Genocide and constructions of Hutu and Tutsi in radio propaganda

Radio and the Rwandan genocide

In 1959, there was a Hutu revolution to end Tutsi monarchy in Rwanda, during which some Tutsi fled from Rwanda to Uganda. Some of those Tutsi refugees created and joined the Rwandan Patriotic Front (RPF), a political group. On 1 October 1990, members of the RPF came from Uganda to attack Rwanda – an attempt by the RPF to return to Rwanda after living in Uganda for a generation. ¹⁰ After the 1990s, opponents of the RPF referred to it as the Inkotanyi.

Following the RPF attack, Rwandan elites suggested that all Tutsi – even those inside Rwanda – were supporters of the RPF, and between mid-October 1990 and the genocide in 1994, there were five massacres of Tutsi. ¹¹ On 6 April 1994, Rwanda’s President Juvenal Habyarimana died when the plane carrying him was shot down. Although it remains unclear who was responsible for the shooting, a small group of Habyarimana’s supporters blamed the RPF and set in motion the planned extermination of Tutsi. ¹² Habyarimana’s death has generally been understood as the triggering event for the genocide.

RTLM was the main voice of extremist hate propaganda thought to have contributed to the genocidal violence. ¹³ Launched as a moderate station in April 1993, RTLM’s views had turned extremist by October 1993. RTLM was known for controlling the actions of the Interahamwe militia, directing killings of Tutsi. ¹⁴ By the time the plane carrying Habyarimana was shot down on 6 April 1994, ‘RTLM was established as the propaganda arm of the Hutu government’. ¹⁵ (Even after the genocide, it broadcast warnings to Hutu who had left Rwanda, saying they would be massacred by the Tutsi government. ¹⁶ )

RTLM was owned and controlled by supporters of Habyarimana, himself Hutu; they also tended to be members of the National Revolutionary Movement for Development (MRND) party. ¹⁷ RTLM’s station was located near the presidential palace and guarded by the government. ¹⁸ President Habyarimana was the largest shareholder. ¹⁹ Other backers of the station included Habyarimana’s wife, Agatha Kanziga, and the father-in-law of one of Habyarimana’s sons. ²⁰

Although many scholars have focused on the role of RTLM, it was not the only radio station broadcasting in Rwanda during the genocide. When survey respondents were asked which stations they listened to during or prior to the Rwandan genocide, about 60 per cent did not cite RTLM: 34 per cent named RTLM and at least one other radio station, while 6 per cent said that they listened exclusively to RTLM. ²¹ In addition to foreign radio stations, alternatives to RTLM included Radio Rwanda and Radio Muhabura.

Radio Rwanda, a state-owned station, was considered the ‘official’ government station and a loyal part of Habyarimana’s regime. It had operated since the eve of Rwanda’s independence in 1962, ²² and its broadcasts tended to be more moderate than those of RTLM. By December 1990, however, Habyarimana’s regime had lost confidence in Christophe Mfizi, the station’s director, ²³ who was replaced by Ferdinand Nahimana, ‘an extremist intellectual’, who used radio broadcasts to manipulate public opinion. ²⁴ His experiments with manipulating public opinion were used during the genocide. ²⁵ By 1993, those outside Rwanda realised that Radio Rwanda was dangerous and controlled by the president.

Radio Rwanda was important not only because it had taught Rwandans much about their country’s past, ²⁶ but also because of its broadcast range: while RTLM was limited to the capital, Kigali, Radio Rwanda reached all of Rwanda. ²⁷ Radio Rwanda became more powerful when Habyarimana provided small transistor radios to each region of Rwanda. ²⁸ Also, ‘Radio Rwanda was located directly across from, and had a direct line to, the presidential palace, and was fed by a generator at the palace in case of power outage.’ ²⁹ When RTLM started using Radio Rwanda’s airwaves, however, Radio Rwanda broadcast less and less. ³⁰

Tutsi-led rebels fighting the Rwandan government ran another radio station, Radio Muhabura. Its appeal was limited, perhaps because its ideological vision was too radical for many Rwandans. The station’s view of politics, according to Darryl Li, ignored ethnicity (or suggested it was an imperialist plot). ³¹ Overall, I have chosen to focus on Radio Rwanda and RTLM broadcasts because both stations were relatively popular during the genocide, at least compared to Radio Muhabura.

Ethnicity: the dominant explanation of the Rwandan genocide

Foreign media coverage tended to characterise the 1994 Rwandan genocide between Hutu and Tutsi as tribal conflict or fighting between ethnic groups. ³² The Christian Science Monitor described ‘tribal bloodshed’ and ‘ethnic slaughter’ between ‘Hutu and Tutsi tribes’. ³³ The Washington Post reported ‘tribal bloodletting’. ³⁴ An article in The Independent was headlined ‘Rwanda’s tribal warfare claims 37’. ³⁵ A headline in The Telegraph read, ‘Survivors of Rwanda genocide still live in fear as rival tribes prepare for first presidential vote’. ³⁶

Several scholars have critiqued such descriptions of the genocide. After all, typical characteristics of ethnicity did not divide Hutu and Tutsi even at the time of the genocide: they had built Rwanda together, had a common language (Kinyarwanda) and religious views (mainly Catholic), and recognised the same heroes. ³⁷ In fact, it was the Twa who spoke a ‘distinctive form of Kinyarwanda’. ³⁸ The Twa were originally forest dwellers, but by the time of the genocide they worked in various professions. They tended to be physically smaller, and they were generally separated from both Hutu and Tutsi.

Despite these similarities between Hutu and Tutsi, during the 1920s the Belgians, the colonial power, in the belief that Tutsi (who looked more European) were superior to Hutu and Twa, offered administrative posts only to Tutsi. Needing, therefore, to distinguish Tutsi from Hutu and Twa, the Belgians introduced identity cards during the 1930s that identified each Rwandan as Hutu, Tutsi or Twa. When the cards were introduced, the population was approximately 84 per cent Hutu, 15 per cent Tutsi, and 1 per cent Twa. Although some people changed their classification after the cards had been introduced, Hutu, Tutsi, and Twa identities became more fixed as a result. Postcolonial governments maintained the identity card system and required that citizens carry the cards with them. ³⁹ Although Hutu and Tutsi shared many characteristics associated with ethnicity (e.g. a common language, history, culture, and religion), Tutsi, Hutu and Twa became understood as separate ethnic identities. ⁴⁰

Although ethnicity became the dominant explanation of the genocide according to foreign media sources, of the eighty-six Radio Rwanda and RTLM transcripts I analysed, only one considered how the differences between Hutu and Tutsi were truly down to ethnicity. During a 12 December 1993 RTLM broadcast, an interviewee, Jean-Bosco Barayagwiza of the Coalition for the Defence of the Republic (CDR), said: ‘Wealth is no longer one of the parameters for being a Tutsi. Not at all!’ ⁴¹ The difference between the two groups instead, he suggested, was ethnic and cultural: ‘The indicator [of ethnicity] that existed in the past and which is still valid today is not only the inborn element. Education is also involved. That is to say, culture.’ ⁴² Barayagwiza concluded that culture and education were ‘indicators of one’s belonging to an ethnic group’. ⁴³ This broadcast was unique insofar as it not only mentioned ethnicity, but also attempted to indicate how one could recognise a member of the Tutsi ethnic group qua ethnic group.

This is not to say that ethnicity was not mentioned during Radio Rwanda and RTLM broadcasts. Both Radio Rwanda and RTLM broadcasts frequently described the conflict between Hutu and Tutsi in ethnic terms. The violence was described as ‘ethnic trouble’, ⁴⁴ ‘ethnical clashes’, ⁴⁵ ‘ethnic clashes’, ⁴⁶ ‘inter-ethnic hatred’, ⁴⁷ ‘ethnic unrest’, ⁴⁸ ‘inter-ethnic tension’, ⁴⁹ ‘interethnic confrontations’, ⁵⁰ ‘ethnic rifts’, ⁵¹ ‘inter-ethnic strife’, ⁵² ‘a war among the ethnic groups’, ⁵³ ‘ethnic cleansing’, ⁵⁴ ‘ethnic problem’, ⁵⁵ ‘ethnic conflicts’, ⁵⁶ and ‘ethnic war’. ⁵⁷ But the broadcasts used ‘ethnic’ and ‘ethnicity’ loosely: they noted divisions between Hutu and Tutsi, but they did not explain how these differences were ethnic (instead of, say, occupational, racial, or political). Both stations, then, relied on listeners having a common understanding of Hutu and Tutsi as ‘other’.

It should be noted that on several occasions both Radio Rwanda and RTLM warned listeners not to kill their Tutsi neighbours just on the basis of ethnicity, or at least not without proof that their neighbours were part of the RPF-Inkotanyi. ⁵⁸ The Rwandan Minister of Defence, for example, told listeners to ‘live together peacefully and not unjustly attack their neighbours’. ⁵⁹ Sometimes the broadcasts were extremely direct: ‘So, stop attacking each other for ethnic, regional or other reasons’, ⁶⁰ or ‘I repeat, make no mistake about the identity of the enemy; the enemy is not your neighbour who has nothing to do with the armed forces of the RPF’. ⁶¹

In short, although foreign media stations tended to describe the genocide as one between ethnic groups, with the exception of one transcript, both Radio Rwanda and RTLM exploited other constructions of Hutu and Tutsi far more frequently.

In response to such foreign media coverage (as well as to some scholarship) there have been attempts to offer alternative and often more nuanced descriptions of Hutu and Tutsi identities, suggesting that Hutu, Tutsi and Twa were castes or classes ⁶² or socioeconomic groups, ⁶³ rather than ethnic groups. Some have argued that the genocide was the result of social interaction and group dynamics ⁶⁴ or the immediate social context, ⁶⁵ rather than ethnic hatred. Others contend that the Rwandan genocide is an example of polarisation based on overlapping inequalities (Kuper), the result of overlapping socioeconomic and ethnic divisions (Stavenhagen), or an instance of ethnoclass conflict (Gurr). ⁶⁶ By drawing attention to the importance of other explanations – especially economic or class inequalities – these perspectives both critique and add to ethnicity-only explanations.

In addition, some scholars note that divisions other than those between Hutu and Tutsi existed in Rwanda. Uvin explains that the division between the haves and have-nots was regional and social rather than ethnic. ⁶⁷ In other words, there were not only divisions between Hutu and Tutsi but also between Hutu in the north and south, and sometimes regional particularities were more important than cultural generalisations. ⁶⁸ D. S. and C. M. Newbury argue that dominant visions of Rwanda have ‘sublimated alternative visions, which expressed regional particularity, diverse forms of identity (based on kin, class, occupation, and friendship networks), and the interaction of local agency with elite policy’. ⁶⁹ In other words, ethnicity-only explanations of the Rwandan genocide miss not only the complexity of Hutu and Tutsi identities, but also a host of other divisions that existed in Rwanda.

Loose and/or one-dimensional constructions of Hutu and Tutsi identities

Despite such critiques of ethnicity-only explanations, existing scholarship tends to suggest that radio broadcasts defined the enemy in one particular way only (e.g. as racial groups or political groups, but not both), or is conceptually loose in its characterisation of the broadcasts. For example, Kellow and Steeves say that RTLM’s agenda was one of ‘ethnic hatred’, ⁷⁰ saying that the ‘war was framed in ethnicity – Hutu versus Tutsi’ and that the media contributed to this ‘heightened ethnic hatred and encouragement of violence during the ethnic wars between 1990 and 1994’. Yet at other times they argue that RTLM was carrying out a ‘full-fledged campaign with the goal of political solidarity for Habyarimana’s party’. At one point they say that the media was a pawn of ‘political and ethnic strife’. Elsewhere, they comment that the media frequently appealed to ‘racial hatred’. Overall, there is conceptual looseness in suggesting that RTLM’s agenda was one of ethnic hatred simply because the war and later the genocide were framed in terms of Hutu versus Tutsi. The problem is not that their characterisations of the broadcasts as fuelling ethnic, political, and racist hatred are irreconcilable, but that they fail to explain why they classify the agenda as one of ethnic rather than political or racial hatred.

Though Darryl Li classifies the radio broadcasts as racist, ⁷¹ he says that RTLM broadcasts dominated public discourse about democracy and that this discourse of democracy ‘was built around a congruence between ethnicity and nation (specifically the Hutu as the Rwandan nation, with the Tutsi transformed from exotic aristocrats to parasitic outsiders)’. ⁷² Li identifies key tropes such as the democracy trope, but they remain divorced from historical constructions of Hutu and Tutsi identities. (The democracy trope, for example, comes from political constructions of Hutu and Tutsi identities and the invader trope comes from racial and political constructions of Hutu and Tutsi identities.)

Rothbart and Bartlett also classify the radio propaganda as racist. Yet subsequently they hope ‘to show how the political propaganda that precipitated the 1994 genocide was not a new phenomenon’. ⁷³ They say that such an ideology perpetuated myths of Tutsi as anything from ‘power-hungry’ to ‘intelligent’. By tracing these myths to historical constructions of Hutu and Tutsi identities, it becomes evident that stereotypes of Tutsi as power-hungry are most attributable to occupational/status and political divisions, while the stereotype of Tutsi as intelligent is most attributable to racial divisions.

Kellow and Steeves, Li, and Rothbart and Bartlett, who all examined the radio propaganda, did not set out to determine which constructions of Hutu and Tutsi identities radio stations exploited. The way in which they attempt to classify the broadcasts is useful, however, because it demonstrates the impossibility of classifying the constructions as only ethnic or racial or political. After all, Hutu and Tutsi have been defined in many ways during Rwanda’s history. Over time, ‘Hutu’ and ‘Tutsi’ have been equated not only with ethnic groups but also with occupational/status groups, racial groups and political groups. The range of ways in which Hutu and Tutsi identities had been constructed during Rwanda’s history offered organisers of the Rwandan genocide many ways to identify the target group.

In what follows, I examine the historical foundation and evolutions of Hutu and Tutsi identities and which construction(s)/identities Radio Rwanda and Radio-Télévision Libre des Milles Collines (RTLM) exploited.

Hutu and Tutsi as occupational/status groups

The division between Hutu and Tutsi is often traced to ancient occupational/status divisions. Over 2,000 years ago, the majority of those who settled in Rwanda were cultivators who raised some small stock and sometimes some cattle, while the minority were pastoralists who relied mainly on cattle to make a living. ⁷⁴ Although cultivators and pastoralists lived together, there were regional differences: more pastoralists lived in the east where it was dry and hot, and fewer lived in the north where it was cool and wet. ⁷⁵

By the eighteenth century, this occupational division was loosely associated with power. Wealth was gauged by the number of cattle one had and one’s wealth tended to be associated with one’s power. By the nineteenth or twentieth centuries, ‘Hutu’ referred to an individual’s status of being a subordinate, while ‘Tutsi’ referred to an individual’s status of having many cattle. ⁷⁶ Over time, Hutu came to refer to the mass of ordinary people as a group, and Tutsi came to refer to the elite as a whole group. ⁷⁷ In other words, an individual’s occupation became associated with wealth, which became associated with power, which was then generalised to a group of either cultivators or pastoralists.

Examining the practice of cattle clientship further illuminates the power dimension of this occupational division. A client could offer his labour in return for the use of a cow. Some early accounts suggested that cattle clientship was voluntary and drew Hutu and Tutsi closer together, but others have since argued that the relationship was divisive and appeared ‘to have been extended and reinforced by colonial power’. ⁷⁸ Tutsi who contracted clientship ties tended to have high political positions so that clientship was an example of elites’ use of political power.

By the time of the genocide, however, these occupational/status differences between Hutu and Tutsi were not so clear. Even by the mid-1950s, an investigation by Philippe Leurquin had concluded that there were no significant differences between the incomes and food consumption of Hutu and Tutsi: ‘No longer … could one assume that all Tutsi were wealthy cattle-owners, and all Hutu poor agriculturalists.’ ⁷⁹ His study was significant because it contended that ethnic identity resulted from, rather than preceded, rural class differentiation. ⁸⁰ Despite the fact that no significant income differences were found between Hutu and Tutsi, Rwandans openly expressed their class-consciousness. Particularly by the late 1980s when Rwanda’s economy was poor, people openly used the word umukungu, meaning rich person, as an insult, ⁸¹ and ‘[e]conomic insecurities … contributed to political mobilization along ethnic lines’. ⁸² In other words, even if more imagined than real, these class divisions shaped social relations between Hutu and Tutsi.

Although not all Hutu were poor peasants and not all Tutsi were rich masters by the time of the genocide, some suggest that the historical occupational/status division between Hutu and Tutsi played a role in it. Howard, for instance, describes the Rwandan genocide as a conflict ‘between (former) landowners and (former) peasants’. ⁸³ Verwimp argues that Habyarimana’s regime targeted the Tutsi minority because the Tutsi were not considered real peasants. Habyarimana praised peasants as the ‘productive forces of [Rwanda]’, and condemned non-productive citizens, saying: ‘… what we want, is to ban once and for all, the spirit of intrigue and feudal mentality … Because, we say it again, the one who refuses to work is harmful to society.’ Tutsi as (former) pastoralists were considered consumers rather than producers, ⁸⁴ and herding large numbers of cattle that grazed on pastures was considered an inefficient use of space. ⁸⁵ Habyarimana’s speeches drew on a past when Tutsi had been masters of Hutu peasants. Verwimp contends that this distinction between Hutu peasants and Tutsi masters formed the ideology of the genocide, which he summarises as follows:

Rwanda should be a peasant society (ideology) and only the Hutu are the real peasants (ethnicity) … to make Rwanda, once and for all, a real peasant society, all Tutsi had to be killed. This annihilation would allow a redistribution of wealth from Tutsi to Hutu. ⁸⁶

In other words, the ideology drew on a (somewhat real and somewhat imagined) past when Hutu were poor peasants and Tutsi were wealthy masters.

This ideology manifested itself in propaganda leading up to and during the genocide. Propaganda between 1990 and 1994 glorified peasants. ⁸⁷ For example, a popular song that asked Hutu to kill Tutsi was addressed to ‘descendants of cultivators’. ⁸⁸ Moreover, it appears that the occupational/status division that was exploited by propaganda translated into action on the ground: in 1991, during a massacre of Tutsi, victims’ cattle were killed and eaten. ⁸⁹ There were massacres in 1992 that were designed to drive pastoralist groups from the land. ⁹⁰ Ohlsson argues that scarcity of agricultural land created conditions leading to the genocide. ⁹¹ There was violence against cattle again during the 1994 genocide. ⁹² And Uvin notes that the Hutu who were killed during the genocide tended to be wealthier than Hutu who were not killed. ⁹³ In fact, Verwimp argues that participation in the genocide was correlated to one’s relation to the land and labour markets, suggesting that peasants sought to gain land from Tutsi. ⁹⁴ In other words, part of the killers’ motivation could have been to rebel against wealthy people, whom killers assumed were former pastoralists or Tutsi.

My analysis of Radio Rwanda and RTLM broadcasts reveals that both stations exploited constructions of Hutu and Tutsi as occupational/status groups, even if those divisions were more imagined than material by the time of the genocide. Referring to historical occupational divisions when Hutu were cultivators and Tutsi were pastoralists, both radio stations often suggested that Hutu were producers and Tutsi were consumers. For example, Tutsi were accused of being lazy and incapable of farming: ‘We know it, Tutsis are lazy. You do not know how to farm.’ ⁹⁵ Because it was said that Tutsi could not farm, they were often accused of stealing food produced by Hutu or living off the work of Hutu cultivators. ⁹⁶ For instance, an RTLM broadcaster asked: ‘Why are you running to eat our flour … were not you eating pleasant food? Your relatives were not bringing you milk there?’ ⁹⁷ Tutsi were accused of consuming ‘our flour’, or flour grown by Hutu. The reference to milk could also be deliberate, given that Tutsi pastoralists were known for keeping cattle.

The broadcasts suggested that as consumers (rather than producers), Tutsi were weak. RTLM told Tutsi (referred to as the RPF-Inkotanyi) not to be so ‘arrogant and haughty’ because ‘before looking down upon someone in his home you must first think and consider that he does not depend on you for his food’. ⁹⁸ A later broadcast also considered Tutsi weak because of their dependence on Hutu:

[T]he cockroaches Inkotanyi who came killing us and eating our things saying that they will take power … asked the assistance of children, white men and sorcerers … So I think that Inkotanyi will continue to die in our potatoes. ⁹⁹

This portrayed Hutu cultivators triumphing over weak Tutsi consumers or pastoralists.

By May and June 1994, the association between Tutsi and pastoralists was manifesting itself in violence. There were reports of killings involving traditional farm implements. It was reported that ‘a good number of people were systematically eh … massacred eh … with hoes and pruning knives’. ¹⁰⁰ Later there were reports of violence involving cattle: ‘… usually they come driving cattle in front of them … Then, the cows are killed and the Inyenzi continue to advance’. ¹⁰¹ Inyenzi, meaning ‘cockroaches’, referred to ‘Tutsi who invaded Rwanda in the 1960s’. ¹⁰² ‘After 1990, opponents of the RPF called its troops Inyenzi.’ ¹⁰³

Overall, Radio Rwanda and especially RTLM used historical constructions of Hutu and Tutsi as occupational/status groups. Both stations exploited the inequalities associated with those divisions. People might have acted on these constructions of Hutu as cultivators and Tutsi as pastoralists as evidenced by the violence involving farm implements and cattle. Some Tutsi were likely targeted because they were considered former (wealthy) pastoralists.

Explaining the Rwandan genocide (and radio propaganda) using only occupational/status divisions is problematic. Although wealthier Hutu were killed, this could have been because of killers’ opportunism (e.g. to loot) rather than because of the view that only Tutsi were wealthy (because of being [former] pastoralists). ¹⁰⁴ As well, peasants also participated in the genocide to preserve what they already had and to avoid sanctions for not participating. ¹⁰⁵ Moreover, the Habyarimana regime tried drawing attention away from class divisions. In a 1991 memorandum, the Habyarimana regime criticised the enemy for ‘sharpening conflict between individuals and regions’ by ‘turning public opinion from the ethnic problem to the socioeconomic problem between the rich and the poor’. ¹⁰⁶ This seems to suggest that the enemy was drawing attention away from the real issue of ethnicity to the irrelevant issue of class. Finally, to focus only on occupational/status divisions is to overlook the fact that RTLM often referred to RPF or Inyenzi (used to refer to political rebels or sometimes all Tutsi ¹⁰⁷ ) even when exploiting constructions of Hutu and Tutsi as occupational/status groups.

Hutu and Tutsi as racial groups

European race theories provided another understanding of Hutu and Tutsi identities both in Rwanda and abroad. During the nineteenth century, the Hamitic hypothesis emerged in Europe, as an offshoot of the ‘scientific’ racism of the period, suggesting that ‘African “civilization” was due to racially distinct Caucasoid invaders from the north/north-east of Africa’. ¹⁰⁸ The hypothesis was that ‘the incoming Hamites were pastoral “Europeans” – arriving wave after wave – better armed as well as quicker witted than the dark agricultural Negroes’. ¹⁰⁹ By the 1930s, Seligman (the author of Races of Africa) was publishing racial typologies to distinguish Hamites from other racial groups ‘based on physical measurements such as skin colour, hair form, stature, head shape, face shape and nose size’. ¹¹⁰ These theories supported the idea that ‘group members were taxonomically and inter-generationally similar’. ¹¹¹ During the 1950s and 1960s, writers built on Seligman’s work and applied the Hamitic hypothesis to Rwanda. Social anthropologists concluded that Tutsi pastoralists were ‘invaders who conquered and dominated the local and agricultural Hutu’. ¹¹² Writers such as Louis de Lacger (a French historian) claimed that Tutsi had Caucasian features. ¹¹³ Europeans praised Tutsi for their physical characteristics, describing them as ‘tall and slender, with aquiline noses and fine hair: “black Europeans”’. ¹¹⁴ Although once associated with the aristocracy, these features became associated with all Tutsi.

Influenced by these theories, the Belgian colonialists believed that Tutsi were superior to Hutu and that only Tutsi should be government officials. The Hamitic hypothesis and a mythical history resulting from it were taught in Rwandan schools and seminaries. History was rewritten to say that Twa were the first in Rwanda, followed by the Hutu, and then the Tutsi (the so-called Hamites) used their power to conquer Rwanda. ¹¹⁵ This version of history, Straus contends, made Rwandans think that Hutu were naturally and historically connected to the land (since they were in Rwanda before Tutsi), that Tutsi were a foreign Hamitic race that had enslaved Hutu, and that racial differences were intergenerational. ¹¹⁶

Although critiques of the Hamitic hypothesis emerged from the 1960s onwards, Straus suggests that ideas from the Hamitic hypothesis influenced the 1994 genocide. Hintjens also notes the role of national foundation myths based on race during the genocide. ¹¹⁷ The idea that Hutu had been in Rwanda before Tutsi ‘congealed into a logic of extermination’ because there was an attempt to return to a pre-Tutsi state. Tutsi were considered foreigners who did not have Rwandan origins, and Rwanda was said to belong to the Hutu ‘race’. Furthermore, the intergenerational aspect of race meant that Rwandans thought that Tutsi remained the same from generation to generation, and that the only solution to the danger of Tutsi domination (given the feudal past) was extermination. ¹¹⁸

There is some evidence that these theories resurfaced during the genocide. From the radio transcripts I reviewed, it appears that RTLM relied more heavily than Radio Rwanda on constructions of Hutu and Tutsi as racial groups. Some broadcasts, such as the following, were obvious in their exploitation of European race theories: ‘For a long time, the white man has harboured the erroneous conception that the Tutsis are the good people. In the white man’s view, the Tutsi is more handsome and more intelligent.’ ¹¹⁹ Concluding that Radio France Internationale helped the Inkotanyi, an RTLM broadcaster said: ‘The wish of these white men is that the preferred race, by God, must rule Rwanda.’ ¹²⁰ In June, an RTLM announcer said that ‘white mercenaries … demanded that the Tutsis should represent the majority in the administration … They said that since they had the good luck of being Tutsis, they had to kill the Hutu gorillas.’ ¹²¹ The broadcaster’s reference to ‘Hutu gorillas’ drew on historical racial typologies that made Hutu appear ‘gorilla-like’ by exaggerating their short stature and flatter noses. The broadcaster referred to Hutu gorillas while mentioning that foreigners demanded that Tutsi become administrators. Of course, this is consistent with a past when racial typologies and the Hamitic hypothesis led colonists to insist that the Tutsi – as the Hamites or the more European-looking, superior ‘race’ – should rule. RTLM, however, resisted historical racial constructions of Hutu and Tutsi. For example, an RTLM announcer said: ‘The Inyenzi should know that they are not more clever than anybody.’ ¹²² When these racial theories were popular, Europeans praised Tutsi for being clever, because Tutsi looked more European.

Perhaps because of Rwanda’s history of ‘double colonialism’ – defended by the Hamitic hypothesis – both Radio Rwanda and RTLM referred not only to Tutsi wanting power and control, but also obtaining that power with the help of collaborators. For instance, a Radio Rwanda announcer said:

the Inkotanyi went out, knowing they had accomplished their mission to assassinate President Habyarimana. They told themselves that they could then take the country … thinking that they were going to take the country and power and the people … assisted by their usual collaborators. ¹²³

In other words, the Inkotanyi (who are equated with Tutsi) were considered foreign invaders who wanted to conquer Rwanda. Rwanda does not belong to Tutsi and Tutsi will only obtain power with assistance from collaborators. Similarly, RTLM noted that ‘the enemy attacked us again with the help of … white mercenaries too … [and] demanded that the Tutsis should represent the majority in the administration’. ¹²⁴ Again, Tutsi were associated with collaborators who would put them in power. These broadcasts created fear by recalling the practice of ‘dual’ colonialism: the Belgians had insisted that Tutsi should run Rwanda’s administration and rule over the Hutu.

There were also more subtle ways in which racial constructions of Hutu and Tutsi were drawn upon during the broadcasts. For instance, both Radio Rwanda and RTLM encouraged Hutu to remain united. Telling listeners to protect Rwanda from Tutsi and not to loot, Clement Kayishema said: ‘We must avoid at all cost setting people against one another … you should first protect your country and then share its wealth.’ ¹²⁵ Rwanda was assumed to belong to Hutu and they needed to remain united to protect it from Tutsi invaders. Although both stations stressed the importance of Hutu unity, RTLM more clearly drew upon racial theories and a mythical history when emphasising the importance of Hutu unity. For example, RTLM asked listeners: ‘why then should these people chase us from our country?’ ¹²⁶ From the rest of the broadcast, it becomes clear that ‘these people’ were Tutsi and that ‘our country’ referred to a Rwanda that belonged to Hutu. Later during the same broadcast, an RTLM announcer said to the enemy: ‘This country is not yours, there is no land for tutsis [sic] in Rwanda. Rwanda is one, it is 26,336 square kilometers big and nothing else can be subtracted from it.’ ¹²⁷ This drew on the Hamitic hypothesis insofar as it suggested that Hutu were in Rwanda before Tutsi invaded and that Rwanda – particularly its land – therefore belonged to Hutu, not Tutsi. In short, there were clearly political motivations for encouraging unity amongst Hutu, but one can find another meaning to these broadcasts by examining the construction of Hutu and Tutsi as racial groups (and the mythical history that stems from that construction).

Both Radio Rwanda and RTLM presented unity as a way to protect against infiltration by the Inkotanyi or Tutsi. Radio Rwanda, for instance, warned listeners ‘to stop those Inyenzi from infiltrating our areas’ ¹²⁸ and said that ‘the Inkotanyi are looking for a way to infiltrate into the South of the country’. ¹²⁹ By June 1994, the station was thanking youth for ‘push[ing] back the enemy’ ¹³⁰ and encouraging listeners to ‘work together … to push back and defeat the enemy’. ¹³¹ All of these broadcasts suggested that the enemy was trying to infiltrate Rwanda from another country. Similar warnings were also broadcast by RTLM. An RTLM announcer said: ‘The truth is that we defend the town besieged, in part, by RPF and we defend it against unverifiable elements.’ ¹³² Both stations made it seem as though the RPF – equated with Tutsi – was an external force trying to infiltrate Rwanda. Constructing Tutsi as foreign invaders fuelled fears of infiltration. Although references to infiltration could stem from constructions of Hutu and Tutsi as political groups (with the RPF comprised of Tutsi refugees living in Uganda who wanted to return to Rwanda), they could also be related to the Hamitic hypothesis and the history of Rwanda that stemmed from it: the Hamitic hypothesis had fostered a mythical history in which Tutsi pastoralists were ‘invaders who conquered and dominated the local and agricultural Hutu’. ¹³³

In sum, there is evidence that both Radio Rwanda and RTLM exploited constructions of Hutu and Tutsi as racial groups. Combined with personal beliefs about race, broadcasts that told listeners how to identify Tutsi by physical features such as little noses ¹³⁴ might explain why some Hutu were accidentally killed because they looked like Tutsi. ¹³⁵ In other words, race theories such as the Hamitic hypothesis might have been influential during the genocide.

Yet the 1994 genocide cannot be understood as (only) putting into practice ideas from the Hamitic hypothesis. Racial typologies connected to the Hamitic hypothesis were based on physical, measurable characteristics. These typologies were perhaps more accurate years earlier when the Hutu and Tutsi had tended to marry within their own occupational group; as a result, Hutu tended to look like other Hutu and Tutsi like other Tutsi. ¹³⁶ By the time of the genocide, however, intermarriage was common and it was difficult to distinguish Hutu from Tutsi. In fact, RTLM sometimes warned listeners that these theories could not be trusted, asking: ‘Isn’t there any Hutu with a long nose? What stops a Hutu from being noble? You understand that it is a shame to think that.’ ¹³⁷ Finally, to focus only on racial divisions is to overlook the fact that Radio Rwanda and RTLM often referred to RPF or Inyenzi (used to refer to political rebels or sometimes all Tutsi) ¹³⁸ even when exploiting constructions of Hutu and Tutsi as racial groups. Thus, race theories cannot be understood as the sole motivation for the genocide.

Hutu and Tutsi as political groups

As occupational/status divisions between Hutu and Tutsi were no longer materially apparent in Rwanda and as racial theories began to wane, Tutsi became synonymous with a political group, the RPF. Tutsi refugees in Uganda, supported by some moderate Hutu, formed the RPF to overthrow Habyarimana and return to Rwanda. Habyarimana exploited this threat by suggesting that all Tutsi – even those inside Rwanda – were supporters of the RPF. Habyarimana’s regime did not warn people against confusing supporters of the RPF with Tutsi as an ‘ethnic’ group. ¹³⁹ Political interest was equated with ‘ethnic’ identity, with the RPF being considered Tutsi and vice versa. ¹⁴⁰ While Tutsi were equated with RPF supporters, Hutu were seen as the democratic majority. ¹⁴¹ ‘Rwandans’ were considered supporters of President Habyarimana. ¹⁴² In short, Hutu and Tutsi became equated with opposing political groups.

In the radio transcripts I analysed, both Radio Rwanda and RTLM frequently equated Hutu and Tutsi with political groups. RTLM, for instance, made it seem as though an identity card could be used to determine someone’s political affiliation. When speaking about someone who had a Tutsi mother but a Hutu identity card, an RTLM broadcaster instructed: ‘If someone has a false identity card, if he is Inkotanyi, a known accomplice of RPF don’t accept anything in exchange, he must be killed.’ ¹⁴³ Identity cards indicated if someone was Hutu or Tutsi, but the broadcaster suggested that identity cards could indicate political affiliation. This association between Hutu or Tutsi identity and political affiliation was further emphasised when the broadcaster said: ‘If you are a cockroach you must be killed, you cannot change anything, if you are Inkotanyi you cannot change anything.’ ¹⁴⁴ Although political affiliation is fluid, it is equated with Hutu and Tutsi ‘ethnic’ identities that had become more fixed with the introduction of identity cards.

More subtly, the radio stations equated Hutu and Tutsi identities with political parties by pitting the Inkotanyi (the RPF, a political group) against Hutu (an ‘ethnic’ group). For instance, an RTLM broadcaster explained that the RPF killed Hutu, ¹⁴⁵ and said: ‘Inkotanyi is my enemy, the enemy of Hutu.’ ¹⁴⁶ As well, Hutu as an ‘ethnic’ group was equated with the democratic majority: ‘Even the members of those parties, MDR, PSD, I am talking about the Hutus, the Hutu majority, they too realized that.’ ¹⁴⁷

Drawing on this construction of Hutu and Tutsi as political groups, both Radio Rwanda and RTLM exploited the threat of infiltration by groups that were equated with Tutsi, including the Inyenzi, Inkotanyi, Tutsi refugees and cockroaches. The Habyarimana regime had come to associate all Tutsi – even those in Rwanda – with the RPF, which was composed mainly of Tutsi refugees living in Uganda. Radio broadcasts warned Hutu that RPF supporters were going to infiltrate Rwanda. Tutsi refugees and RPF supporters were often mentioned together as infiltrators. Radio Rwanda reported: ‘Another disturbing fact is that several Tutsi refugees and the Inkotanyi are reported to be present in Mabayi.’ ¹⁴⁸ Although there were Tutsi living in Rwanda, the broadcasts associated Tutsi with Tutsi refugees and RPF supporters who were trying to infiltrate Rwanda from other countries.

Although less frequently, both Radio Rwanda and RTLM suggested that the enemy was also infiltrating political parties, further strengthening the link between Hutu and Tutsi identity and political loyalties. A Radio Rwanda broadcaster said: ‘the Inyenzi were hiding inside the Parti Liberal, that it was an Inyenzi party … We managed to know that the RPF was attempting to infiltrate the emerging party.’ ¹⁴⁹ Also speaking to the enemy, an RTLM broadcaster said, ‘you infiltrated the political parties, you infiltrated through the accomplices, you infiltrated through so many places’. ¹⁵⁰

To classify Radio Rwanda and RTLM broadcasts as only political propaganda, however, would be to overlook the ways in which constructions of Hutu and Tutsi as political groups interact with other constructions of Hutu and Tutsi identities. For instance, the trope of Tutsi infiltration stems not only from political constructions of Hutu and Tutsi but also racial theories (especially the Hamitic hypothesis). In sum, just as Radio Rwanda and RTLM transcripts should not be understood as exploiting only occupational/status divisions or only racial divisions, they should not be understood as exploiting only political divisions between Hutu and Tutsi.

Conclusion

Overall, while foreign media tended to report about genocide between ethnic groups, Radio Rwanda and RTLM broadcasts exploited constructions of Hutu and Tutsi as occupational/status, racial and political groups instead. Although many authors have criticised the ethnicity-only explanations of the genocide that so frequently appeared in foreign media coverage, proposing more complex and sometimes multi-faceted explanations, the way in which radio broadcasts defined the target group still tends to be characterised in one-dimensional ways. I argue instead that Radio Rwanda and RTLM broadcasts ‘othered’ the target group or enemy by simultaneously drawing upon multiple constructions of Hutu and Tutsi identities from many periods in Rwanda’s history, to also shape key tropes. Broadcasts did not speak with one voice, but instead were polyphonic, sometimes even presenting dissenting perspectives. Although the constructions of Hutu and Tutsi identity exploited by the radio stations were often murky and layered, a key theme emerged: the unequal power relationship. Tutsi – either as pastoralists, Hamites and/or RPF supporters – were presented as dominating, or wanting to dominate, Hutu.