, arguably importing a “but for” causation requirement into the jurisprudence. Ordering, essentially instigation in the context of a superior-subordinate relationship, has been deficient as it permits the superior to escape liability if the command is not carried out (clearly problematic when juxtaposed with incitement to genocide, where there is no superior-subordinate relationship between speaker and listener but the speech utterance itself – regardless of resultant violence — carries liability).

In addition to such individual offense problems, I was also beginning to realize these modalities did not function well together as an ensemble, thus creating significant liability loopholes. For example, in the law’s current state, liability for “incitement” — an inchoate crime — is limited to genocide. Crimes against humanity and war crimes are also horrific atrocity offenses. Why was there no incitement liability connected to those crimes? Similarly, speech uttered in support of contemporaneous mass violence — and with knowledge that the violence is occurring — is limited to the offense of persecution as a crime against humanity. Why was there no speech-specific liability for rhetoric uttered knowingly in support of ongoing acts of genocide or war crimes? The answer could not lie in generic accomplice liability, as it does not recognize the unique power of speech to provoke mass atrocity in the first place.

Upon deeper reflection, it occurred to me that this problem owed to the piecemeal development of the entire body of the law from its inception. On an ad hoc basis, according to individual exigencies at different times, this doctrinal assemblage had been cobbled together by taking a hodgepodge of legal concepts, such as inchoate or accomplice liability, and willy-nilly fastening them to different speech activities. Thus, it is only by historical chance that incitement, a form of inchoate liability, only applies to genocide and not to crimes against humanity or war crimes. The resulting gaps frustrate prevention efforts and help encourage repressive regimes to take advantage of the ambiguity and suppress legitimate speech.

So, in the full measure of time, the growing body of jurisprudence was clearly not allaying my initial sense of frustration. Perhaps, I began thinking, fellow academics and other experts were also noticing these problems and offering solutions. But a review of the literature also left me disappointed – it was as fragmented as the doctrine. Despite excellent individual pieces by talented scholars such as Susan Benesch, Carol Pauli, Wibke Timmermann, Diane Orentlicher and Richard Wilson, among others (myself included – guilty as charged!), there had been no comprehensive study of this body of law. Some articles and books had dealt with parts of it but no single volume had furnished a comprehensive analysis of the entire jurisprudential output and the relation of each of its parts to one another and to the whole (although Wibke’s monograph Incitement in International Law, published after I started writing my book, came closest). No one had yet bothered to step back, systematically consider what has been produced, and provide holistic, constructive analysis and suggestions for change.

And thus was the Atrocity Speech Law project born. After completing my research, I found the book logically divided into the three components of its subtitle. Part 1, “Foundation,” begins with a brief history of atrocity speech, focusing on the Armenian genocide, the Holocaust, and mass killing in the former Yugoslavia and Rwanda in the 1990s. It then looks at the modern treatment of hate speech in international human rights treaties (such as the International Covenant on Civil and Political Rights) and in domestic jurisdictions. This serves as a bridge to a history of atrocity speech law focusing on its origins at the Nuremberg trials. Flowing from this, the book examines the development of speech crimes as formulated in the Genocide Convention and the statutes of the ICTR, ICTY and ICC. It then analyzes the relevant decisions issued by these courts, including the seminal ICTR Akayesu, Ruggiu, and Media Case judgments as well as the ICTY’s Kordić decision. It concludes by considering the general framework and the elements of the crimes established by these decisions.

Part 2, “Fragmentation,” goes on to identify the discrepancies within that framework, its inconsistent applications and other problems the framework engenders, as discussed above. Finally, Part 3, “Fruition,” recommends how the law should be developed going forward to deal with these issues. It begins by proposing how to fix the various problems within each individual speech offense. Then it suggests a more comprehensive and elegant solution: a “Unified Liability Theory” that would replace the current patchwork of speech offenses surrounding atrocity — e.g., incitement only applying to genocide, speech-specific inchoate liability not applying to crimes against humanity or war crimes, and speech offenses related to war crimes being limited to instigation/ordering — and create four general categories of speech offenses: (1) “incitement” (an inchoate mode of liability applying to all three core crimes — genocide, crimes against humanity and war crimes — but eliminating the “public” element from the liability portion of the crime and attaching it to sentencing considerations — while removing “direct” from the title only, not the prima facie elements to help protect free speech); (2) “speech-abetting” (a type of accomplice liability for speech knowingly delivered simultaneously with commission of atrocities, and also applying to all three core crimes); (3) “instigation” (a form of commission liability applying to all speech causally related to subsequent atrocity and thus also linked to the three core crimes); and (4) ordering (criminalizing commands to commit atrocity within a superior-subordinate relationship and incorporating inchoate liability).

And all of these reforms can be operationalized through promulgation of a new treaty, “The Convention on the Classification and Criminalization of Atrocity Speech Offenses,” and/or through amendment of the Rome Statute to include Article 25bis — “Liability Related to Speech.” That new provision, whose equivalent could also be placed in domestic statutes as well as the constituent instruments of existing and/or new ad hoc international tribunals, would contain all four types of speech liability — incitement, speech abetting, instigation and ordering — connected to all three core international offenses — genocide, crimes against humanity and war crimes. The point is that speech’s unique and potent contribution to atrocity should be recognized and criminalized in its own right. It should not become lost in a set of scattered provisions, relegated as a functionally invisible adjunct to other criminal law concepts in the general “modes of responsibility” sections of statutes, charters and codes. And the set of principles it gives rise to should have a name commensurate with its elevated status. That name should capture the entire range of the doctrine and its intimate relationship with mass violence. And this book coins that name: “atrocity speech law.”