Author Archive for
Susan Benesch

When Does Music Incite Genocide?

by Susan Benesch

[Susan Benesch is a Fellow at the Center for Applied Legal Studies at Georgetown Law Center and a former guest blogger here at Opinio Juris.]

Simon Bikindi, the Rwandan pop star whose two-year trial at the ICTR was apparently the first attempt to criminalize music in international law, was just convicted of incitement to genocide but not, after all, for his songs, even though Rwandan genocidaires sang them like anthems while hacking people to death. The ICTR did find, notably, that the songs “amplified” the genocide, but it missed an important chance to develop jurisprudence on incitement to genocide.

ICTR prosecutors indicted Bikindi, inter alia, because songs that he “composed, performed, recorded or disseminated” were used in a propaganda campaign to incite the Hutu population to kill Tutsi. The indictment named three songs in particular, and the Trial Chamber’s first challenge was to pin down the meanings of the songs, since they were written in metaphorical (and possibly ambiguous) Kinyarwanda. For instance Bikindi composed Twasezereye (“We Bade Farewell”) in 1986 for the Silver Jubilee of Rwanda’s Independence, so it referred to freedom from Belgian colonialism, and also to freedom from the Tutsi-dominated feudal system that the Belgians left behind….

Vile Crime or Inalienable Right: A Reply to Professor Gordon

by Susan Benesch

I am indebted to Professor Gregory Gordon not only for his comments now, but for his own published work on incitement to genocide, and for fruitful debates that we are continuing here. As he knows, I disagree with his contention that the ICTR jurisprudence has identified or even “gleaned” as he puts it, a four-part test for incitement to genocide. The section of the “Media” judgment from which Gregory gleans his test [Nahimana or “Media” judgment, paras 1004-1015] is simply a rambling discussion of “general principles” that “emerge from the international jurisprudence on incitement to discrimination and violence [and] serve as a useful guide to the factors to be considered” in defining incitement to genocide. The decision lists three principles: purpose, context, and causation.



Under “purpose,” the judgment conflates purpose with intent, mentions hate speech cases only (not incitement to genocide cases), and completely ignores the Genocide Convention’s specific intent requirement, which makes the question of “purpose” moot, in my view. One point on which the jurisprudence is clear is that the inciter to genocide must have the specific intent to bring about genocide. The next “general principle,” “context,” is discussed in three disjointed paragraphs. [1004, 1005, 1006]. The first one refers to historical context, i.e., previous violence, the second notes that courts may use context to ferret out a speaker’s concealed intent, and the third discusses considerations for and against restricting speech in the jurisprudence of the European Court of Human Rights. Again, there is no reference to incitement to genocide. Finally, under “causation,” the judgment finally mentions incitement to genocide, and notes that the crime carries no causation requirement.



Gregory notes that the ICTR’s jurisprudence has given examples of discourse “falling between [the] two extremes” of historical research and news reporting, on the one hand, and “explicit calls for violence” on the other. That’s true, but that shows only that the ICTR has correctly identified the two (obvious) ends of the spectrum. The necessary task, which the ICTR did not accomplish, is to distinguish between adjacent points along the spectrum: between hate speech and incitement to genocide.



It is no surprise that the Canadian Supreme Court did not apply the four-part test at all, much less “explicitly and systematically,” since it is not in the jurisprudence. I take Gregory’s point that a Canadian appeals court seems to have run off the rails when it understood Mugesera’s November 1992 speech to be about “elections, courage, and love,” but it is worth noting that there was fervent debate about how to translate the speech from the original Kinyarwanda, which made room for expert disagreements about its meaning. Also, even if three other Canadian courts reached the correct conclusion, criminalization of speech is such a delicate, dangerous, and important operation, in my view, that it must be rigorously explained.



Just two final points. I didn’t say (or didn’t intend to say) that since seventeen months elapsed between Mugesera’s speech and the Rwandan genocide, the speech cannot have been incitement to genocide. What I meant to argue is that some length of time would be too long, making the connection between speech and genocide too attenuated for criminal responsibility. For this reason, it is more logical to ask whether a speech created a reasonably possibility of genocide when the speech was made, than whether the speech influenced a genocide that took place much later.



Finally, Gregory argues that my reasonable possibility test is at odds with the inchoate nature of the crime, and that the test “opens a conceptual fissure” that might improperly admit a causation requirement. On the contrary, the reasonable possibility test allows for incitement to genocide to be identified (and prosecuted) whether genocide ensues or not.



In sum, we agree on the goal of prosecuting incitement to genocide vigorously while protecting speech as much as possible. The only question is precisely how to accomplish this, if the law – that blunt instrument – can manage such an exquisite balance at all.








Vile Crime or Inalienable Right: A Reply to Drumbl and Keitner

by Susan Benesch

Professor Mark Drumbl has put his finger on a key conundrum: that early, “entrepreneurial” speech offers the best opportunity for genocide prevention because it is the speech that primes a society for genocide, but it is also far more difficult to define than blatant incitement, uttered on the brink of genocide. The problem was beautifully captured in metaphor by a witness at the ICTR’s “Media” trial, who said that the notorious radio station RTLM had “spread petrol throughout the country little by little, so that one day it would be able to set fire to the whole country.” [Mugesera or “Media” judgment, ¶ 436]. As the witness implied, the crime that matters most is spreading the petrol, not striking the match. I wrestled with this, and constructed a definitional model that captures the later drops of petrol, but not the first ones. Incitement to genocide must be limited to speech that calls for genocide, albeit in coded language, and it must be distinguished from hate speech, which is not an international crime. Wibke Kristin Timmerman has suggested that hate speech become an international crime, but I don’t agree – at least not for prosecution by international criminal tribunals. Hate speech is criminalized quite differently in various bodies of municipal law, often in idiosyncratic response to national history. International criminal law should not attempt to supplant this, in my view. And in response to Professor Chimène Keitner’s question, I would not argue for a customary international law prohibition against incitement to genocide, since a customary norm would likely be imprecise, and subject to the usual debate over when it has crystallized. Incitement to genocide should be clearly defined in international criminal law as the extraordinary crime that it is.



I was disappointed that the appeal decision in the Media case seems to set the threshold higher than I suggest, by finding that only RTLM broadcasts after April 1994, when the genocide began, constituted incitement to genocide. The appeals panel did not make it clear, however, whether it imposed this limitation simply for lack of evidence that the pre-April broadcasts “contributed significantly to the commission of acts of genocide” or for analytical reasons.



I share Mark’s skepticism that courts and tribunals will prosecute before a genocide takes place, and I agree that other methods, such as “information intervention” like radio jamming, have a much better chance of preventing or at least limiting genocide. Chimène suggests that before my test can be used for ex ante interventions, one would have to explain how that would work. When could state sovereignty be breached by radio jamming, who would identify incitement to genocide, and so on? Chimène is quite right. I admire the proposal that Jamie Frederic Metzl outlined in his article “Rwandan Genocide and the International Law of Radio Jamming,” as well as his arguments that the end of the Cold War removed some longstanding obstacles to such relatively low-cost, high-tech humanitarian interventions, so I punt to him. Metzl found it important, notably, that “a relative consensus can be maintained regarding the international definition of incitement.”



The criminal law should not be left out of the picture even if it cannot prevent genocide, as Mark points out, as it also has important expressive and didactic goals. At their best, international tribunals and courts take part in the contemporary effort to understand how and why atrocities are committed. So tribunals should focus not only on the defendants who bear great criminal liability and responsibility, but also on the crimes that did the greatest damage. Defendants should be prosecuted for the acts they committed that contributed to bringing about genocide and other atrocities – more than for the acts that didn’t really make much difference, or that must be stretched to fit a criminological template. For example, civilian political leaders should be prosecuted for conflict entrepreneurship, rather than for civilian superior liability, which is often an attempt to adapt command responsibility to a civilian context where it inherently doesn’t apply. This is why tribunals should focus on speech as a crime – difficult though that is – as the ICTR continues to do in the trial of the pop star Simon Bikindi, and as the ICTY is now doing in the trial of the propagandist Vojislav Seselj.



Chimène also asks how my reasonable possibility test compares with other criminal laws tests regarding probable consequences, and I cannot yet answer adequately – that useful question is now on my list for future inquiry.

Vile Crime or Inalienable Right: Defining Incitement to Genocide

by Susan Benesch

Many thanks to Opinio Juris for the invitation to blog, to the Virginia Journal of International Law for publishing my article “Vile Crime or Inalienable Right: Defining Incitement to Genocide,” and to Mark Drumbl, Chimène Keitner, and Gregory Gordon for commenting.



The article argues that incitement to genocide demands keen attention because it is a precursor to genocide, and may be a prerequisite for it. Public speech is often the means by which ‘conflict entrepreneurs’ – to use Mark’s thought-provoking term – ‘exacerbate discriminatory divisions which they then commandeer.’ (Atrocity, Punishment and International Law, p. 25).



However the crime remains alarmingly ill-defined. Courts have begun to decide cases on incitement to genocide during the last decade, beginning with the ICTR’s conviction of Jean-Paul Akayesu in September 1998, but have failed to explain adequately what incitement to genocide is, or how to identify it. The confusion is so great that a Canadian federal appeals court found that a November 1992 speech by the Rwandan Hutu activist Léon Mugesera showed him to be “a fervent supporter of democracy” who spoke of “elections, courage, and love” – and then the Canadian Supreme Court concluded that the same speech constituted incitement to genocide.



The Genocide Convention, which simply describes incitement to genocide as “direct and public,” does very little to limit the crime and, especially, to distinguish it from hate speech. Without a reliable distinction, a mere racist could be convicted of a crime tantamount to genocide, and speech may be unduly and dangerously restricted. Another knotty problem that courts have so far side-stepped is the temporal one. Since it takes time to persuade a group of people to condone and/or participate in genocide, incitement to genocide must not be limited to statements made on the immediate brink of genocide, or once it has already begun. But Mugesera’s speech was given seventeen months before the Rwanda genocide started, and there must be some lapse of time that would be too long, even though courts have found that there is no causation requirement for incitement to genocide. To solve this problem, I propose that a speech be considered incitement to genocide if there is a reasonable possibility that genocide can occur when the speech was given – irrespective of whether genocide actually takes place later. Note that the “reasonable possibility” standard is not inconsistent with criminal law, since it is not to be used as a standard of proof, but rather as an aid in defining or recognizing a crime.



To evaluate when there is (or was) a reasonable possibility that a speech will lead to genocide, I propose a six-part test. The first inquiry is whether the speech was understood by its audience, at the time it was made, as a call to commit genocide. (It is not enough to examine the plain meaning of the speech, since coded language is often a feature of incitement to genocide.) The second inquiry is whether the speaker had some form of influence over the audience. This emerges from my observation that anyone can commit hate speech, but incitement to genocide requires some form of influence over the audience. Political or state authority is not necessary and may in fact be weaker than the influence of someone like Simon Bikindi, the Rwandan pop music idol who is now a defendant, charged with incitement to genocide, at the ICTR.



A third inquiry is whether the speaker used what I call hallmarks of incitement to genocide – techniques that prepare the audience psychologically for genocide. One such technique is to announce that the intended victims are plotting to massacre the audience: the Jews will annihilate you Germans if you don’t kill them first, the Tutsi are coming to wipe out you Hutu. This makes genocide seem necessary, like homicide in self-defense. The other three ‘prongs’ narrow incitement to genocide by describing a social context in which conflict entrepreneurs have already taken other damaging steps: disseminating hate speech, quashing dissent so that the poisonous speech cannot be neutralized by the marketplace of ideas, and promoting violence against the intended victims. If all six prongs of the test are satisfied, it is reasonably possible that the speech will lead to genocide.