Study 1

Part II

 Part I  Part II  Part III  Part IV  Part V Part VI  Part VII

The land question

Already in 1984, 57% of rural households in Rwanda farmed less than one hectare of land and 25% had less than half a hectare, from which they had to feed an average family of five people. With the growth of the population, the inheritance laws - dividing a family's rights to use land among all the remaining sons - ensured that the size of the holdings would continue to fall and be increasingly fragmented into small plots, scattered over wider areas. Thus, in the beginning of the 1990s, the average Rwandese household farmed at least five plots of land, each with its specific characteristics of fertility, accessibility and form of tenancy.

On the diverse plots, the household must produce a constant supply of food throughout the year. This involves highly sophisticated decisions. Preferably, the crops should be of more than one type so that, for example, carbohydrate-rich crops such as potatoes can complement protein-rich crops such as beans. In addition, the farmer must also take into account the degree to which the different crops require fertile soils or can tolerate poorer soils etc. Thus, a study (quoted by Waller, 1993) showed that in order to preserve soil fertility, in 1993 farmers in one area in southern Rwanda grew 14 different crops in almost 50 rotations.

Under increasing population pressure, such a complex system is difficult to maintain, and during the 1980s more and more families could no longer afford to let their plots rest and recover through periods of fallow. The result was reduced soil fertility and short-term survival strategies, such as farming on the steepest slopes, even though the peasants knew that such practices were not sustainable. In the beginning of the 1990s, half of Rwanda's farming was done on slopes of more than 10% inclination, where rainfall often washed away both the soil and the crops. In turn, for an increasing part of the peasantry this translated into malnutrition and deeper poverty. According to a report by the Ministry of Agriculture, in 1984 the agricultural population of Rwanda numbered some 5.5 million. On average each person consumed 49 grammes of protein per day (which should be compared with the internationally recommended minimum of 59 grammes). By 1989, the agricultural population had risen to 6.5 million, but the average daily intake of protein was now down to 36 grammes per person (Waller, 1993).

The position of women

As in other African countries, the legal position of women in Rwanda is ambiguous. The 1991 constitution stipulates that all citizens are equal, while at the same time accepting the validity of traditional law in areas where there is no written code. This includes the important question of inheritance. The main problem is that the law does not consider the woman legally "competent" and only recognizes the man as the head of the household. A woman can acquire land for usufruct by settlement from her parents, or by inheritance if she has no brothers, but upon marriage it becomes the husband's property and if the marriage ends in divorce she cannot claim it. If her husband dies, the wife inherits nothing. In effect, a woman can own nothing legally, neither house, tools, livestock, nor crops. This lack of legal status causes particular problems in rural households headed by single women (22% of the total in 1984). In the modern sector of the economy, a woman's legal incapacity means that she cannot open a bank account without the permission of the husband, or - if unmarried - a male relative. Combined with her inability to own assets, this makes it almost impossible for her to obtain any loans.

In the area of government and administration, there were no women ministers until the coalition government of 1992. Nor were there any women pr‚fets or bourgemestres; 97% of all economically active women are farmers, responsible for feeding their families and running almost all aspects of the household, including farming activities such as sowing, weeding and harvesting, in addition to fetching wood and water. In the beginning of the 1990s women did 54% of all agricultural work and had on average 20% less free time than men. In spite of this, 38% of rural women had never had any contact with a government agricultural extension agent (Waller, 1993).

Of late, Rwandese women have become increasingly conscious of the injustice of their position in society. Thus, associations of women working together in the rural areas grew in strength throughout the 1980s. Within these associations women have acquired a de facto legal status through which they can gain access to land and credit. In times of political turmoil and upheavals - such as in 1994 - under the traditional, male-dominated and conservative political culture, women in general, however, have not exercised a moderating influence.

The economy

Except for under-exploited hydroelectric sites, Rwanda has very limited natural resources and the economy has almost exclusively been built around the two cash crops, coffee and tea.The manufacturing sector, however, has grown in importance since Independence, rising from virtually nothing to around 16% of the gross domestic product (GDP) in 1992. Before the 1994 upheavals - which badly hit the industrial infrastructure, management and workforce - the most important sub-sectors comprised the production of beverages and food, detergents, textiles and agricultural tools, such as hoes and machetes.

According to World Bank figures, Rwanda's GDP grew in real terms by an impressive annual average of 4.7% in the period 1970-1979, but slowed down to a 2.2% in 1980-88. In 1989, GDP fell significantly due to a sharp fall in coffee earnings. The decline continued in 1991, 1992 and 1993, and was for obvious reasons particularly devastating in 1994. GDP per capita was estimated at US$200 in 1993, compared with US$330 in 1989, i.e. a decrease of about 40% in only four years (The Economist Intelligence Unit, 1995).

Coffee is by far Rwanda's most important cash crop. The variety grown is predominantly arabica and is classed with "other milds" on the world market. It was first introduced into the country by the Belgian administration in the 1920s. The Belgians planted coffee extensively and decreed that taxes be paid in cash rather than in kind, to further force its cultivation. The colonial authorities later made coffee cultivation compulsory in many areas, a law that exists virtually unamended until today (The Economist Intelligence Unit, 1995). From its introduction, coffee production expanded until by 1986 exports - peaking at more than 42,000 tonnes - accounted for 82% of Rwanda's total export earnings. However, following the collapse of international coffee prices in 1989 and the war from October 1990, this share has constantly fallen. It stood at 51% in 1992 and should be considerably lower after the tragedy of 1994, which left coffee bushes damaged, untended and diseased all over the country. It has been estimated that it will take at least three years for them to recover.

All the coffee producers in Rwanda are smallholders, who are obliged to grow some coffee on their plots. (In Rwanda, the land belongs to the state. Individual farmers have the right only to use the land, not to own it, and the State can reclaim land for its own use, without compensation for the losses (Waller, 1993).) In 1989, there were almost 700,000 coffee farmers - or about 60% of all smallholders in the country - each growing an average of 150 bushes. During most of the 1980s, the government assured them a guaranteed price of 125 Rwanda francs (RWF) per kilo, which until 1987 meant that the price paid to the producers was less than the strong world price and, consequently, that the government made huge profits from the coffee trade. However, the world price began to fall and in 1989 Rwanda, like other small coffee-producing countries, was severly hit by the collapse of the International Coffee Agreement, which in turn led to a fall of the price on the London market to half its 1980 level. Against this background, the Rwandese government cut the price to the producers to 115 RWF per kilo and - within the 1990 Structural Adjustment Programme with the World Bank and the International Monetary Fund (IMF) - devalued the national currency by as much as 40%. An additional devaluation of 15% was made in June 1992.

While the restructuring was inevitable, from the point of view of the Rwandese peasant farmer it made coffee cultivation even less attractive than before. In one year - from 1989 to 1990 - the average farmer actually responded to the crisis by increasing production considerably, only to earn about 20% less. As the coffee bushes yielded a crop of substantially less value per hectare in local currency than, for example, bananas or beans, many Rwandese peasants were desperate to rip up their coffee plants in favour of other crops (Waller, 1993).

In addition to coffee, tea has developed into a significant foreign exchange earner, rising from 9% of export receipts in 1986 to 30% in 1992. Unlike coffee, tea is primarily grown on big estates, of which all except one are government-owned. Between them, the plantations cover 1% of Rwanda's cultivated area. In some places, such as Nkuli (Ruhengeri), tea estates have been introduced in areas previously settled by peasant farmers. The annual harvests during the period 1988-1992 fluctuated around 13,000 tonnes. However, the cultivation of tea, like that of coffee, was badly hit by the war from October 1990.

Taken as a whole, the 1980s saw a dramatic decline in Rwanda's economic fortunes and at the end of the decade the economy worsened in every key area, such as GDP growth, balance of payments, terms of trade and indebtedness (tables 1-8 in Appendix 1). The stagnating GDP has been commented upon above. The balance of payments also deteriorated from 1985, and in 1989 the value of imported goods was 3.5 times higher than the value of goods exported. This was largely a consequence of the decline in Rwanda's terms of trade, or international purchasing power, which fell by 47% between 1980 and 1987. Very few countries experienced such drastic declines over the same period. Finally, the ensuing external debt - which in 1980 stood at US$189 million - rose virtually constantly throughout the 1980s and had in 1992 jumped to US$873 million (Waller, 1993; Vassall-Adams, 1994; and The Economist Intelligence Unit, 1995).

The 1990 and 1992 structural adjustment programmes coincided with the war that started with the invasion by the Rwandese Patriotic Front in October 1990. It is, thus, difficult to assess its macro-economic impact. In addition to the hardships experienced by the coffee-growing peasants, there is, however, ample evidence that the introduction of higher fees for health and education, among other things, added to the already heavy burdens of Rwanda's poor (Vassall-Adams, 1994; Marysse, 1993 and 1994).

The war had a devastating effect on Rwanda's economy. First, it displaced hundreds of thousands of peasant farmers in northern Rwanda, with a dramatic impact on both coffee and food production. Second, it cut off the road to the Kenyan port of Mombasa, Rwanda's main overland route to the outside world. Third, it destroyed the country's fledgling tourist industry, which had become the third largest earner of foreign exchange. Finally, it prompted the government to enlarge its armed forces dramatically, thereby reducing national resources available for other purposes (Vassall-Adams, 1994).

Chapter 2

Pre-Colonial Period

This presentation, based on documentation by historians on the development of pre-colonial Rwanda, points in different directions. On the one hand, it indicates an "ethnic" diversity based partly on occupational status, partly on a patron/client relationship, but also on Hutu/Tutsi ancestry. On the other hand, it is said that until written descriptions by the first European travellers, people identified themselves according to clan rather than to ethnic affiliation, and that the description of so-called ethnic groups was laid down by those travellers. As in many other areas in Africa, in the absence of written source material, archaeological remains etc., reliable historical data from before the mid-19th century are scarce.

Most historians agree that the first inhabitants of Rwanda were hunter-gatherers and forest-dwellers, whose modern-day descendents are the Twa, today's small minority who have inhabited the country from as early as 2,000 B.C. Besides hunting, they practised pottery and basketwork. Around 1,000 A.D., a migration of farmers, Hutu, began to displace them. This migration was part of the so-called Bantu expansion, which, in the case of Rwanda, can be followed from the savannahs of present Cameroon to the Great Lakes area. They cleared the forests and cultivated the dark and rich volcanic soils (Vansina, 1962 and de Heusch, 1966).

The immigrant Bantu-speaking agriculturalists grew sorghum, kept livestock and bees, hunted and developed village industries. They wore goatskins and bark cloth, and organized themselves into lineages and clans under the leadership of heads or chiefs, respectively (d'Hertefelt, 1962). Hutu co-existed with Twa, and bartered skins and meat in exchange for salt and iron goods.

By the 15th century, many Hutu were organized into "statelets". Each of them was controlled by a dominant clan and composed of several different lineages under a ruling lineage (which over time became dynastic) headed by a mwami (chief or king), who was a land chief as well as a ritual leader in charge of rain-making (Vansina, 1962). There is evidence that some lineages had already acquired cattle at that time and that several states had emerged before the immigration of Tutsi (principally the Nyiginya clan). According to Kagame (1972), at this time seven major clans had pre-Nyiginya status.

Present-day Rwanda, seen as a geopolitical entity of many statelets, emerged, according to different historians, some time between the 11th and 15th century, largely through the pastoralist immigration and settlement of Tutsi. From about the 15th century, the number of pastoralists increased sharply in the existing states.

Tutsi seem to have been part of a larger pastoralist migration southward into the Great Lakes region (Bauman, 1948; d'Hertefelt, 1962). Whether the immigration into Rwanda was gradual or sudden has been contested. However, over time the Tutsi settlement was achieved through both conquest and peaceful assimilation (Lemarchand, 1970; d'Hertefelt, 1962; Kagame, 1972; Vansina, 1962; Ogot, 1984; and Reyntjens, 1985).

Two different phases of interaction between Hutu and Tutsi have been distinguished. The Tutsi immigration into Hutu areas is described as beginning with a gradual and peaceful infiltration. Cattle products were exchanged for agricultural products, forming the basis of social interaction. Peaceful co-existence, however, was usually followed by Tutsi conquests, resulting in the establishment of direct Tutsi military rule and administration (Lemarchand, 1966; Vansina, 1962). This phase was followed by a process directed towards the control of the factors of production, involving gradual restriction of access to land, cattle and labour (C. Newbury, 1974; d'Hertefelt, 1962; and Vidal, 1969).

Thus, over a period of 400 years, a number of independent Hutu political units were reduced to administrative entities, and Hutu transformed into what was to be described as an "ethnic" category (Lema, 1993). One, often repeated, assertion is that about 20 generations ago, one Tutsi clan, the Nyiginya, achieved political dominance in eastern Rwanda. Over several centuries, they came to form the core of a state that expanded westwards to cover most of the modern-day territory (D. Newbury, 1987). The history of the periphery differs, therefore, from that of the centre. Rwanda is, consequently, a country of strong regional variations. In particular, the northern areas or present-day Gisenyi and Ruhengeri pr‚fectures, as well as parts of the south-west, lay outside the Rwandese core state.

In the process, Tutsi were assimilated by Hutu. They took over the language spoken by Hutu (kinyarwanda) and incorporated Hutu traditions and cults. Moreover, they shared the same hills - there was no segregation of people - and they intermarried and bore the same names (Lemarchand, 1970; Rennie, 1972; Oliver, 1977 and Reyntjens, 1985).

In large part, during the pre-colonial period or before the 19th century, Tutsi, Hutu and Twa roughly corresponded to occupational categories. Cattle-herders, soldiers and administrators were mostly Tutsi, while Hutu were farmers. Twa were marginalized and often mistreated by the others. Hutu and Tutsi were less sharply distinct, and individuals could and did move between the categories as their fortunes rose or fell. Though there is no doubt that the early Nyiginya Tutsi were dominant, a range of institutions mediated social relations, notably the clan system, which spanned the entire Rwandese society. Nineteen clans encompasssed members of all three groups. Some argue that up to about the middle of the 19th century these clan identities in fact overrode the Tutsi-Hutu-Twa categorization (d'Hertefelt, 1971; D. Newbury, 1980; C. Newbury, 1978).

The first European travellers who reached central Rwanda noted a socio-economic and "ethnic" stratification between Tutsi, Hutu and Twa. Tutsi were described as distinct in terms of origin, economic activities, social status and physical appearance, although sharing the language, religion and settlement with Hutu (von G"tzen, 1895; Kandt, 1921). This description of Rwandese "ethnic groups" - partly based on indigenous mythology - was upheld and diffused by outsiders, colonial agents, ethnographers, anthropologists, historians etc. and came to represent the generalized Western view of the Rwandese people. It seems, however, that the people themselves identified each other rather according to clan affiliation. In a number of studies, David Newbury has shown that while the terms "Hutu" and "Tutsi" existed in pre-colonial times, they did not have the same significance as in the recent era, and the meaning of an "ethnic" identity varied from place to place and over time. There was no single universal definition of ethnic identity, valid for all regions at one time (D. Newbury, 1979, 1980; C. Newbury, 1988).

The amalgamation of the statelets into a united Rwanda was a process spread over several hundred years. The core Nyiginya state in eastern Rwanda slowly expanded by conquest and by giving protection, in return receiving tribute. Not until the second part of the 19th century under mwami Kigeri Rwabugiri was Rwanda united as one country. Under Rwabugiri, the mwami was the source and symbol of all authority in the politically-centralized state. Some smaller states, however, stayed autonomous until 1910-20. This was, for example, the case of the northern region near Ruhengeri, which was only incorporated into the Rwandese monarchy under German colonial rule. It took several military expeditions by the German Schutztruppe, assisted by Tutsi from central Rwanda between 1910 and 1912, before the northern Hutu - also known as Kiga - were defeated, leaving considerable bitterness towards both the Tutsi and the southern Hutu, called Banyanduga, who came with them (Dorsey, 1994; Waller, 1993). Thus, to this day, the northerners form a distinctive sub-culture, in which contacts with Tutsi have been less frequent and the awareness of a pre-Tutsi past more pronounced than in other parts of Rwanda (Lemarchand, 1970 and 1995).

The reign of Rwabugiri, or Kigeri IV, lasted from 1860 to 1895, i.e. just before the arrival of European colonialism, and marks an important watershed in the history of Rwanda. Rwabugiri broke through traditional restraints and increased the prerogatives of the throne. He is considered the last of the great reformers and is also referred to as the great warrior king. His domestic policies reflected two complementary goals, namely centralization of power and extension of the central political structures to peripheral areas of the kingdom. In foreign policy, he led a series of military campaigns against the smaller Hutu statelets in both western and eastern Rwanda, eventually incorporating them under his crown. The northern and south-western parts, however, remained largely autonomous. To undermine the hereditary power held by old Tutsi families, Rwabugiri dismissed incumbent officials and appointed men who were directly dependent on him, notably in regions that previously had been relatively independent, thereby increasing the material resources available to the monarchy (Dorsey, 1994).

What is of importance is that the state-building efforts during the reign of Rwabugiri heightened awareness of "ethnic" differences in Rwanda. As C. Newbury explains,

with the arrival of central authorities, lines of distinction were altered and sharpened, as the categories of Hutu and Tutsi assumed new hierarchical overtones associated with proximity to the central courxxxxx " proximity to power. Later, when the political arena widened and the intensity of political activity increased, these classifications became increasingly stratified and rigidified. More than simply conveying the connotation of cultural difference from Tutsi, Hutu identity came to be associated with and eventually defined by inferior status (C. Newbury, 1988).

Patron/client relationships

What appears to have kept the people together is the institution of the ubuhake - a highly personalized relationship between two individuals of unequal social status (Maquet, 1954). This patron/ client relationship involved reciprocal bonds of loyalty and exchange of goods and services. It provided a place,

a status, within an hierarchical system. The patron was mostly Tutsi, but the client could be Hutu or Tutsi of inferior social status. One person could be a client as well as a patron. Even Tutsi patrons of Hutu could be clients of yet another Tutsi. Theoretically, the only person ultimately not a client of this system was the mwami himself. Thus, most Tutsi were clients and some Hutu patrons. At the top, however, there were always Tutsi and at the bottom always Hutu and/or Twa. This institutionalized relationship was reinforced under colonial rule and lasted until it was brought to an end in the 1950s (Saucier, 1974; C. Newbury, 1988).

The ubuhake system and social order were predominant in central Rwanda, where Tutsi had their strongest influence. In the regions dominated by Hutu in the northern and south-western areas, different systems, mostly based on land-lease contracts or donation of agricultural products, were developed; the patrons were often Hutu, and in the north exclusively so (d'Hertefelt, 1962; Vansina, 1962). However, the dominance of cattle as a form of disposable wealth meant that the Tutsi cattle chiefs were able to dominate central Rwanda. Mobilizing an army required capital, which came only in the form of livestock, and Tutsi controlled the cattle. In these parts, Hutu was almost synonymous with client.

The ubuhake (and other forms of patron/client relationship such as the uburetwa) did have some important effects, viz. 1) it institutionalized the economic differences between the mainly cultivating Hutu and the cattle-breeding Tutsi; 2) it was an instrument of control, and turned Hutu into socio-economic and political clients and Tutsi into patrons; and 3) it led to a process of "ethnic" amalgamation, particularly among Hutu. The result was an "ethnic" Hutu-Tutsi dichotomy, following the socio-economic and political process engendered by Tutsi extension and occupation (Lema, 1993).

However, a number of historians question the assumption that the patron/client relationship was an important cornerstone of the traditional Hutu-Tutsi social formation as the Rwandese clans were both multi-class and multi-"ethnic". Hutu and Tutsi, they point out, shared membership in all the 19 main clans of Rwanda (Vidal, 1985; d'Hertefelt, 1971; C. Newbury 1978; D. Newbury, 1980). In particular, Hutu did not feel as one people or identify themselves as such. In this view, the system rather resulted in an economic differentiation and stratification between various occupations.

Ethnicity in pre-colonial Rwanda

As head of the late 19th century Rwandese state, the mwami owned all land and cattle. He ruled despotically, but had a political board of great chiefs and a permanent council of abiru (ritual specialists) who advised him about the divine obligations connected to his office (Vansina, 1962). According to the 500 years of mwami chronology, all bami (plural of mwami) were Tutsi (Kagame, 1957; Vansina, 1962). None of them was married to a Hutu woman, an important fact as the queen mother played a crucial role in the traditional society. Also, the great chiefs all appear to have been Tutsi (Lema, 1993), while the ritual specialists of the abiru seem to have been Hutu and based themselves on old Hutu rituals (de Lacger, 1939; Vansina, 1962).

The mwami was the supreme court/judge in traditional Rwandese society. Lower courts were the administrative court and the military court. The first dealt with land tenure disputes and was led by the land chief, the second dealt with disputes concerning cattle and was led by the army chief. The mwami and all army chiefs were Tutsi, and so almost without exception were the cattle chiefs (Vanhove, 1941). As regards the army, although it had a multi-"ethnic" composition, it was clearly stratified in the way that all higher military offices were held by Tutsi, followed by Hutu and, finally, Twa in the lowest ranks. There was, thus, no power-sharing over the army's activities: the army command, like most other institutions of the state, was mono- or uni-ethnic Tutsi (Lema, 1993; Adekanye, 1995).

By the end of the 19th century, many areas of the Rwandese kingdom had developed a complex and highly-organized administrative structure encompassing provinces, districts, hills and neighbourhoods (Vansina, 1962). The provinces were normally administered by high chiefs or army commanders, who always were Tutsi. The districts were administered by two chiefs appointed by the mwami - one cattle chief in charge of cattle taxes and one land chief responsible for agricultural levies (PagŠs, 1933; de Lacger, 1939; Kagame, 1952 and Maquet, 1961). Tutsi were normally appointed as cattle chiefs and Hutu as land chiefs (Kagame, 1957 and 1975). The districts were divided into hills, administered by chiefs responsible for handing over the levies to the two district chiefs. Rwanda did not and still does not have villages in the sense of concentrated homesteads (C. Newbury 1978). The hill was the basic administrative unit and had normally not one, but three chiefs, namely:

• the "chief of the pastures" (always a Tutsi), in charge of delimiting grazing rights;
• the "chief of the lans" (almost always a Hutu), in charge of agricultural matters and land taxes; and
• the "chief of the men" (usually a Tutsi), who was a tax collector and a kind of census agent for the abagaba, the mwami´s army recruiters (Kagame, 1975).

The three functions were often intertwined: the same person could hold all three, but on different hills. Or he could hold only one or two. (The peasants played on inter-chief rivalries, a fundamental feature of peasant survival, which was destroyed by the reforms of Governor Voisin from the late 1920s, when each hill was to be administered by only one chief) (Prunier, 1995).

It follows that the Rwandese state formation developed into a Tutsi-dominated structure, built to consolidate political power. Hutu participated only in the middle and lower levels of the administration. They were receivers of orders and norms, not norm-makers. Consequently, there was only very limited scope for "ethnic" integration in the upper echelons of the state apparatus (Lema, 1993).

Chapter 3

Colonial Period and Independence

The colonial era

When, in 1916, Belgium occupied Ruanda-Urundi as a result of the World War I East African campaign against Germany, the two kingdoms of Rwanda and Burundi had only been marginally administered from Berlin (via Dar-es-Salaam) since 1899. In 1914 there were just six German civil servants in Burundi and five in Rwanda, i.e. a total of eleven officials for a territory twice the size of Belgium. Having discovered that the existing mwami kingdoms already functioned as fully-fledged nations before the arrival of the Europeans and also, undoubtedly, because of a shortage of colonial personnel, the Germans decided from the very beginning to favour a policy of indirect rule. The occupation came about through protectorate "treaties" negotiated between the Germans and the mwami (Reyntjens, 1994). This meant that full use was to be made of the existing political system, which was much stronger and more centralized in Rwanda than in Burundi (Louis, 1963).

Belgium continued this policy: A decree of 6 April, 1917 stated that "under the authority of the Resident Commissioner the Sultans (bami) exercise their political and judicial powers to the extent that these are in accord with indigenous customs and the instructions of the Royal Commissioner" (Rumiya, 1992).

After World War I, the League of Nations mandated Belgium to administer Rwanda and in 1946 the country became a Belgian trust territory under the United Nations. During 40 years of Belgian administration, as under most colonial dispensations, we observe the disintegration, distortion or bastardization of indigenous social and political structures and their consequences. For example, while the indigenous pre-colonial patron/client relationship was flexible and contained an important element of reciprocity, the Belgian colonizers actually rigidified the system and did away with mutual obligations. By "reinforcing" a Rwandese institution, the colonizers in this way introduced forced labour and strengthened the socio-economic divisions between Tutsi and Hutu. The same abuse of other pre-colonial institutions could be quoted. Balandier has described this phenomenon in terms of the following features: the falling into abeyance of traditional political entities, the overall deterioration as a result of depoliticization, the breakdown of traditional systems of power control, the incompatibility of the two systems of power and authority and, finally, the abuse of power (Balandier, 1978). What is of interest here is the extent to which such developments affected ethnic inter-relations in Rwanda.

Among the European civil servants and missionaries operating in the Great Lakes region at the turn of the century, the so-called Hamitic thesis became generalized. According to this thesis, "everything of value in Africa had been introduced by the Hamites, supposedly a branch of the Caucasian race" (Sanders, 1969). When the well-known British explorer John Speke arrived in the Buganda kingdom (in present-day Uganda) with its sophisticated political organization, he attributed this civilization to an indigenous race of nomadic pastoralists related to the (Ethiopian) Galla "Hamites". For Europeans, the attractiveness of this hypothesis lay in the fact that it allowed for linking physical characteristics with mental capacity: the "Hamites" were supposed to be born leaders and, in principle, had the right to a history and a future almost as noble as that of their European "cousins" (Linden, 1977). In Rwanda, the "Hamites" were Tutsi: "they resemble the negro only in the colour of their skin" (Jamoulle, 1927); "before becoming black these people were tanned" (de Lacger, 1961); "his stature resembles more closely that of a white person rather than that of a negro - in fact, it would not be an exaggeration to state that he is a European who happens to have a black skin..." (Gahama, 1983). This racist thesis was expressed in innumerable ways, but, in short, Tutsi were considered to be related to Europeans and, therefore, the Europeans could easily work with them. It, thus, also served the colonial policy of divide-and-rule (Adekanye, 1995).

By the end of the 1920s, the Hamitic hypothesis was to be utilized with far-reaching consequences for ethnic relations in Rwanda. Within the framework of an administrative reform process (culminating in the Programme Voisin of 1926-1931), especially a regrouping and enlarging of chiefdoms (out of around 200 chiefs only 40 remained in the new system), it was decided to give preferential treatment to Tutsi when recruiting indigenous political authorities. It would seem that the particular position taken on the matter by Monsignor Leon-Paul Classe, the Vicar Apostolic to Rwanda, was of considerable influence. In a letter dated 21 September 1927, he wrote to Georges Mortehan, the Belgian Resident Commissioner:

If we want to be practical and look after the real interest of the country we shall find a remarkable element of progress with the Mututsi youth [...] Ask the Bahutu whether they prefer to be given orders by uncouth persons or by nobles and the answer will be clear: they will prefer the Batutsi, and quite rightly so. Born chiefs, the latter have a knack of giving orders. [...] Here lies the secret of how they managed to settle in this country and hold it in their grip (de Lacger, 1961).

Faced with what he sees as "hesitations and foot-dragging of the colonial administration regarding the traditional hegemony of the well-born Batutsi", Monsignor Classe - in 1930 - issues a stern warning:

the greatest harm the government could possibly inflict on itself and on the country would be to do away with the Mututsi caste. Such a revolution would lead the country straight into anarchy and towards a viciously anti-European communism. Far from achieving progress, this will annihilate any action taken by the government for the latter would be deprived of auxiliaries who are born capable of comprehension and obedience. [...] As a rule, we cannot possibly have chiefs wh- - d be better, more intelligent, more active, more capable of understanding the idea of progress and even more likely to be accepted by the population, than the Batutsi (Classe, 1930).

The Vicar Apostolic's message was seen as a strong plea in favour of - at least in principle - a Tutsi monopoly. His intervention put an end to the "hesitations and foot-dragging" of the administration. The Hutu chiefs and deputy-chiefs were removed and replaced by Tutsi. Furthermore, a policy favouring protection and strengthening of the Tutsi hegemony was vigorously pursued. Therefore, and given that traditionally Hutu, and even Twa, exercised certain political power, albeit at lower levels, the "Tutsification" of the 1930s resulted in a monopoly of political and administrative power in the hands of Tutsi. With the abolition of the three-fold hierarchy of the chiefs (army chief, cattle chief and land chief), this policy accentuated the ethnic divisions (Reyntjens, 1985). It was also reinforced by the introduction of identity cards in 1933. Every Rwandese was henceforth (on the basis of quite arbitrary criteria) registered as Tutsi, Hutu or Twa (Reyntjens, 1985).

Finally, the possibilities of most Hutu were further limited by the discrimination introduced in the Catholic schools, which represented the dominant educational system throughout the colonial period. Tutsi who had resisted conversion became increasingly enrolled in the Catholic mission schools. To accommodate and further encourage this process, the Church adjusted its educational policies and openly favoured Tutsi and discriminated against Hutu. With some exceptions, Hutu received only the education required for working in the mines and in industry (C. Newbury, 1988).

In summary, the monopolization of power in the hands of Tutsi constituted a crucial and undisputed factor in firmly establishing ("structuring") the ethnic cleavage. This colonial intervention caused the groups to become distinct political categories. In a certain sense, we have here an instance of ethnogenesis (Roosens, 1989), which in the case of Rwanda would inevitably lead to a reaction on the part of Hutu that they had been excluded from power. Tutsi discourse has drawn inordinate conclusions from the alleged ethnogenesis by claiming that, before the arrival of the Europeans, the people of Rwanda (and Burundi) were quite homogeneous and that, through their policy of divide-and-rule, the colonial authorities deliberately introduced ethnic cleavages. Yet the ethnic groups existed before colonialism. Colonial policies were merely grafted onto a foundation that already contained a potential for conflict (Reyntjens, 1994).

From the mid-1950s, political demands in Rwanda were formulated in ethnic terms. The opposing theses were expressed, rather stereotypically, in three main documents: on the one hand, the Bahutu Manifesto of March 24, 1957 and, on the other, two letters by the great Tutsi chiefs ("Abagaragu b'ibwami bakuru") (Nkundabagenzi, 1961). Putting the ethnic problem in a social context, the Bahutu Manifesto demanded Hutu emancipation as well as democratization. Starting from the colonial thesis that Tutsi were outsiders/foreigners and claiming that Hutu (in majority) were true Rwandese nationals, and thus the rightful rulers of Rwanda, the manifesto was a significant statement for both the social revolution from 1959 and the deepening ethnic cleavage. This important document, originally published as "Notes on the Social Aspect of the Racial Native Problem in Rwanda" and aiming to influence a United Nations Trusteeship mission to the territory, was drafted by nine Hutu intellectuals. Among the signatories was the future president, Gr‚goire Kayibanda. It attacked the whole concept of Belgian administration and maintained that the basic problem of Rwanda was a conflict between Hutu and Hamitic - i.e. foreign - Tutsi (Dorsey, 1994; Prunier, 1995). The two letters written by the conservative great chiefs (and which did not necessarily express the point of view of the whole Tutsi political elite) rejected Hutu participation "because our kings conquered the land of the Bahutu, killed their 'little' kings and thus subjugated the Bahutu; how, then, can they now pretend to be our brothers?" (Reyntjens, 1994).

When political parties were set up in the late 1950s, political structures had already been established along the ethnic cleavage: Parmehutu (Parti du mouvement de l'‚mancipation des Bahutu) and APROSOMA (Association pour la promotion sociale des masses) were essentially Hutu, whereas UNAR (Union nationale rwandaise) and RADER (Rassemblement d‚mocratique rwandais) were essentially Tutsi. At the parliamentary elections of September 1961, the cleavage was confirmed: the Hutu parties obtained about 83% of the vote, corresponding roughly to the proportion of Hutu among the population. In other words, a demographic majority came to be matched by a political majority. From 1965 onwards, following the elimination of the opposition (partly by physical, partly by political means), Rwanda was a de facto single party state; and in essence (Hutu) mono-ethnic (Reyntjens, 1985).

From the time of mwami Rwabugiri until the monarchy was abolished in 1961, the kingdom of Rwanda was a highly organized and stratified state. This was reinforced by communal reforms during the colonial period. The latest major communal reform took place in 1960, once again confirming the very organized structure of the Rwandese state. The country was divided into 10 pr‚fectures, each divided into a number of communes. These, which numbered 143 in total, formed the basis for development. The communes were each divided into 4-5 secteurs and these into "cells" (10 "cells" per secteur). Taking after the Tanzanian model, the final organizational unit was the 10-household cell comprising some 80 people. Few African countries were so well organized and also used the structures set up so extensively as Rwanda (Reyntjens, 1985).

Transition to independence

The revolution of 1959-1961, with the support of the Belgian administration (Harroy, 1984; Logiest, 1988), led to the abolition of the monarchy and to the removal of all political and administrative Tutsi structures on which, for decades, Belgium had based its policy of indirect rule. The peasants' (or Hutu) revolt was largely provoked by the intransigence of a conservative political and administrative elite, which flatly refused any democratization, demanded not only by an emerging Hutu elite, but also by a Tutsi counter-elite, far more progressive than the one in power (Reyntjens, 1994). Though, initially, the number of victims was rather small, the attempts on the part of the Tutsi-led traditional power-elite to maintain authoritarian rule led to violent clashes. The Belgians supported the revolt. The abolition of the monarchy and the rise of a Hutu elite became definitive in September 1961 when, at a referendum, 80% of the electorate voted in favour of a republic. At the same time, the results of the parliamentary elections showed a correspondingly clear victory for the Hutu-dominated parties.

 Part I  Part II  Part III  Part IV  Part V Part VI  Part VII