This year, Rwanda commemorates the 25th anniversary of the genocide against the Tutsi. Americans would do well to consider the sobering similarities between the Rwandan “hate radio,” or “Radio Machete”—which helped to incite that genocide—and President Trump’s tweets. This is not to say that Trump is marching the United States toward genocide, nor that hateful speech necessarily catalyzes hateful action. But the “hate radio” in Rwanda reminds us that the words of political and cultural leaders—spread through sensational and often fake news—can normalize violence.
Many warnings about Trump’s behavior and rhetoric have compared him to Hitler. Such a comparison can lead to the impression that a leader must explicitly advocate violence in order for genocide to occur. But Rwanda teaches us that violence does not have to flow through a direct chain of command (as was the case in Nazi Germany). Mass violence can also be highly decentralized. This is what happened in Rwanda, where violence was employed by centralized, state-sanctioned, and state-trained militias, but also by informal groups—for instance, bands of rural civilians—who turned on their neighbors with machetes and other makeshift weapons.1
One reason that ordinary civilians participated in the genocide was a popular radio station of the day, Radio Télévision Libre des Milles Collines (RTLM), later nicknamed “Radio Machete.” Known at first for its pop music and witty broadcasters, RTLM slowly but surely fostered the ideology to support genocidal logic, wielding fake news, ethnic scapegoating, and the denigration of other news channels. By the time the genocide started, RTLM was broadcasting outright lies and encouragement to exterminate the “cockroaches”—that is, the Tutsi minority—all the while remarking that the circumstances were “sad. Really sad.”2
A close look at Trump’s tweets in relation to RTLM not only sounds yet another alarm about the dangers of hateful rhetoric spread via modern technology; it also points to the specific strategies used by “hate radio” and its corollaries to bypass the legal and social mores that otherwise might limit, prevent, or repudiate such speech. Once we understand those strategies, we can begin to determine how truth-based rhetoric might compete with the rhetorical power of hate.
In one striking similarity, RTLM employed the language of sadness in the way that Trump uses it. Both Trump’s tweets and RTLM’s rhetoric toggle between different meanings of the word “sad,” resulting in dangerous political effects. Having designated a particular event or phenomena as somber or tragic, using the traditional language of sadness, they then transform said sadness into outrage, directed at those who made the leader sad.
Consider, for instance, Trump’s and RTLM’s responses to massacres in their respective countries. After the October 2018 mass shooting at the Tree of Life synagogue, in Pittsburgh, Trump tweeted: “Melania and I were treated very nicely yesterday in Pittsburgh. The Office of the President was shown great respect on a very sad & solemn day. We were treated so warmly. Small protest was not seen by us, staged far away. The Fake News stories were just the opposite—Disgraceful!”
Beginning with the word “sad” to mark a tragic event, Trump pivots his tweet, so that what is upsetting is not the violence itself, but the news stories that covered anti-Trump protests. By the end of the tweet, Trump is not sad about the Jews who were massacred, nor is he condemning the violence that led to a sad day. Instead, he is outraged by those who criticize him. As one can see in a keyword search of “sad” and its cognates in the Trump Twitter archives, this is typical of how Trump wields the rhetoric of sorrow and offense.
This is also how RTLM maneuvered from expressing sadness over deaths to venting outrage at the broadcasters’ political opponents. For example: on April 1, 1994—just six days before the genocide against the Tutsi began—one of RTLM’s most popular broadcasters, Noël (Noheli) Hitimana, implicitly accused Tutsi violence of causing Hutu sorrow. In so doing, Hitimana flagrantly flipped the positions of victim and perpetrator, so as to further the Hutu Power movement’s political ends.
Both Trump’s tweets and RTLM’s rhetoric toggle between different meanings of the word “sad,” resulting in dangerous political effects.
He said, “It’s ridiculous to hear them declaring on their radio: ‘That murderer, Noël!’ Ha! Ha! If you go to Ruhengeri or Byumba and walk around to the communes, … you’ll understand who has suffered and who’s the murderer. You’ll understand how deep their sorrow is. Their sorrow! When you go on … you find sad people everywhere [hose nta ho batababaye]. Sad. Really sad [rwose wababaza]. Ask them who’s responsible for their sorrow.”3
For both Trump and RTLM, such language is used to empathize with the sorrow of their political followers—and its effect, intended or not, is to inspire outrage, even violence. In short, their words incite hatred, while wearing the unassailably sympathetic cloak of sadness.
This is just one rhetorical strategy among many used by both Trump and RTLM to transform the self-perceived identity of a powerful and populous group—American white conservatives and loyalist Rwandan Hutus, respectively—into something akin to endangered minority status. That is, to identify white conservatives and loyalist Hutus as victims deserving of sympathy and also as warriors who must rise up against those who have caused them and their leader sadness.
In Rwanda, RTLM worked to mythologize a stereotypical vision of Rwandan social life that made killing Tutsis not sad or disgraceful, but the other way around: killing Tutsis was shown to be the only way to save the majority Hutu from the impending threat of the Tutsis, who had oppressed them and their forefathers. RTLM accomplished this through repeated references to the greatness of the Hutu past and the threat the Tutsi minority posed to its continuation. This will sound familiar to those of us who have been listening to Trump, over the last few years, valorize the greatness of the America that once was and describe immigrants, Muslims, Mexicans, people of color, and the liberal elite as threats to the nation.
Of course, Rwandans did not simply listen to RTLM and run to grab their machetes. Nor are Americans likely to instigate a genocide against Trump’s targets of criticism after reading his tweets. With the benefit of hindsight, we can see that there were many other factors affecting Rwandans’ behavior, and it remains a matter of debate just how much RTLM did to spur on or exacerbate the genocide.4 But we do know that Rwandans’ interpretation of the radio content was influenced by how their friends, families, neighbors, and enemies spoke about the broadcasts: how, in other words, the radio’s messages flowed through its social network of listeners.5
Take a look at Trump’s Twitter feed: you will see a proliferating network of information and misinformation, as followers and bots—in addition to protesters—respond to Trump’s original message and to each other. It doesn’t take long to find tweets supporting white supremacy, whether from Trump himself or from followers. And while Twitter weeds out many of the more egregious responses, some Trump supporters have taken to online platforms—like 8Chan and Gab—whose looser regulations on hate speech allow them to respond more explicitly to Trump’s tweets. Following Trump’s July 14 tweet that certain Democratic congresswomen of color should “go back and help fix the totally broken and crime infested places from which they came,” one 8Chan commenter posted that they anticipated an increase in supporters who “blindly emulate the President’s example.”
Just as RTLM enjoyed deadly dissemination across communities, Trump’s Twitter feed risks spreading hateful ideology through the contributions of other users, both on Twitter and other social media platforms.
That’s a far cry from the American president advocating genocide. And the strong presence of anti-Trump voices responding to the president marks a distinct difference between American social media in our current era and Rwanda in the early 1990s, when opposition voices were granted only token slots on RTLM. But it’s nonetheless difficult to counter the hateful narratives that circulate below Trump’s tweets, in part because Trump has consistently eroded the value of truth. And it’s especially difficult when so much of what Trump and his followers do and say is considered entertaining.
In Rwanda during the genocide, RTLM was also considered entertaining. In fact, RTLM was so much more entertaining than the competing radio stations that even soldiers of the Tutsi opposition force were said to prefer it. Read through a few transcripts of the opposition’s Radio Muhabura, or the government’s Radio Rwanda, and you’ll get an idea of why those dry, monotone news channels were less popular. That RTLM broadcast primarily in Rwanda’s native language of Kinyarwanda, rather than the colonial languages of English or French employed by Radio Muhabura and Radio Rwanda, that it took a populist approach to interviewing people on the street, and that it responded quickly to breaking news only served to heighten its appeal.
Just as RTLM enjoyed deadly dissemination across communities, Trump’s Twitter feed risks spreading genocidal ideology through the contributions of listeners.
It wasn’t only Rwandan listeners who perceived RTLM as sensationally entertaining. When foreign dignitaries called on the international community to block RTLM’s airwaves, these requests were denied—not only on the grounds of freedom of speech and national sovereignty, but because RTLM wasn’t taken seriously.6 As then Canadian Ambassador to Rwanda, Lucie Edwards, later said of RTLM, “There were so many genuinely silly things being said on the station, so many obvious lies, that it was hard to take seriously. It was like relying on the National Enquirer to determine your policy in outer space.”
Talk about ironic foreshadowing. It can be hard to take Trump’s tweets seriously, when so many of them seem to belong on the pages of the National Enquirer: from his petty personal feuds to his infamous typos to his outrageous political insults. But the political implications of their entertainment value must not be underestimated. It’s not only that winning the media war can mean winning an election. It’s that when public discourse becomes entertaining due to sensationalism and lies, truth is degraded. And when truth is degraded, it’s easy to unleash one of the great tactics of dictatorships worldwide: “accusation in a mirror.”
“Accusation in a mirror” means accusing the enemy of a wrong that you yourself have committed. In the Rwandan context, one-third of RTLM’s calls to rouse the Hutu to fight and/or defend themselves were accompanied by allegations of Tutsi atrocities—allegations that mirrored the actions of the Hutus themselves. These allegations ranged from political assassinations (which the Hutu Power militants likely carried out themselves) to descriptions of Tutsi civilians burning, shooting, and hacking Hutu to death (the very methods that Hutu militants employed to kill the Tutsi). Long before RTLM began blaming the Tutsi for Hutu violence, they were using accusation in a mirror to vilify competing radio stations.
In the American context, Trump is a master of accusation in a mirror, especially when it comes to vilifying the media. He spreads fake news and then accuses “the other side” of fake news. The danger of this is not only that lies spread and citizens may believe those lies, but that truth itself ceases to be grounded in reality. As journalist Masha Gessen has written, what’s happened in Putin’s Russia and Trump’s America is that the leader seeks to become king of reality itself.7 Put that way, it is easier to understand why many Rwandans took up arms to decimate their neighbors: they had ceased to own reality.
Trump’s targets for criticism and discrimination are clear: immigrants, Muslims, people of color, liberals. Even if he does not intend or desire to provoke genocide, that does not mean that his rhetoric doesn’t sow the seeds of violence. Whether or not Trump supports the language or the actions of white supremacists, those groups perceive him as supporting them. They act accordingly. Consider Charlottesville; or an angry misfit in Florida, in a van plastered with Trump stickers, mailing bombs to liberal leaders; or the mass murderer of Muslim worshippers in New Zealand: all found inspiration and validation in Trump’s words, and in the ways that others have similarly interpreted Trump.
An argument that is made over and over again by concerned citizens in the Trump era is that our society must increase its commitment to the truth if we want to combat Trump’s style of rhetoric. But what has yet to be determined is how the truth can compete with the sensational entertainment of fake news.
We can’t draw a lesson from Rwanda here, because the opposition’s Radio Muhabura was notoriously boring. We are at a new juncture in history, and we will have to create our own model to follow. Is the answer to make facts entertaining? And if so, how? Does entertainment value necessarily cloud the truth? Would Radio Muhabura have had to stoop to RTLM’s methods to sustain an audience and change hearts and minds? Or can truth entertain without losing its truthfulness?
The public conversation about Trump (and about Rwanda, for that matter), is dominated by people trained in politics. But how we can make truth compete with the lurid appeal of National Enquirer–style rhetoric is a matter of urgent debate and action for everyone. Moving forward, we will need all hands on deck to forge means of truth-telling that can cut through the noise of hate radio in all its manifestations.
This article was commissioned by Sharon Marcus and Liz Bowen. 
- The Rwandan genocide was both remarkably centralized and remarkably decentralized. As Darryl Li remarks, “Although it was spearheaded and guided at the local level by bureaucrats, party cadres and armed elements (military, police, militia), what sets the Rwanda genocide apart from many other contemporary mass atrocities was the participation of such a sizeable and heterogeneous portion of the population.” “Echoes of Violence: Considerations on Radio and Genocide in Rwanda,” in The Media and the Rwandan Genocide, edited by Allan Thompson (Pluto Press, 2007), p. 91. The hate radio is one way that scholars have made sense of this heterogeneous, decentralized participation. Alison Des Forges writes in her authoritative account of the genocide, Leave None to Tell the Story (Human Rights Watch, 1999): “The density of the administrative and political hierarchies, characteristic of Rwanda for many years, gave genocidal leaders rapid and easy access to the population, but did not guarantee mass participation in the slaughter. … Both on the radio and through public meetings, authorities worked to make the long-decried threat of [Rwandan Patriotic Front] infiltration concrete and immediate. Throughout the country they disseminated detailed false information. … Authorities counted on such news to convince Hutu that their Tutsi neighbors were dangerous agents of the RPF who had to be eliminated.” ↩
- “Cockroach” was a racial epithet for both the Tutsi and the Tutsi army, the Rwandan Patriotic Front. See the International Criminal Tribunal for Rwanda’s Judgment and Sentence Report on the hate media indictments, “The Prosecutor v. Ferdinand Nahimana, Jean-Bosco Barayagwiza, and Hassan Ngeze,” especially pp. 133–165, for a detailed assessment of RTLM broadcasts after the genocide began. The ICTR report demonstrates that RTLM not only defined the enemy as the Tutsi, but that it conflated the Tutsi and the Rwandan Patriotic Front army (calling them “cockroaches/inyenzi”), such that the radio station justified the extermination of Tutsi civilians as a form of self-defense against the Rwandan Patriotic Front. ↩
- Quoted in English in Darryl Li’s “Echoes of Violence,” p. 102. It is evident that Hitimana is describing Hutu, rather than Tutsi, “sadness” not only because of his position as broadcaster on a Hutu Power radio station, but also because of the predominately Hutu locations that he names. Also available in the Kinyarwanda original here. The Kinyarwanda section on sadness reads: “Wajya … hose nta ho batababaye. Ntaho batababaye, nbaho batababaye rwose wababaza.” The word being used in English translation as “sad” or “sadness” or “sorrow” comes from the verb kubabara, to suffer, to hurt, to be in pain. The language of sadness recurs throughout RTLM transcripts. For example, RTLM on November 24, 1993: “Or are you sad [mubabajwe] because of people who died? That was five days ago, and you, officials of political parties, you did not say anything about that.” RTLM on March 23, 1994: “A sad thing [une chose est triste]. The international community passively supports the atrocities being committed by these criminals” (in reference to alleged Tutsi violence in Burundi). ↩
- Scott Straus is especially skeptical of causal claims linking RTLM to the genocide. See Straus, “What Is the Relationship between Hate Radio and Violence? Rethinking Rwanda’s ‘Radio Machete,’” Politics and Society, vol. 35, no. 4 (2007). Charles Mironko’s “The Effect of RTLM’s Rhetoric of Ethnic Hatred in Rural Rwanda,” in Thompson, The Media and the Rwanda Genocide, overviews the debate (see especially p. 125) and discusses Mironko’s own interviews with over a hundred perpetrators. He concludes that “while some Rwandan villagers may have listened to broadcasts, many stated that they did not. They heard the messages from others, however, and understood the ideological significance of certain songs, speeches, and the reporting of current events from others. Nevertheless, this information alone did not cause them to kill. It is, therefore, necessary to explore other reasons why these Rwandans took part in the genocide.” Like many researchers, Mironko outlines the potential impact of RTLM as one of many factors that incited people to participate in the genocide. ↩
- Darryl Li’s “Echoes of Violence” describes in detail how RTLM “performatively” created “a dynamic relationship with and among listeners” (p. 90). He reports that “broadcasts were often reincarnated elsewhere as rumour, where the possibilities for exaggeration or reinterpretation could only expand … Across the country, thousands of listeners were relaying, embellishing, and even misrepresenting RTLM’s broadcasts” (pp. 99–100). Collective listening also fed into this dynamic; many people would listen to the radio in “groups as large as 100, closely following the information relayed to plan the next day’s activities” (p. 101). As Charles Mironko puts it, essentially playing out Benedict Anderson’s theory of imagined communities, “The radio audiences were well aware that they were audiences (not individuals), that millions of others were simultaneously listening as well” (p. 129). ↩
- For a description of the cries to block the airwaves, and the international response, see Frank Chalk, “Hate Radio in Rwanda,” in The Path of a Genocide: The Rwanda Crisis from Uganda to Zaire, edited by Howard Adelman and Astri Suhrke (Transaction, 1999), pp. 101–102. ↩
- Masha Gessen, “The Putin Paradigm,” New York Review of Books, December 13, 2016. Gessen emphasizes: “Lying is the message. It’s not just that both Putin and Trump lie, it is that they lie in the same way and for the same purpose: blatantly, to assert power over truth itself.” ↩
