5—
Tribalism in the Political History of Malawi^[1]

Leroy Vail and Landeg White

So far as I am concerned, there is no Yao in this country; no Lomwe; no Sena; no Chewa; no Ngoni; no Nyakyusa; no Tonga; there are only 'Malawians'. That is all.
President Kamuzu Banda^[2]

I am a Chewa.
President Kamuzu Banda^[3]

Introduction

Between 1964, when the government of newly independent Malawi was torn apart in the so-called Cabinet Crisis, and 1975, when the Secretary General of the Malawi Congress Party, Albert Nqumayo Muwalo, and the head of the police's Special Branch, Focus Gwede, were arrested for plotting to assassinate President H. Kamuzu Banda, a traumatic rearrangement of the Malawian political order occurred. The language in which the politics of this period was discussed increasingly drew upon a store of ethnic symbols and stereotypes. The restructuring of relationships of power that occurred was seen explicitly as a campaign against the Yao-speaking peoples of the southern part of the country and all the peoples of the Northern Region. These attacks were coupled with an affirmation of the special authenticity of the culture of the country's Chewa-speaking people. These events, accompanied by repeated purges of the Party's leadership and a steadily declining real income for Malawi's workers and peasants, were responsible for the destruction of sentiments of national unity which the campaign against the Federation of Rhodesia and Nyasaland had inspired during the late 1950s and early 1960s and encouraged the fragmentation of the country along ethnically defined fault lines.^[4]

Why political discourse should have been carried out in this way requires explanation, the more so as the thrust of Malawian anti-colonialism was seen in virtually all the substantial literature it generated as having tended teleologically towards ultimate independence and national unity.^[5] From the earliest written 'tribal histories' and 'tribal associations', which were judged to have been manifestations of pride in African culture and hence as 'resistance' to colonial control, the jump to a fully nationalist perspective was assumed to be merely a matter of greater education and modernization, a process finally consummated by independence in 1964.

Yet in many ways earlier forms of parochial consciousness have proved more enduring. Tensions expressed in terms of ethnicity exist in Malawi, as they do elsewhere in the region, and they have assumed a potent reality, focusing attitudes and specifying actions. To understand why they possess such power, one must go beyond the older historiography, with its emphasis upon the nationalist dimension of resistance against colonialism and its stress on ultimate national unity. One must instead seek the varied origins of current ethnic and regional consciousness in the uneven nature of the country's colonial experience. Malawi is a particularly apt country for such an approach. Since 1921 it has been divided into Northern, Central and Southern Regions. These divisions have reflected not merely administrative convenience, but also different economic, social, and intellectual experiences dating from before the turn of this century. They thus provide a useful framework for a study of the manner in which ethnic politics has varied from area to area and from period to period.

Our hypothesis is that colonial authorities, for reasons of administrative convenience, imposed from above parameters of political debate that centred upon the assumed reality of the 'tribe' as a taxonomic unit and accepted the existence of local powers for chiefs and their advisers. A full-blown ethnic identity came into being, however, only when and where a group of African intellectuals were available to give specific cultural definition to the supposed 'tribe' and to communicate this vision through education. Because of the unevenness of the availability of education in the country, such crafting of ethnic identities and propagandizing for ethnic consciousnesses were themselves necessarily uneven.

An Ethnographic Mélange in the North

Signs of specifically ethnic thinking first appeared at the beginning of the twentieth century in Malawi's Northern Province, an isolated area far removed from the country's centres of production, trade, and industry, and characterized by great cultural fragmentation. In the far north of the province, there was a narrow band of culturally related peoples, stretching out from east to west: Ngonde, Sukwa, Ndali, Lambya, and Nyiha. Three other, substantially larger population groups were located south of this northernmost ribbon. First, there were the Tumbuka-speaking autochthones of the area, scattered widely and loosely organized under largely autonomous village headmen. Second, there were people who dwelt in a set of small chiefdoms along the shore of Lake Malawi and who spoke Lakeside Tonga, a language similar to Tumbuka in both grammar and lexicon. Finally, centred on the Kasitu river valley, there were the Ngoni under Chief Mbelwa, an amalgam of peoples of diverse cultural origins ruled by a small aristocracy.

These Ngoni derived from refugees who had fled from the wars sparked off by the creation of the Zulu empire in Natal during the 1820s. Their first leader was Zwangendaba.^[6] Around 1855, under Zwangendaba's successor, Mbelwa, the group arrived amongst the Tumbuka, whom they conquered easily. The most prominent of the Tumbuka chiefs, Chikulamayembe VIII, was slain and his people captured.^[7] The Ngoni invaders then settled, established large villages, and devastated the surrounding areas so as to create a defensive buffer zone of wilderness that only wild animals inhabited.^[8] Large numbers of defeated Lakeside Tonga and Tumbuka were incorporated in the new villages, and the Ngoni made concerted attempts to suppress the captives' own culture. The old Tumbuka religious cult of the spirit Chikang'ombe died out, and earlobes were perforated in Ngoni fashion to serve as 'a sign of baptism from the Ngoni. . . . We could not

then desert them'.^[9] Ngoni-pattrned patrilineality and an acceptance of bridewealth payable in cattle (lobola ) gradually replaced local Tumbuka matrilineal inheritance and uxorilocal brideservice practices, eroding many of the distinguishing characteristics of Tumbuka culture.

Despite these policies, the captives nonetheless retained a sense of self-awareness because of their subordination to the Ngoni ruling elite as bafo, 'slaves' or 'serfs'.^[10] In the late 1870s, moreover, several groups of Tonga and Tumbuka captives escaped successfully from Ngoni rule. One of these groups, known as the Henga, settled amongst the culturally distinct Ngonde people at the northern end of Lake Malawi, where they dwell to this day, an island of Tumbuka language and culture in a sea of Ngonde people.^[11] These rebellions eventually yielded an heroic tradition useful in the formulation of later ethnic consciousness.

In 1878, then, when the missionaries of the Scottish Presbyterian Livingstonia Mission established their first posts in northern Malawi, they encountered a Babel of linguistic confusion, the heritage of decades of movement by thousands of refugees and captives. In addition to the local languages of the region—Nyiha, Ndali, Lambya, Sukwa, Ngonde, Tumbuka, and Ngoni—the missionaries had to deal with Bisa, Bemba, Swahili, Senga, Nsenga, Sukuma, Fipa, and Nyanja, languages spoken by those whom the Ngoni had incorporated on their long anabasis through southern and eastern Africa.

For reasons of strategy and logistics, the first mission station was situated amongst the Lakeside Tonga, but, after a short time, it became apparent that Tonga, the local language of the first converts, was not feasible as a medium for further expansion of the Mission. The missionaries then decided to employ two languages: English, the language of 'high culture', and Nyanja, the language spoken on the southwest shore of Lake Malawi.^[12] Nyanja was chosen for both preaching and teaching, partly because it was a lingua franca throughout large areas of East Central Africa and partly because there was already a substantial body of religious publications in Nyanja.^[13] Nyanja was also the sole local language in the Protectorate's civil service examinations, and the missionaries hoped that it would become the lingua franca throughout the entire area of their work.^[14] The very success of the Mission, however, soon made the use of Nyanja impossible.

The Seeds of Tumbuka Ethnic Identity

The reasons behind these successes lay in the Mission's profound attractions for different Tumbuka groups.^[15] The Ngonde people at the northern end of Lake Malawi had suffered little from Ngoni incursions and had been saved by British intervention from serious disruption by late nineteenth-century slave raiders. They were thus able to maintain a coherent culture and economy, and visitors frequently described them as content to live their 'idyllic lives' uninterested in any change.^[16] The Tumbuka-speaking Henga who lived in Ngonde territory were, however, despised and resented by the Ngonde both as refugees and as former allies of Swahili slavers. Their formal religion was largely dead, and they were both ready to experiment with a new one and eager for the sort of mission education that would enable them to become successful traders, clerks and teachers.^[17] The Henga responded enthusiastically to the new educational opportunities, and in 1911 a Mission report distinguished clearly between the two local African groups:

The Henga are a keen, vigorous, progressive people, the great majority of the church members are from among them; their schools are well attended, the pupils alert, and the boys and girls in about equal numbers.

The [Ng]onde, on the other hand, are slow to move, extremely conservative, and suspicious of the new movements going on all around them.^[18]

Through their embrace of an education which had a substantial English language component, the Henga were well on their way to developing an educated petty bourgeoisie with values shaped by Victorian missionary teaching and example.

For the Tumbuka under Ngoni domination the situation was somewhat more complex. In the 1880s and 1890s, the Ngoni leadership invited missionaries to live amongst them out of economic and political considerations.^[19] At the same time, however, the Ngoni feared the corrosive impact of Christian teachings upon their military ethic. Rather than educate their own children, therefore, the Ngoni permitted the children of their Tumbuka slaves and serfs to attend the Mission's schools.^[20] The subordinate Tumbuka were thus the earliest converts to Christianity in large numbers. For the Tumbuka, an added attraction of the Mission was that, from 1894 on, its main station at Kondowe was situated between the Ngonde to the north and the Ngoni to the south, in the very heart of empty territory then gradually being reoccupied by returning Tumbuka refugees. They thus saw the Mission as politically neutral.

The Mission's educational work was remarkable. In 1893 there were ten schools with 630 pupils. By 1901 there were 55 schools which had an average attendance of 2800 pupils. By 1904 one station alone maintained some 134 schools with over 9000 students.^[21] As evidence of the Mission's educational impact, when Nyasaland's governor visited it in 1911, he was told that in 1910 the area's people had purchased 1200 lbs of writing paper and 30,000 envelopes from its shops.^[22]

As the Tumbuka embraced western education, their language gained respectability. Because they had been the first to grasp the new educational opportunities, their language could no longer be seen as the language of slaves and serfs only. Rather, it was the language of a rapidly expanding group of educated people. Moreover, because it was the language of most of the Ngoni ruling elite's wives and concubines, it was the language which even Ngoni young people learned as they grew up. In this way it displaced Ngoni and the variety of other languages spoken by the original immigrants. The area became linguistically more homogeneous and by 1900 it was clear that Ngoni was a language of the past. By 1909 Tumbuka had also displaced other languages in the Mission's Ngonde area, becoming the medium of instruction in its primary schools.^[23] It became clear to the Mission, therefore, that Tumbuka, not Nyanja, would best serve as its language for instruction and preaching, and the Bible and other religious writings were translated into it.^[24] The sole exception to this policy of using Tumbuka in Mission work was amongst the Lakeside Tonga, where, to avoid alienating its earliest converts, the Mission continued to use the local language.

A further component was required, however, to give a political thrust to the shift in Tumbuka's local standing and to release the process of myth-making that would culminate in the forging of a new, specifically ethnic, ideology. The defeat of the old Tumbuka chiefly elite by the Ngoni opened the way for the creation of a new form of leadership, and the British provided the opportunity for its establishment. Throughout Nyasaland the British administration, which gradually established its authority in the country after 1890, sought African political leaders to assist it in tax-collecting and general administration. The Tumbuka already under the authority of Ngonde or Ngoni chiefs posed no problem to the British. But for those Tumbuka refugees who had settled north of the Ngoni, in areas without chiefs, it was necessary to establish a new structure of chiefly authority.

Two groups of local Africans were eager to influence the British in this process. First, in Nkhamanga, the area of the long defunct Chikulamayembe chiefdom, there was a broad desire to resuscitate that chieftainship. According to custom the revived office should have gone to a member of the lineage of the last of the chiefs, Majuma Gondwe. Village leaders, however, felt that they needed an educated chief to voice their complaints and wishes, for, as one put it, 'For an uneducated man to speak with Europeans was an impossible dream.' When the revival of the chieftainship was raised, therefore, they did not seek it for the heir of the Majuma line, but for Chilongozi Gondwe, an offspring of a collateral lineage.^[25] This was an apt choice. Chilongozi Gondwe came from among the Henga people dwelling among the Ngonde and had attended the Livingstonia Mission's school. He had entered the colonial civil service as a policeman at Deep Bay, near the Mission headquarters at Kondowe.^[26] Although Chilongozi was neither a 'common man' nor from the area of the chieftaincy itself, the people supported him because, as he was educated, 'they felt that he would understand Europeans better'.^[27]

The second group seeking to influence the British was composed of articulate, educated Tumbuka teachers, clerics and clerks. They also supported the idea of a revived Chikulamayembe chieftainship held by Chilongozi Gondwe, one of their number. Aware through their studies at school of the potency of European nationalism and through their own personal experience of the strengths of the local Ngoni state, they recognized that political unity would also be useful to the Tumbuka in their dealings with the colonial administration. It seemed to them that the cultural symbols surrounding the revived chieftainship could be used to give ideological form to this political unity.^[28] They also lobbied the British on Chilongozi Gondwe's behalf.

In 1907 the British bowed to such pressure and recognized Chilongozi as 'Chief Chikulamayembe IX'. A coronation ceremony was put together for the occasion and an embryonic ethnic ideology based on historical symbols was created, offering Tumbuka-speakers who had formerly identified themselves primarily by their membership in lineages and clans the potential of reformulating their identity into that of belonging to a wholly new ideological construct, the 'tribe'. The new chief and his educated supporters set about building an historical image for their new Tumbuka 'tribe'. A series of articles devoted to Tumbuka history appeared in the Mission's local newspaper, Makani, and elaborate celebrations were held annually to commemorate the anniversary of Chilongozi's appointment.^[29] Saulos Nyirenda, a telegraph clerk educated at the Mission who was later to be considered the Tumbuka's 'Father of History', produced a lengthy political history in 1909. Related by marriage to the chief, he wrote with but one purpose: to glorify the history of the Chikulamayembe chieftainship and to denigrate the Ngoni for having 'spoiled our country'.^[30] Another Mission graduate, Andrew Nkhonjera, apparently with the aim of convincing the British of the supreme importance of the Chikulamayembe chieftaincy in the history of all Tumbuka-speakers, produced a similar history.^[31]

It was one thing for intellectuals to craft an ideological core of Tumbuka identity centring around the revived chieftainship. It was quite another to have the new identity widely accepted by Tumbuka-speakers and become a genuine ethnic consciousness. That it did can be explained, we suggest, by relating it to two important factors that affected the local situation after World War I. The first of these was the active intervention of the Livingstonia Mission in education and local politics, activities that encouraged people to think ethnically. The second was the profound change in the political economy of the Northern Province which prompted men to rely upon the chiefly elite to maintain order in the village while

they were far away, working as labour migrants in Southern Rhodesia or South Africa. This reliance bolstered chiefly authority and opened the way for a general acceptance of an identity and consciousness defined in terms of 'tribe'.

Ethnic Ideology and the Livingstonia Mission

Perhaps the best organized force working towards specifying the terms of the new Tumbuka identity and expanding it into a living consciousness was the Livingstonia Mission itself. Two of its missionaries had special significance in the process. The first was Thomas Cullen Young, a Scot who worked in the area between 1904 and 1931. The second was Edward Bote Manda, a Lakeside Tonga who was both a teacher and an ordained minister at the Mission headquarters. These two men, supported by the Mission's prestige and working amongst people whose respect for the printed word was immense, took up Saulos Nyirenda's earlier work and, as active culture brokers, propagated a myth based upon the alleged historical glories of the Chikulamayembe chieftainship.

Soon after his arrival, Young began studying the customs of the local people. He was in an area free from Ngonde and Ngoni political interference, and he was witness to the beginnings of the new formulation of Tumbuka history. For his historical research he depended largely on Nyirenda's history and on oral evidence gathered in Chief Chikulamayembe's area. His data was thus substantially biased towards the new chief's 'official' version of the past.^[32] His principal thesis was the same as Nyirenda's: that in the pre-Ngoni period there had existed a large Tumbuka empire, founded by the first Chikulamayembe, Mlowoka, and sustained by his successors.^[33] This empire, it was argued, included not only all speakers of the Tumbuka language, but also the Lakeside Tonga and the Ngonde peoples as well as some Chewa-speakers, extending from the Dwangwa river in the south to the Songwe river in the north, from the Luangwa valley in the west to the Lakeshore in the east, an area of some 20,000 square miles. The actual historical reality was, however, quite different. The original Chikulamayembe chieftainship had been territoriall' small, and there had been no such thing before the coming of the Ngoni as a unified empire, state, or 'tribe' encompassing the Tumbuka.^[34]

This new version of a Golden Age of the Chikulamayembes provided a heady vision of the past to a people who had been scattered and oppressed by the Ngoni conquerors and who were discontented with the colonial reality. Young's books were widely available through the Mission's shops and, as English was central in the local school curriculum, they were widely read by the rising group of Tumbuka intellectuals, helping to shape their historical consciousness. Even more importantly, his history was utilized as the basis for pamphlets about local history that were published in Tumbuka for use in the area's many primary schools. In this way, then, the heroic accounts of Tumbuka history written by Saulos Nyirenda and Cullen Young passed into popular consciousness through the Mission's educational work.^[35]

Edward Manda was at work at the same time. Manda was not of Tumbuka origin, having been born a Lakeside Tonga in an area under Ngoni hegemony, his father a captive of the Ngoni. He began his studies at the Mission in 1885. In 1905 he became a teacher, and in 1918 was ordained to the ministry, afterwards remaining at the Mission headquarters.^[36] In theory, Manda might have become a kind of Saulos Nyirenda for the Lakeside Tonga, the culture broker for a distinctly Tonga ethnic consciousness. The parallels were striking. Like the Tumbuka, the Tonga were divided into a host of small chieftaincies. Like the Tumbuka, they had

known defeat and humiliation at the hands of the Ngoni. As with the Tumbuka, some of them had rebelled and migrated to form autonomous communities, and, like the Tumbuka, they had in Chief Mankhambira, who defeated an Ngoni attack in 1880, a figure around whom myths of past greatness could have accumulated. Finally, again like the Tumbuka, they had been among the Mission's earliest converts and possessed a substantial well-educated petty bourgeoisie who could serve as culture brokers.

Yet no comparable version of Tonga ethnic consciousness was formulated. Unlike the Tumbuka refugees for whom a chief had to be found to deal with the new British administration, the Tonga did not lack political institutions, and other Tonga chiefs, eager to retain their authority, were unwilling to submit to the claims of Mankhambira's successor to have jurisdiction over all Tonga-speakers. Furthermore, Lakeside Tonga was a language so similar to Tumbuka that it could not readily be used as a distinctive cultural symbol for an ideology of Tonga ethnic identity. Young Tonga intellectuals banded together in 1919 to form the West Nyasa Native Association, as their educated counterparts had done elsewhere in the Northern Province.^[37] The Association's members were hostile to colonial injustice and oppression, and they, like their counterparts in other areas of the Northern Province, co-opted local chiefs as members to gain support from ordinary villagers. Of the three northern associations, however, the West Nyasa Native Association was the most detached from local ethnic issues, the most affected by intellectual influences from South Africa and the United States, and the most firmly committed to a generalized denunciation of colonialism in an idiom of Justice and Civilization.

Edward Manda held similar views regarding colonial injustice. While living amongst the Tumbuka he had joined the first African political association in Nyasaland. This was the North Nyasa Native Association, formed in 1912 by Tumbuka men educated at the Livingstonia Mission with the encouragement of Robert Laws, the Mission's head.^[38] By 1925 Manda had become the Association's chairman. He had also become exceedingly unpopular with the British because of his unending protests. Although a firm believer in the Victorian virtues of Improvement and Uplift, he also felt that British cultural imperialism required the development of a rival 'traditional' mythology that would at the same time incorporate the values of the educated African petty bourgeoisie.^[39] For him, the Chikulamayembe chieftainship was an obvious point of departure and, with all the ardour of a convert, he set about to bolster its status and power.

In this work, he enjoyed two real advantages. First, he was geographically well situated. As one official noted, the chiefs court was far from district headquarters at Karonga and

there is a tendency for things to fall into the hands of the Mission natives, particularly the Rev. Edward Manda at Livingstonia, who is of necessity a liaison between Chikulamayembe and the D.C. as he is in telegraphic and postal communication with Karonga.^[40]

Second, his involvement occurred at an opportune time for the strengthening of specifically ethnic institutions, for his work intersected with structural changes in the local economy and reinforced the local role of chiefs.

Chiefly Powers and Social Control

From the turn of the century onwards, Nyasaland and Northern Rhodesia, as the new British colonies came to be known, were linked to the emergent capitalist

economy of southern Africa and were especially oriented towards the gold mines of the Witwatersrand.^[41] The transformation of the rural areas of Central Africa into satellites of South Africa was gradual, but it was real, eradicating distances and dissolving territorial boundaries. The solvent in this process was money. The British administration imposed an annual hut tax both to finance its operations and to press Africans into wage labour. The missionaries who operated the local school system required the payment of substantial school fees, and local people were ardent for education. Bridewealth, which eventually came to be payable in either money or cattle, constituted a third use for money. Finally, people needed money for their discretionary wants, such as clothing, salt, tools and similar items.

Because there was no way of earning money locally—no markets, no industries, no farms, no plantations—the men had no option but to walk to regional labour markets to earn, as migrant workers, the money they required. Initially, in the years before World War I, people worked out strategies to sustain village life. Certain men would go to the labour markets of the south to earn enough money for several of their fellows to survive upon, for example, and those left at home would attend to village agricultural requirements, going for their share of migrant labour later, when their fellows returned.^[42]

Between 1912 and 1923, however, a succession of blows were dealt to such attempts to come to terms creatively with the necessity of labour migrancy. An outbreak of bovine pleuro-pneumonia in 1912 ended the trade in cattle which had become important in supplementing earnings from labour migrancy.^[43] During World War I men were drafted to serve as porters in the British army for periods of up to three years and the resulting shortage of male labour adversely affected food production in the village gardens. The army also commandeered large amounts of food and many cattle from the people, and great burdens were placed upon the women remaining in the villages as they tried to produce enough food for their families and themselves. The war was followed by the great worldwide Spanish influenza pandemic of 1919 which killed thousands of Tumbuka and Ngoni. This in turn was followed by a heavy round of inflation, an outbreak of Bubonic plague, a doubling of the tax rate in 1920, and a severe famine in 1923.^[44]

The impact of all these events resulted in an abrupt emigration of the male population from northern Nyasaland, with as many as 70 per cent of the men absent from home at any one time. This absence placed great strains upon village life in general and upon relations between men and women in particular. A crisis within the family resulted, with women whose husbands were absent bearing children conceived in adultery or deciding to seek divorce so as to remarry.^[45] As a consequence, men sought to assert control over women through recourse to institutions of African 'customary' law. They undertook this strategy at an opportune moment.

During the 1920s, the British were especially eager to implement a system of formal Indirect Rule. Chiefs and headmen in all districts of Nyasaland already played a considerable part in the collection of taxes and in general administration: in 1912, the passage of the District Administration (Native) Ordinance (DANO) had provided for the appointment of Principal Headmen and subordinate village headmen to whom would be delegated minor responsibilities for the general conduct and welfare of village life and for keeping the district officer informed of births, deaths, crimes, disputes and disturbances, and immigration. The village headmen had the additional important duty of 'allocating village gardens and pasturage' under the officer's direction.^[46]

In 1924 an amended version of DANO that strengthened the African role in controlling village society was passed. Principal Headmen and section councillors

were to administer 'sections' made up of 'village areas'. The delegated powers were substantial, with Principal Headmen in particular gaining power. They could hear cases referred to them by the village headmen and could charge a fee; they could officiate at weddings and grant or refuse divorces; they were responsible for tax collection which conferred advantages in the control of labour; they issued beer licences which brought a major industry engaged in by women under their jurisdiction; they controlled afforestation, which involved house building and much local industry; and they acquired for the first time clearly defined powers over village headmen, with profound consequences for the allocation of land. In 1929 a further amendment established courts at which the Principal Headmen could hear civil cases. All these provisions were consolidated and a system of Indirect Rule was formally established in 1933 with the promulgation of the Native Authorities and the Native Courts Ordinances.^[47]

From the British point of view the advantages of Indirect Rule in cutting costs, saving work and dividing the Africans into competing communities were obvious. But there were advantages for certain Africans as well, and not only the chiefs, the 'Native Authorities', who were to assume the new powers. One group who benefited were migrant labourers. The British used local chiefs and headmen as adjudicators in marital disputes, and consequently 'traditional' chiefs came to be central in the concerns and calculations of ordinary men who were eager to preserve their interests at home while working outside Nyasaland.^[48] In the context of the new political economy of migrant labour, then, chiefs had an enlarged role to play and men were willing—even eager—to accept chiefly authority and the historical mythology that legitimized it. Another group who could benefit were educated Africans, for they could count on being advisers of the chiefs. Given the powers ceded, however, the question of who actually secured the post of Native Authority was crucial, and, in the case of the Tumbuka, Edward Manda's intervention proved decisive.

Chilongozi Gondwe, Chief Chikulamayembe, had been in poor health for several years, allowing Manda's own influence to increase. In 1931 the old chief died. Although many elders initially supported the claims of Gogoti Gondwe, his appropriate successor in terms of 'custom', Manda persuaded them to accept John Gondwe, the son of the deceased chief and the Mission's candidate.^[49] He had been educated at the Mission and, being only 26 years old in 1931, he fulfilled Manda's requirements for a young, pliable, educated chief.^[50] A Tumbuka elder of the time, while acknowledging the appointment's irregularity, explained that when 'appointing John we knew that John had education, knew white men and modern affairs. It was because of this that we appointed John.'^[51] Confirming the strong linkage between education and growing ethnic consciousness, the new chief introduced compulsory education for all children in his chiefdom, the first (and, to our knowledge, the only) Nyasaland chief to do so. Within six years, a district officer could comment that 'compulsory education in Chikulayembe's country has ceased to be an experiment and is becoming an accomplished fact'.^[52]

Functioning both as a theocratic éminence grise and as a progressive, Manda enthusiastically manipulated the Past to bolster the public image of his chosen instrument, the Chikulamayembeship.^[53] In 1932, he proposed that the new chief should rule over all 'Utumbuka', 'the Land of the Tumbuka'. Although this included territory never under past Chikulamayembes, Cullen Young's work was cited to justify the claim.^[54] In the following year the Chikulamayembe and his supporters 'invaded' the neighbouring Mwafulilwa area to 'annex' it, earning an official reprimand.^[55] Soon afterwards, the chief, aided by Manda, attempted covertly to establish a Tumbuka chiefdom in the heart of Ngoni country, near

Hora Mountain, where Tumbuka rebels had fought the Ngoni in 1879, seeking, as the District Commissioner explained, 'to make this a centre from which Tumbuka influence would spread and eliminate Angoni rule'.^[56] The Tumbuka would then be able to live 'without the stigma of subservience to Angoni rule'.^[57] These efforts failed, but they revived bitter anti-Ngoni feelings, especially amongst the partisans of Tumbuka lineages whose authority the Ngoni invasion of the mid-nineteenth century had extinguished.^[58]

By the early 1930s, then, largely because of the Mission's educational work and Tumbuka men's willingness to accept increased chiefly authority, Tumbuka ethnic consciousness had become a reality, focusing attitudes, instilling pride, and helping to shape actions. The census of 1931 noted that there was a marked decline in the number of people identifying themselves as 'Ngoni', caused by

the tribal consciousness in the peoples amongst whom they live. In the days of their greatest power, it was politic and long afterwards fashionable . . . to claim membership in the all-powerful Ngoni tribe. . . . The indigenous tribes are no longer afraid or ashamed to call themselves by the proper names.^[59]

Ngoni Ethnic Consciousness

The upsurge in ethnic pride in Nyasaland's Northern Province during the inter-war years also affected the Ngoni. In 1933 the British recognized Lazaro Jere as Native Authority for the area of Mbelwa's Ngoni and as Paramount Chief of the Ngoni, reviving a title which had been banned since 1915. Superficially, this triumph for Ngoni ethnic consciousness is difficult to distinguish from the Tumbuka version just to the north. An alliance of chiefs, African intellectuals and ordinary villagers, coinciding with British concern for the social health of villages affected by labour migrancy and moves towards Indirect Rule, had produced a political solution identical to the Chikulamayembeship in everything except the title itself. Like the Tumbuka, the Ngoni had their culture brokers—for the Reverend Edward Manda, the Reverends Charles Chinula and Yesaya Chibambo; for Thomas Cullen Young, the British anthropologist Margaret Read.^[60] Behind such similarities, however, lay profound contrasts in meaning and causation.

For the Tumbuka building their ethnic self-awareness, the opposing 'they' were primarily the Ngoni. Tumbuka attitudes to the British as liberators and educators, oppressors and exploiters were necessarily ambiguous, and the specifically anti-colonial ingredient was a comparatively late addition to Tumbuka ethnic consciousness. For the Ngoni, however, the problems of ethnic awareness were at once far simpler, yet also more complex. On the one hand, the Ngoni enjoyed a reputation throughout southern Africa as effective soldiers and administrators. Like the British, they had come to Nyasaland as conquerors, and they had been able to maintain their independence until 1904, long after their neighbours had conceded sovereignty. In short, the Ngoni knew that they were a people with whom one had to reckon.

On the other hand, the final quarter century of Ngoni independence had been plagued with problems. Their heartland had been severely damaged by overgrazing, the destruction of forest cover, a declining water table, and falling soil fertility.^[61] This damage was aggravated, first, by the extension of the pax Britannica which ended the annual raiding by which the Ngoni had supplemented their food supplies, restricting them to their exhausted heartland, and, second, by the devastation of their cattle herds by the rinderpest epizoötic of the

mid-1890s.^[62] In 1903, famine finally pushed large numbers of people to move out into the depopulated lands around them, thereby violating an earlier agreement with the British. Urged on by the Scottish missionaries, the Ngoni accepted British sovereignty in September 1904, the governor agreeing that the Ngoni should retain their Paramount Chief and enjoy other freedoms denied to all the other peoples of Nyasaland. In the eyes of the Ngoni, their chief was equal in status to the local British district officer.^[63]

It was the abrupt imposition of a heavy tax burden on an impoverished people in an impoverished land that is still most vividly remembered:

Taxes were the main problem. . . . A hut tax of 3 shillings was introduced in 1906. That year, the people paid the tax, but when taxes were demanded afresh in 1907, the people rebelled, saying 'Should we pay taxes a second time? No!! That cannot be so!! We invited in government, not repeated tax collection.' It was after the tax rebellion of 1907 that the government sent Reuben, Madondolo, and Pickford to bum the huts of those who refused to pay.

There was wailing . . . villages were burnt here in Embangweni, in Engalaweni, everywhere. The main complaint was about the system of taxation which was bad in those days. They arrested anyone who defaulted in the payment of their taxes. Those who defaulted were often subjected to ill-treatment. Even if they were girls, they were tied up with ropes and beaten with the sjambok . The people had to migrate to Harare [Southern Rhodesia]. There was no money in this country, so they had to walk all the way to Rhodesia.^[64]

This testimony is revealing in two ways. First, it shows well over half a century after the event the intensity of Ngoni anger over the tax issue. For the Tumbuka, accustomed to paying tribute to the Ngoni or Ngonde chiefs, British taxation involved primarily—although by no means only—a shift of allegiance. For the Ngoni, by contrast, it meant the surrender of their sovereignty to a new conqueror, a humiliating sign of tributary status for a people who had migrated all over southern and central Africa to avoid just such a fate.

Second, the testimony records that although tax had to be paid in money, there was 'no money in this country'. Where the soil was exhausted, as in Ngoni country, and where there were no markets for food or cash crops, as was the situation everywhere in northern Nyasaland, and when an outbreak of bovine pleuro-pneumonia soon made it impossible for the people to sell cattle to raise money, the imposition of the tax brought to the Ngoni the demands of a political economy based on a general labour migrancy that had already affected the Tumbuka north of them.

On the material level the results of labour migrancy for the Ngoni were the same as for the Tumbuka, but for the Ngoni, accustomed to being rulers, they were more galling. The British found the discontented Ngoni difficult to control, unlike the Tumbuka, and, reflecting their administrative exasperation, the acting governor sourly observed in 1913 that nothing would 'benefit them or the country more than to shoot a few down, burn their villages, deport the so-called chiefs and confiscate their cattle'.^[65] In July 1914, when the governor met with the Ngoni leaders, mutual recriminations filled the air. The governor asserted that the chiefs were encouraging 'their people to evade payment of hut tax, deceived the Resident when he applied to them for information and assistance', and did not merit their governmental subsidies. The Ngoni chiefs complained of the lack of markets in northern Nyasaland and the oppressive weight of taxation.^[66]

Shortly afterwards the Ngoni hierarchy was dealt a traumatic blow. In 1915,

when a British official tried to raise men to serve in the much-feared Carrier Corps in the East African campaign of World War I, Paramount Chief Chimtunga forbade it.^[67] For this, he was removed from office and banished to the Southern Province, and DANO was applied to Ngoni country for the first time, reducing the chiefs to little more than assistants to the District Commissioner for mobilizing labour and collecting taxes. In the 1970s people still recalled this shattering event in an ingoma song:

This blow to Ngoni self-respect was followed by the deaths of a great many of the Province's people in the war, in which they were forced to serve as porters.^[69] The severe reverses of the post-war years followed.^[70] The imposition of colonial control thus coincided with a rapid and general deterioration of conditions of living in the area, generating profound resentment. For the Ngoni, the 'they' in opposition were clearly the agents of the new colonial political economy.

In 1919 a group of Livingstonia-educated clerks, teachers and clergymen established the Mombera Native Association, modelled on 1912's North Nyasaland Native Association. Over the years, this association was dominated by the Reverend Charles Chinula, who, though of Tumbuka origin, came to be as ardent a protagonist for the Ngoni ruling elite and advocate of Ngoni history as Edward Manda was for the Tumbuka and their version.^[71] Chinula was assisted in this work by the Reverend Yesaya Chibambo, also of the Mission, who in a speech of 1920 succinctly expressed the aims and attitudes of the new organization:

The country is now in a new era with a new life, new knowledge, new resolutions, new laws, new customs which can be learned through education: it would be foolish and ridiculous if people of the country dislike the civilization. The old life differs greatly from the present life, and it would be wise for the people of this country to aspire to have education, which alone leads to civilization.^[72]

Significantly, however, the path to this progressive future was seen as passing through a celebration of the Past.^[73] The Ngoni past was vivid and, when compared with the grim realities of ecological decay, labour migrancy, and the undermined authority of chiefs, it seemed indeed glorious. The migration from Natal, the deeds of Zwangendaba, the victories of Ngoni armies, Ngoni skill in state-building, their high level of culture—all these were celebrated and publicized by Yesaya Chibambo in his role as historian in the book Midauko: Makani gha waNgoai (History: The Deeds of the Ngoni ) (Livingstonia, n.d.), which was prepared specifically for use in the local Mission schools.^[74] To talk of the Ngoni past was, then, to speak of real and continuing structures of power.

In dealing with the British Administration, however, the Ngoni chiefs found it useful to use a less traditional face. As the District Commissioner minuted in 1930:

My experience of the District Councils in the Northern Province is that the Chiefs are so tired of trying to get their grievances redressed that they turned to

the Native Associations in the hope that the latter, being more educated, would bring greater pressure to bear upon the Government.^[75]

In practice, the Mombera Native Association had from its inception been fulfilling this role. Among its first campaigns, the Association took up the cause of the banished Chief, Chimtunga Jere, persuading the colonial authorities to permit his return home, even though he was allowed no political role. After Chimtunga's death, it championed the claims of his son, Lazaro Jere, a Mission-educated clerk employed in the Northern Rhodesian administration, to return to Nyasaland as 'Paramount', a title still outlawed.^[76] Once Lazaro returned in 1924, a well orchestrated campaign to revive the Paramountcy itself began, resulting in popular excitement so great that the District Commissioner contemplated calling in troops to suppress the movement.^[77]

The adoption of Indirect Rule, however, was approaching, and in 1928 the government recognized Lazaro Jere as Principal Headman. He immediately confirmed the Association's role in promoting his advancement by becoming its chairman.^[78] Not surprisingly, the Association continued its campaign in terms appropriate to a specifically ethnic consciousness, fighting off an attempt in 1929 to transfer the northern fringe of Ngoni territory to Chikulamayembe, and in 1930 arguing once again that the 'desire to have a paramount chief in Mombera still rings in the hearts of the people, for the present policy of equalizing all the Principal Headmen is contrary to the law of the country'.^[79] In 1933 the government yielded and recognized Lazaro Jere as the new Paramount, Mbelwa II. Like the Chikulamayembe, he immediately became expansionist. Within a year he succeeded in annexing part of the area of the neighbouring Chewa chief Kaluluma and unsuccessfully attempted to acquire the adjacent Northern Rhodesian Ngoni chiefdoms of Magodi and Pikamalaza.^[80] In 1938, he petitioned the state that the entire area of Northern Rhodesia between the Nyasaland border and the Luangwa river be placed under his administration, but this manoeuvre also failed.^[81]

In sum, then, by the early 1930s two strong ethnic ideologies had been created in northern Nyasaland. In the Tumbuka case, the strongest creative influence was the intellectual input of graduates of the Scottish Mission's schools; in the Ngoni case, the strongest influences were the still-living memories of past independence and prosperity. For both groups, however, it was the demands of the colonial political economy and, most especially, the fact of widespread labour migrancy that provided a firm underpinning to these culturally defined revivals. While men were in Southern Rhodesia or South Africa earning money as labour migrants, they thought it essential that their local interests in land, cattle, and, particularly, control over women and children should be protected. Indeed, in recognition of the roles of the chiefs in preserving social order in the villages, migrants customarily presented the chiefs with gifts on their return home. Financial considerations, therefore, reinforced the chiefs' natural desires to maintain social order in their areas.^[82]

In essence, Tumbuka and Ngoni ethnic ideologies during the inter-war period were products of a dialogue between labour migrants who wanted social controls enforced and African intellectuals who sought to shape these controls so as to encourage what they saw as Progress for their people. Ethnic consciousness in the Northern Province was, then, a form of resistance to colonialism that asserted the validity of the African way of life and the African past through a stress on 'tradition' while still looking forward to a future of 'Progress' through western education and training.

Language Policy and the Creation of Northern Regionalism

During the 1930s the Northern Province's two major ethnic movements, with their ideological core of history and chiefly authority, partially coalesced into a regional movement that was defined by common interests and specified by a new concern, language. The issue that precipitated this change was the government's language policy. In 1918 one of the administration's junior officers resurrected the old idea that Nyanja should be made the official language of the country and taught in all its schools. As a proposal it made good sense, for Nyanja, or its dialectal variants, Mang'anja and Chewa, were spoken by a majority of the country's people. The governor, Sir George Smith, rejected the suggestion totally, however, pointing out that:

Though the spread of one dialect through the country would be advantageous . . . it would tend to merge the various tribes in the Protectorate at a greater rate than at present, and this I consider undesirable. One of the chief safeguards against any combined rising is the individualism of the various tribes, and with a small and scattered white population, this I think should be postponed.^[83]

Divide-and-rule was to be British policy. In the north, the language that was granted official status was Tumbuka. Throughout the 1920s the Mission's presses confirmed the status of Tumbuka by pouring forth school texts in the Tumbuka language in editions of between 7000 and 10,000 copies.^[84] For the Tumbuka themselves, it was a symbol of their reviving respectability and self-esteem. For the Ngoni, it was their adopted language within the context of a larger society of potentially competing languages.

In the late 1920s, the government's language policy shifted. Governor Sir Shenton Thomas was not haunted by visions of uprisings nor preoccupied with strategies of divide-and-rule. He was eager to streamline the colony's administration, and he argued that the adoption of a single official lingua franca would both help unite the country and save money.^[85] Following Thomas's suggestion, the Advisory Committee on Education adopted a proposal that Nyanja 'be introduced as the medium of instruction not later than Class 4 in all Government and Assisted schools'.^[86] To obtain state aid missions would be obliged to follow this policy.^[87]

This decision distressed the Scottish missionaries of the north. On 15 July 1933 the Livingstonia Mission announced that it was 'unable to accept this ruling', attacking Nyanja on several grounds. First, it was educationally unsound and simply would not work. Second, it would inconvenience the Mission, which would have to find Nyanja teachers. Third, Nyanja was a bad choice because it was not a 'language of higher cultural and linguistic value'. Fourth, it was a politically unsound decision, because the people themselves opposed it.^[88] Rather than introduce Nyanja into the north, therefore, the Mission would forgo governmental aid. The head of the Mission, W.P. Young, the brother of Cullen Young, put the argument in somewhat less reasoned terms:

Politically, it is an unfortunate moment to choose to attempt to turn back the pages of history. When the Livingstonia Mission began work the local people were under the domination of the Angoni. The Tumbuka . . . were a scattered and subject people, whose language was proscribed. Yet they clung to it as the symbol of their identity as a people . . . to them in a peculiar sense, their language is their life.^[89]

The government had indeed chosen a peculiarly inappropriate moment to implement its decision. Indirect Rule had been introduced only recently in the north, and African leaders were quick to echo the Mission's objections. They argued that it was 'unfair to force a people to accept a language which they do not wish' and, tellingly, that 'people go to schools to learn their own vernacular books, after which they wish to learn English, which is more profitable'.^[90] The language issue resulted in a merger of local Tumbuka and Ngoni ethnic consciousnesses into a new Northern regional coalition glued together by the fact that these groups now possessed a common language in a country of many languages. Thus, when a new governor, Sir Hubert Young, visited the north to convince the people that the new policy was in their best interests, fierce opposition greeted him. At a gathering of Ngoni leaders on 1 October 1933, Charles Chinula remonstrated that 'Chinyanja is not wanted in this Tumbuka-speaking area.' When Young travelled further north to speak with Tumbuka leaders, their spokesman told him that 'Tumbuka should be preserved for future generations just as seed for native produce, domestic and wild animals are preserved for them.'^[91] In a minute of 19 October, Levi Mumba, the ranking Tumbuka-speaker in government service and the first African to sit on the influential Advisory Committee of Education, aligned himself with the anti-Nyanja forces, arguing that it was much too early to have a lingua franca in Nyasaland and that, if ever one were adopted, it should be English.^[92]

The governor was stubborn, however, and drawing support from the Roman Catholic and Dutch Reformed Churches, both of which used Nyanja as their medium of instruction in the Central Province, he appealed to Whitehall for support. In the far north of the country Ngonde-speakers, resenting their long social subordination to Mission-educated Henga, joined in attacking Tumbuka, asserting that 'We Bangonde would like our children to learn Nyanja in schools and not Tumbuka.'^[93] The Colonial Office initially agreed with the governor, commenting that the Livingstonia Mission must face facts and accept Nyanja. If they did not, their students' careers would be blighted because of their ignorance of the official language of the country and their consequent ineligibility for positions in the civil service.^[94] Therefore, in 1935 Young's successor, Sir Harold Kittermaster, ordered the immediate implementation of the new language policy.

The Mission, however, refused to give up, carrying the fight for their educated graduates over the heads of Nyasaland's officials to London and succeeding in gaining the sympathy of key Whitehall officials.^[95] As one noted:

It does seem to me a pity—to put it no more strongly—that because of this persistent pursuit of a policy about whose merits there is considerable dispute, Government should run the risk of alienating a Mission in Nyasaland which is doing wonderful work and which . . . is only too anxious to co-operate with Government wherever it can.^[96]

The Mission was victorious. London instructed Kittermaster to hold another conference and to impose no policy against the Mission's wishes.^[97] In mid-1936 this conference was held. The Mission's representative asserted that the mother tongue was the 'soul of the people' and that to impose Nyanja as a lingua franca would be tantamount to the suppression of Nyasaland's other languages. At this conference, Levi Mumba not only deprecated the whole idea of a lingua franca but went so far as to say that Nyanja should never be considered as a subject to be taught in schools, as its introduction 'would interfere with the mental development of the children'.^[98] After World War II Tumbuka gained an additional victory when, in 1947, it, together with Nyanja, was made one of the

two official languages of the country, a position it held until 1968 despite the fact that it was spoken by only a very small fraction of the population.^[99] The alliance of educated Africans and well-connected Scottish missionaries was a potent one, and it ensured that the bulk of African positions in the colonial civil service would be taken by Tumbuka-speaking northerners as rapidly as such positions opened up.

Chiefs, Planters, and Immigrants in the South

Conceivably, events in Nyasaland's Southern Province might well have followed a similar course. As in the north, the south had experienced, well before the British arrival, the invasion of other African peoples and the overthrow of established political systems. During the 1860s, Yao-speaking people from Mozambique had conquered the Shire Highlands, ruling as a group of competing warlords over the Mang'anja-speaking autochthones. The Shire valley was taken over by the Makololo porters of David Livingstone, who ruled from stockades along the riverbanks. The indigenous Mang'anja chiefs, including the Paramount Chief, Lundu, were killed and their people incorporated into the new Makololo chiefdoms.

When Scottish missionaries established the Blantyre Mission in 1876, they attracted their first converts from the defeated and enslaved Mang'anja. Mang'anja became the language of preaching and education, the language used in Biblical translation, and the first language of the country to have a scholarly dictionary. It is possible, then, to sketch out a hypothetical history in terms of which the Blantyre Mission might have become a focus of a resuscitated anti-Yao, anti-Makololo, pro-Mang'anja consciousness, campaigning for the restoration of Chief Lundu's Paramountcy, drawing into its service the administrative and political talents of such educated converts to Christianity as John Grey Kufa and John Chilembwe, and obtaining critical support from the local Scottish missionaries.

Yet events followed quite a different course. Local African leaders who opposed colonialism, perhaps influenced by the distinctly non-particularist, universalist message of such foreign missionaries as Joseph Booth and James Cheek, made no attempt to mobilize African cultural symbols or to formulate a view of the African past in their opposition to British colonialism. It was not, in fact, until after Malawi's independence in 1964 that political leaders appreciated the possibilities of the use of a crafted past as a mobilizer of political support, when President Banda added an historical dimension to the Chewa/Mang'anja ethnic political coalition he had built by reviving the defunct Lundu Paramountcy and crowning the new chief with a ceremonial pith helmet.

The political economy of the Southern Province was crucial in shaping its history in the twentieth century, and pivotal in shaping that political economy was the fact that land alienation to Europeans occurred in the nineteenth century, before the establishment of a British administration. Vast estates comprising almost one million acres were obtained by a handful of settlers and companies during the 1880s as part of their strategy to induce the British government to annex the Shire Highlands and adjacent riverine areas before the Portuguese could do so. One reason such land alienation was possible was that the bulk of the land 'purchased' from the chiefs was relatively unoccupied. In 1861, missionaries had found the Shire Highlands thickly populated with Mang'anja villages. Disruptions from Yao invasions, the expansion of the slave trade, and a disastrous famine in 1862 soon resulted, however, in an entirely different pattern of

settlement wherein the major chiefs ruled from heavily fortified stockades on mountain tops. The plains reverted to secondary forest and were by the 1880s thick with game. With minor exceptions, it was this largely unoccupied and uncultivated land which was alienated to Europeans in the 1880s. The planters were thus in a strong position to defend their interests against both government and missions after the British annexed the area in 1890, successfully insisting upon the de facto right to run their estates pretty much as they wanted, without governmental interference and without an unwanted mission presence. Most planters barred mission work on their estates, and, therefore, in the south no networks of mission schools located in the villages developed as they did in the north. In the Southern Province education remained a relatively rare phenomenon.

From the 1890s onwards, the issue that dominated the politics of southern Nyasaland centred on the nature of the terms on which Africans would be permitted access to the mostly empty lands held by the European planters or the Crown Land that still remained under Yao chiefs.^[100] That this could become a central issue was because of the entry into Nyasaland from Mozambique of groups of 'Nguru' peoples seeking land. This immigration began in 1895 and continued for several decades. Some were slaves freed from the chiefs' stockades, while others were fleeing Portuguese tax and labour policies. They spoke various languages, including Lomwe, Mpotola, and Mihavani, but nothing called 'Nguru'.

When these migrations began, the area's European planters were struggling to find a suitable product for export, and their main problem was finding an adequate labour supply. Though a hut tax had been devised to solve this problem, it had proved inefficient. Even when local people could be induced to pay their tax in labour rather than in cash or kind, it generated only one month's work per man per year, allowing no time for the development of skills and producing a labour surplus in the dry season but virtually none during the rains, when the bulk of the agricultural work had to be done. Moreover, when people needed to earn money, they found it more advantageous to leave the country altogether to seek the higher pay available elsewhere in southern Africa. In the planters' view, Nyasaland's people had too many alternatives.

What made the new Nguru immigrants so valuable was their vulnerability. As immigrants, they lacked land. By accepting land in return for their labour, they could be turned into a captive workforce. Two groups took advantage of their vulnerability: the European planters and the established Yao chiefs and headmen dwelling on Crown Land. As the Nguru migrants crossed the border, the planters had vast tracts of empty land available for settlement which they offered to the newcomers under terms by which they exchanged land for labour—a system known as thangata . According to the legislation of 1904 which defined thangata, workers were to be provided with eight acres of land for settlement and cultivation, the 'rent' on this land being one month's labour per year in lieu of hut tax, plus one month's thangata labour paid at the current rate of tax. The real attractions of the system for the planters lay in its hidden advantages. A month's hut tax labour could be stretched to six or eight weeks simply by withholding a signature from the tax certificate. Thangata agreements were informal and verbal and not subject to government review. Most planters had little difficulty in extending the actual labour service to four or five months. And unlike hut tax or tax certificate labour, it could be demanded in the rainy season.^[101] The Nguru were in no position to bargain. If they refused to work or if they attempted, as others did, to find work in South Africa or Southern Rhodesia, they lost their right to land in Nyasaland. The planters therefore encouraged the Nguru to settle. As

the governor commented a few years later, Nguru immigration had come 'most opportunely. It populated vacant spaces, it enhanced the Protectorate's revenue and most important of all it has provided a ready and permanent labour supply for the extension of European enterprise.'^[102] This situation continued well into the 1920s.

The planters were not the sole beneficiaries of the migrations. Some Nguru also settled under the protection of chiefs and headmen on Crown Land. These chiefs and headmen were Yao-speakers established in the area since the 1860s. Once the British established an administration, people began to move down from Zomba, Chiradzulu and Mulanje mountains and reoccupied the abandoned plains, growing crops. Over the next fifteen years the whole of the Shire Highlands was repopulated.^[103] To clear the land required labour, and it was the Nguru immigrants who supplied it. The chiefs and headmen gave them food in exchange for their clearing fresh land and growing cotton, the area's major cash crop. Their labour was also used to produce maize and vegetables for sale. On the whole, the British approved. Although there were accusations of slave-holding and slave-trading, the chiefs were successful in getting cotton and food production under way and were useful in supplying public works labour to the administration. The consensus was that although the Nguru were 'kept in a certain degree of mild subjection and occasionally perform a little menial labour for the protection of the chiefs under whom they serve, there is no serious interference with their rights and duties'. The chiefs who were to come into prominence in the colonial period were precisely those who attracted the largest numbers of settlers.^[104]

The result of the welcoming of the Nguru by planters and chiefs so as to gain access to their labour was the establishment in the Southern Province of a population of great ethnic complexity, a mélange of diverse peoples and cultures. The Mang'anja and Nyanja peoples had been overlain with—and ultimately outnumbered by—Muslim Yao people in the mid-ninteenth century and by the Nguru immigrants from Mozambique in the early twentieth century. All these groups retained their own cultural practices, and the pattern of scattered settlement throughout the area undermined all possibility of defining geographically discrete ethnic areas.

Tribalism from Above: the Strengthening of the Yao Chiefs

Yet in 1915 an event occurred that was instrumental in imposing a clear, specifically ethnic interpretation upon the south's heterogeneous population, at least in the perceptions of its British rulers. This was the Chilembwe Rising, an upsurgence of popular discontent on certain plantations of the Chiradzulu district under the leadership of the Rev. John Chilembwe, head of the Providence Industrial Mission.^[105] Chilembwe had been influenced by the non-particularist message of such missionaries as Booth and Cheek and then educated in America. He had adopted, as others of the Southern Province's educated petty bourgeoisie had, a universalistic Christian message as the ideological base for his opposition to colonialism.

The Chilembwe Rising was crucial in shaping later British attitudes towards Africans in two ways. First, it reinforced an already well developed colonial distrust of educated Africans in particular and of Protestant mission education in general.^[106] Planters and administrators alike henceforth did all they could to

separate members of the anti-colonial petty bourgeoisie from possible popular village support against the planters' control over the local economy so as to avoid a repetition of the Chilembwe phenomenon.

Second, and related directly to this distrust, the Rising encouraged the British to impose Indirect Rule on the confusing tangle of African people in southern Nyasaland in an effort to check any possible disturbances by dissatisfied Africans through bolstering chiefly control. Indirect Rule was essentially conservative, employing its own concept of tradition and drawing its personnel from chiefs and headmen. For the British the more 'traditional' Africans were, the better, and they warmly embraced the most obviously conservative Africans in the area, Muslim Yao chiefs. In the Southern Province, then, a 'tribal order' was to be largely imposed from above rather than being shaped from below, as occurred in the Northern Province.^[107]

As one reads official documents after 1915, it is difficult to avoid concluding that British officials who were pursuing control through divide-and-rule policies invoked ethnic categories that existed only in their own minds. The contrasts for them seemed to be between law and disorder, between the trustworthy Yao and the untrustworthy Nguru.^[108] The insecure British administration's differentiation of Africans into 'good' Yao and 'bad' Nguru began in the immediate response to the Rising itself. The vast majority of those found guilty and sentenced to death or to long terms of imprisonment were members of Chilembwe's church, and the vast majority of these were Nguru. No declared Muslim was found guilty, although several Christian Yao were sentenced.^[109] Because of the clear Nguru support of Chilembwe, it was obviously dangerous to be labelled an 'Nguru', and many immigrants therefore claimed that they were 'Yao'. It was also dangerous to be a Christian. 'We were arrested in our village by Yaos because we were Christians,' was a frequent remark in the testimony of those brought to interrogation.^[110]

With this opportunity presented to them, the Yao chiefs and headmen speedily acted to protect and, if possible, improve their positions. In the week after the rising, the Yao chiefs and headmen of Chiradzulu presented themselves at the British administrator's office, assuring him that they did not support Chilembwe. They ingratiated themselves with him through gifts of flour and eggs and chickens, and he responded, finding that he enjoyed talking to them. He liked their Islamic robes, so different from the disturbing mimicry of the European suits worn by Chilembwe and many of his followers. Reassured and grateful, he wrote categorically that the Yao chiefs were all completely loyal. Two months later, his successor in Chiradzulu reported that the chiefs were being cooperative in providing road repair labour.

The British decided to seize the opportunity of the Yao chiefs' goodwill and to proceed, years in advance of anywhere else in the Southern Province, with the implementation of the DANO (1912) and with the appointment of Principal Headmen and village headmen to assume the duties laid down by the ordinance.^[111] In their embrace of the Yao chiefs what the British were searching for some analogue to the English class system. The Principal Headmen were, after all, being incorporated into the colonial bureaucracy, some way down the ladder from the district commissioners, but nonetheless needing to display some of the same natural ability to govern.

The problem, given the masses of Africans all remarkably alike in their material poverty, was how to distinguish which ones were the natural gentlemen. With Christianity and mission education threatening to produce an African bourgeoisie hostile to colonialism, and yet with an official embrace of Islam a political

impossibility, the British were forced instead to formulate ethnic theories of African differentiation. Given the mixture of 'tribes' within a single area and the impossibility of finding distinctions between them, they postulated for the south a labour-specific hierarchy of 'tribes' that occupied the same land.

An official publication illustrative of the thought processes of British officials as they defined ethnic labels is S.S. Murray's A Handbook of Nyasaland (1922). Murray correctly comments on the vagueness of such terms as 'Nguru'. The Nguru are, he explains, a number of different peoples loosely allied in the Makua-Lomwe group but bearing separate designations (Atakwani, Akokola, etc.) that refer to different districts of origin in Mozambique. Similarly, with respect to the Yao, Murray comments that because of intermarriage which has robbed them of their 'finer features', the 'majority of the so-called Yaos' in the Shire Highlands 'have little claim to the name'. Yet this recognition of the obvious cultural heterogeneity of southern Nyasaland does not prevent him from detecting quite specific 'Nguru' and 'Yao' traits. The Nguru, he claims, 'are represented among the idle and criminal classes to a disproportionate extent'. The Yao, on the other hand, are 'intelligent and quick', making 'excellent servants' while 'as soldiers they have proved of inestimable value'. They also speak 'perhaps the finest of all Central African languages'. The Yao are seen as being more like Europeans than any other people of the Southern Province. They live in 'square houses' and cultivate habits of 'personal cleanliness'. It was felt that they understood a certain man-to-man equality of address. Unfortunately, and as if it were a by-product of their intelligence and fine features, the Yao are 'poor cultivators of the soil'.^[112]

Before DANO could be implemented, however, 'villages' for Yao headmen to head had to be created from the ethnic soup. Complaints had noted that houses were 'scattered in twos and threes all about the place', making it difficult to collect taxes and to keep good order generally. The British ordered that houses be 'concentrated' into groups of no fewer than twenty. Many thousands of people had to be relocated, and it was impossible to join four adjacent settlements and appoint a headman without political trouble. Yet, despite such problems, this plan to form new villages was implemented.

The majority of those most directly affected by hut concentration were Nguru, for the power of the newly appointed Yao village headmen to allocate land put all immigrants firmly in their power. As Nguru immigrants continued to pour into the area during the 1920s and 1930s, they needed to secure a tax certificate, a document carrying a chief's name, as evidence of their legal residence in the country. This ensured continued advantages to the Yao chiefs and headmen from whom they had to request permission to settle. Chiefly control over land effectively made Nguru labour available to the Yao chiefs and headmen on akapolo ('slave') terms, just as it had been made available to the European planters through the thangata system. When, therefore. World War I created a demand for porters for the Carrier Corps, the newly installed Yao chiefs, in marked contrast to the Ngoni Paramount in the north, Chimtunga, responded with alacrity. From Chiradzulu district alone, 2300 porters had been despatched by the end of 1915, and, the chiefs being 'cooperative', 800 more were provided in January 1916.^[113] Virtually all the 6000 porters supplied from the district during the war were Nguru.

This government-sponsored political differentiation between Yao chiefs and Nguru commoners also had a clear economic parallel. By April 1916, exactly one year after village consolidation had begun, tobacco was being cultivated as a cash crop. The new crop was fire-cured dark leaf tobacco. No reliable production

figures are available until the mid-1920s, but descriptions of the crop make it clear that this was a development of great importance to the country's economy. By 1930 tobacco grown and cured by Africans represented almost 75 per cent of Nyasaland's tobacco exports. It was the ethnically specified redistribution of power on the Shire Highlands which provided much of the opportunity for this expansion in production as chiefs could use Nguru labour to produce the crop in exchange for permission to settle.^[114]

The fallout from the Chilembwe Rising, the demands of the war, and the desire for bureaucratic convenience had all made the Yao chiefs seem indispensable to the British. DANO proved its worth during World War I in suppressing opposition to colonialism, in keeping the supplies of labour flowing, in getting cash crops grown and taxes collected, and in reducing officialdom's bureaucratic burden. In the years after the war, therefore, official support for the political and economic authority of the Yao ruling elite continued to grow. And, as the alliance between the British administrators and the Yao elite deepened, the British came to see the Yao, their chosen instruments in Indirect Rule, as a people with a real history, in marked contrast to other local Africans, who had only customs and folklore. In 1919 Yohanna B. Abdallah's The Yaos was published in Zomba by the Government Printing Office in both Yao and English versions. Abdallah's aim was to 'write a book all about the customs of we Yaos, so that we remind ourselves whence we sprang and of our beginnings as a nation'.^[115] It is significant that Abdallah was a priest of the Anglican Universities Mission to Central Africa, the first African to be so ordained in Nyasaland. Thus his book effectively made the point that some Yao were not Muslims: they were loyal members of the Church of England.

During the 1930s the contrasts between the Yao and Nguru 'tribes' were made in ever starker terms. In 1936, for example, a district officer produced an evolutionary account of Yao history that was even more useful to the British administration than Abdallah's had been. According to him, the Yao had their origins in Mozambique as family units, small matrilineal communities often living many miles apart. Because of threats from the Portuguese and Arabs of the East Coast, they coalesced into larger communities, living under chiefs who were responsible for organizing their security. The chief who led in war came from the largest of the various units that had amalgamated. Then there were coalitions of the larger groupings, the family units having thus evolved into the 'tribe' or the sub-sections of the 'tribe' under powerful chiefs. At this stage of development, which coincided with the migrations into Nyasaland, there were thus three levels of power among the Yao: the paramount chiefs, the subordinate chiefs, and the village lineage head. Although the paramount was basically a military figure, his position 'rested largely on his reputation for fairness'. No Yao paramount had the power of a Zulu or Ngoni chief: if the Zulu chief was like Caesar, the Yao paramount was more like an English prime minister! The Yao, then, were a tribe with a true history: they had evolved through the proper stages into something like a nation.

The Nguru, according to the official's investigations, had evolved in just the opposite direction. Originally united under the chieftaincy of Mwatunga, the tribe had disintegrated into family groups taking their names from their relationship to Mwatunga's respective wives. Thus the Nguru were no longer a 'tribal unit'. Even the name 'Anguru' as a term of unification was only a fiction, this discovery by a twist of colonial thinking being adduced as fresh proof of their intrinsic inferiority.^[116]

This vision of the Nguru as inherently inferior beings was both a reflection of,

and rationalization for, their subordinate social position in the area's political economy. As such, it was reinforced as a result of changes in Nyasaland's economy in the Depression years. European planters and Yao chiefs alike had encouraged Nguru settlement for decades, and as a result of these policies there was a severe overpopulation of the Shire Highlands and a serious threat of soil erosion. In response to this situation, then, certain planters tried to evict what they now viewed as 'surplus' Nguru. While concerned about the ecological problem, the government opposed this plan because of the congestion that existed already on Crown Land and because the Nguru were living in such abiding poverty and insecurity that there was a likelihood that they might stage a bloody revolt if subjected to any further pressures.^[117]

To many it was already clear that the only solution lay in the purchase by the government of the largely unused European-held estates, the abolition of the thangata system, and the general relief of congestion in the Shire Highlands by land settlement schemes. The public debate about the role of the Nguru became extraordinarily virulent. Spokesmen for planters who had begun to fear that they might be displaced by the very Nguru whom they had themselves settled on their land lashed out in letters to the Nyasaland Times . Why should Britons die, raged one, 'to make Nyasaland a safe boozing den for alien Nguru?' There was insufficient space in the newspaper, clamoured another, to show 'from the history of these people the steps by which they became in turn slave-trading gangsters, irregular soldiers, cringing-starving-unclothed refugees, and finally under a safe benign government: drunken, slothful and vicious'. The older colonial historiography had been turned on its head: all along it had been the Nguru who were the slave traders; they were 'candid bandits, their prey human flesh and blood, and having gorged like hyenas, they then returned to Manguru for the most part replete'.^[118] The wheel of European opinion had come full circle: the former 'raw Anguru' who had been welcomed for decades as cheap workers were now cannibals!

While the government considered the question of post-war land reforms, a new ethnic initiative made its appearance. The reasoned African reply to the letters to the Nyasaland Times was coordinated by Charles W. Mlanga. Within two years Mlanga had become the first secretary general of the new Nyasaland African Congress, the first specifically nationalist movement in the country. One of its committee members was an educated Nguru, Lewis Bandawe. Bandawe, who was to serve the Lomwe in much the same way that Edward Manda had served the Tumbuka, had been born in Mozambique in 1887.^[119] He was educated in the schools of the Blantyre Mission and in 1913 he returned to Mozambique to teach, remaining there until 1928. He worked as a teacher, acted as head of the local mission for long periods of time, and translated the New Testament into 'Lomwe'. On his return to Blantyre, he broke with the Blantyre Mission and became a clerk in the Judicial Department, eventually rising to the rank of deputy registrar.

With his long experience of both Nyasaland and Mozambique, Bandawe was ideally placed to become a spokesman for the despised Nguru peoples. He understood the goals of Indirect Rule and the room for manoeuvre they offered. He began to speak of 'a vast country' east of Lake Chirwa, extending from Yao country to the north to Sena territory to the south, from the Nyasaland border east to the Indian Ocean. It was a country populated by 'the mighty Lomwe tribe' and its 'sub-tribes', all of whom looked to the Namuli hills, the heartland of the Lomwe people, as their ancestral home. These were arguments the administration understood.

In 1943 Bandawe founded the Alomwe Tribal Representative Association, a

group which successfully petitioned the British administration to have the word 'Anguru' banned from all official government documents and replaced with the term 'Lomwe'. One of the tactics used in Bandawe's campaign for the rehabilitation of the image of the immigrants from Mozambique was, as elsewhere in Nyasaland, an appeal to History. Not all the 'Yao' chiefs who had invaded the Shire Highlands in the 1860s, it was claimed, were actually Yao—some, like Kawinga, had been Lomwe!^[120] In this way, then, the various disparate groups of refugees and immigrants from Mozambique had come to join the ranks of those of Nyasaland's peoples—like the Tumbuka, the Ngoni and the Yao—who were to be considered a 'tribe' in their own right, even though earlier stereotypes of their alleged inferiority were to remain powerful at least into the early 1980s.

In the Southern Province, then, the elaboration of ethnicity was underscored with appeals and justifications drawn from what, largely at European prompting, was said to be history. The British sought to develop and impose a tribal system useful to them within the structures of Indirect Rule, with loyal Yao chiefs ruling over docile Nguru workers to further the successes of the European plantation economy and to maintain order. The Yao chiefs collaborated with the British, but, in so doing, they were promoting their own personal and economic power rather than any broadly conceptualized notion of Yao unity or identity. The great majority of Yao-speakers remained Muslim and hence were hostile to the establishment of the sort of Christian schools which propagated the notions of Ngoni and Tumbuka identity in the north. Similarly, Bandawe's ideas failed to gain popular acceptance because of the lack of schools under Lomwe control.

Without schools to propagate notions of ethnic identity among the young at the village level, ethnic ideologies remained weak, largely restricted to the chiefly elite and their supporters and possessing little popular force. What existed in the south, then, was a highly stratified African society, with Yao chiefs and headmen ruling over a large population of unhappy Nguru—or 'Lomwe'—whom they held in thrall by control over the land. In this situation, those who had mission education and who sought an end to colonialism were unable—and perhaps unwilling—to mobilize any popular support for their movement among Mang'anja, Yao or Nguru/Lomwe peoples by evoking ethnic symbols.

Dissatisfactions in the Central Province

In Nyasaland's Central Province the situation regarding the development of ethnic identity was wholly different from that in either the Northern or Southern Provinces. In the pre-colonial period, its small scale uxorilocal villages without strongly centralized political structures had survived the political and economic disruptions of the nineteenth century in firm possession of ancestral lands. Only in two enclaves did the Ngoni establish their presence, and to a great extent local Chewa culture speedily overwhelmed Ngoni culture.

Up to the end of World War I it was also comparatively unaffected by colonial rule, suffering neither the large scale labour migrancy that typified the north nor the oppressiveness of a government-supported planter and chiefly elite which had characterized the south. Hut tax had been imposed, but it was not collected rigorously until 1915–16, and little land was alienated to Europeans because the area was too far from transport routes to interest planters. Village life was stable and the fertile soil of the Province ensured subsistence, making it possible for most of the people to meet colonial tax demands through producing crops for market.

Unlike the Northern and Southern Provinces, it" was also remarkably homogeneous culturally, with the Chewa language spoken throughout. The

Chewa were also distinguished by an institution of remarkable resilience and vitality, the nyau societies, which served to unite the people in times of social stress and acted as powerful curbs on the influence both of missions and chiefs or headmen.^[121] Many Chewa desired Western education, but they had a well-founded fear that mission teachers would assail Chewa culture in the classroom and consequently hesitated to send their children to the mission schools.^[122] All education, moreover, was controlled by French-speaking Roman Catholics and Afrikaans-speaking members of South Africa's Dutch Reformed Church, both committed to policies that de-emphasized the use of English because of a fear that its use would encourage labour migrancy. Furthermore, they stressed the duty of moral and intellectual passivity before a body of fixed doctrine and were either hostile to or uninterested in African history and culture.^[123] Virtually no Chewa intellectuals emerged from this educational milieu to serve as culture brokers either for a progressive ethnic ideology along the lines of the north or a universalist nationalism like that explored by Chilembwe in the south.^[124]

Yet despite the continued stability of the Chewa culture and way of life, a mounting resentment against colonialism developed in the inter-war years. This resentment centred on economic issues. In 1920 two planters seeking to expand tobacco production leased 2000 acres near Lilongwe. It was the start of a new industry which grew rapidly. The tobacco grown was the fire-cured dark leaf type which had been first grown in 1916 in the Southern Province, and it quickly attracted African peasant growers. Even in the mid-1930s, when the first tobacco rush was over, only some 30,000 acres had passed into European hands compared to the million acres held by planters in the south. There was no thangata system and, though labour could be tied to planters or buyers through a variety of devices, all labour was paid. Finally, there were no Nguru immigrants from Mozambique to be used by planters or by chiefs and headmen as captive tobacco growers. Principal Headmen who wished to have their tobacco grown cheaply had to resort to using tax defaulters. The tobacco boom in the Central Province was thus a very different affair from that of the south. There was plenty of fertile land available together with large supplies of firewood for curing the leaves, and there was a population experienced in the production of indigenous tobacco and eager to exploit the new opportunities. In 1924 an official observed that there was little likelihood of Africans turning to migrant labour as they now had 'a method of earning money without having to work for somebody else, which is just what the natives were longing for'.^[125] As the Central Province was opened to road traffic, the number of African growers increased from 900 to over 33,000.

The very success of African peasant tobacco production, however, soon led to state intervention. The arrival in tobacco areas of buyers eager to purchase African grown tobacco shattered earlier monopsonistic marketing arrangements and resulted in a quick rise in producer prices. This annoyed government officials, who deeply distrusted African initiative. Listening to the complaints of the local white oligarchy, they determined that European interests should have first priority and that Africans should be 'encouraged' to work for whites: 'The education value to Natives who engage in this sort of work is great and for some time to come better results will on the whole be obtained from this work than by production by Natives working for themselves.'^[126]

Starting from this clear racial and class bias against African endeavour, the officials began controlling what they had had no part in establishing. Asserting that open competition for tobacco was bad for the African growers, they explained state intervention by declaring that they were seeking a 'rationalized'

industry by establishing so-called 'stable' prices. In 1926 a Native Tobacco Board was set up, the work of which was financed by a special tax on African-produced tobacco.^[127] The Board, on which major European planters were prominent members, was clearly intended to protect European interests from African competition, and the measures it employed were crudely straightforward.^[128]

A slump in tobacco prices in the late 1920s brought further state intervention. During the Depression years a general consensus developed throughout European colonial empires that the way out of the Depression was to force up prices by cutting commodity production. In Nyasaland, African-produced tobacco was the main target, and efforts were quickly made to 'stabilize' the industry. Officials talked urgently of the 'moral development' of Africans as well as their material development and pontificated about the need to teach growers 'a sharp lesson now and then'.^[129] It was a generally held official opinion that 'you cannot treat the native as if he were a responsible being'.^[130] So zealous was the Board that it passed regulations permitting the uprooting of growing tobacco.^[131]

The policy succeeded. In 1934 the District Commissioner at Dedza reported that the local industry was 'dying of discouragement and neglect', while the District Commissioner in Lilongwe observed about the African tobacco producer that:

one cannot help feeling that as a primary producer he has been the plaything of the rapacious middleman, and that the Native Tobacco Board has done remarkably little—beyond collecting an enormous revenue for itself—in the way of protecting him from these powerful interests.^[132]

Prices dropped again in 1937, and African growers showed their dissatisfaction by burning tobacco in protest in near-riot conditions.^[133]

Although complaints about governmental agricultural policies were also heard from Africans in the south, it was in the Central Province that such complaints were most clearly articulated. The Central Province (Universal) Native Association was founded in 1927 under the leadership of George Simeon Mwase, a Lakeside Tonga then trading at Lilongwe. The executive committee of the Association was composed for the most part of store owners and tobacco growers. Significantly, none of them were mission teachers, clerics, or government clerks. The Association, then, was far different from those elsewhere in the country. It was not preoccupied with employment opportunities for school-leavers literate in English, nor was it interested in promoting any sort of ethnic consciousness in support of limited local autonomy. They consciously saw themselves as 'universal', above 'tribe' and within the Chilembwe tradition, recapitulating matters which originally had been discussed at the Providence Industrial Mission. Significantly, Mwase became the first Malawian to write a full account of the Chilembwe Rising.^[134] They exchanged visits with the other Native Associations so that the regional distinctions began to be blurred.

Yet it would be wrong to overemphasize the forward-looking, nationally oriented petty bourgeois element in the Association, for their concerns were local concerns. Fully three-quarters of the membership were tobacco growers, and it was the concerns of the small growers and small traders of the Central Province that dominated their meetings. Thus, for instance, the Association complained about the Credit Trade with Natives Ordinance, which made African debts unrecoverable at law. They attacked the Native Foodstuffs Ordinance, which forbade trade in foodstuffs between districts by Africans. They criticized the Forests Ordinance, which regulated access to firewood needed by tobacco growers for curing their product. And they attacked the Native Tobacco Board's policies.

The various ordinances of which they complained, and especially the restrictive pricing policies of the Native Tobacco Board, were real obstacles to the accumulation of capital by Africans. The Association in effect was representing men who were trying to break loose of the restraints of the local Chewa matrilineal system by accumulating capital through their control over new sources of wealth, especially tobacco. Government policies were frustrating this control and checking the growth of a new, rurally-based bourgeoisie.^[135] Thus, in a region where there were over 33,000 tobacco growers, the Association was speaking for a substantial constituency best described as peasant producers. Apart from the absence of women from their constituency, it had claims to genuine mass support in the Central Province. Within a few years, continued unsatisfactory colonial agricultural policies gave this constituency a truly national dimension.

The Focusing of African Discontent

During the 1940s and 1950s, African discontent in the Northern Province over labour migrancy, in the Southern Province over the thangata system and Yao chiefly dominance, and in the Central Province over economic policies all came to be subsumed in a country-wide hostility to the state's agricultural policies. By the end of the 1930s, Nyasaland's agricultural experts had become convinced that there was an ecological crisis in the making. Problems of deforestation, soil erosion, and soil exhaustion loomed ever more prominent in official reports, holding out the prospect that Nyasaland might soon be unable to feed itself. There were various reasons for this problem, some national and some regional. The first was simply that population had increased to a point where in many parts of the country the carrying capacity of the soil had been exceeded.^[136] The country was, in the context of east central Africa, a relatively hospitable territory, and at the beginning of the British occupation the population was already fairly dense. With increased security and improved medical facilities, especially in the form of anti-smallpox vaccine, the population grew. To it were added tens of thousands of immigrants from Mozambique. In 1945, the census reported an average population density of 56 persons per square mile, with as many as 310 persons in the most densely populated areas.^[137]

The second reason was linked to the first and lay in the nature of the country's agricultural systems. Throughout the country, but especially in its northern half, with its dry Brachystegia woodland and relatively lower rainfall, successful subsistence cultivation depended on giving the land long periods of rest, often extending to twenty or thirty years. Without such respite, the humus quickly vanished under strong sun and leaching rains. Such systems were appropriate to the country's ecological demands, but they depended on the abundance of two things: land and labour. In the overpopulated areas, in particular in the Shire Highlands and the lower Shire Valley, there was no longer land available for such fallowing. As a consequence, the soils of most of the heavily populated districts began to decline in fertility. By contrast, in the north of the country, where land remained relatively plentiful, it was labour that was lacking. Labour migrancy had drained the region of its men since the 1890s, and the labour necessary to open fallow land for cultivation, thus ensuring that the land already under cultivation could return to fallow, was simply not available. Again, the soils of the region suffered from excessive use.^[138]

A third reason for the crisis lay in the effects of the tobacco boom itself. Soil erosion and soil exhaustion were occurring both on African land and on the European estates, where, for example, tobacco was planted on ridges descending

hillsides so that unwanted rainwater could be quickly carried away!^[139] Tobacco also required wood for curing, and deforestation had become a serious problem over wide tobacco-growing areas. Districts like Chiradzulu, for example, which older settlers could remember as having been thick with trees, were beginning to look like semi-desert. Streams that had once been perennial now flowed only in the wet season. This deforestation, together with the opening up of tobacco nursery gardens on the banks of streams, furthered soil erosion.^[140]

Finally, as causes had effects, so effects produced further causes. Land shortages and soil exhaustion, reinforced by the hostile pricing policies of the Native Tobacco Board, led to increased labour migrancy and a shift in its sources. Before World War II the Northern Province, with but 14 per cent of the country's population, contributed over 50 per cent of the total number of migrants.^[141] On the other hand, only eight per cent of adult male migrants came from the four districts where Africans produced tobacco or cotton and the two districts where European plantations held the local population through ties of thangata or where they could produce maize or tobacco for sale on Crown Land.^[142] By 1945 the situation was quite different, with only 28 per cent of total migrants now coming from the north. Thus the problems inherent in large scale migrancy came to new areas of the country.^[143]

The significance of this shift lay not merely in the fact that more labour was being removed from the villages, however. Not all the migrants left the country. Some became visiting tenants on estates in the Central Province, while others took jobs on nearby tea plantations. These workers, together with the immigrants from Mozambique who continued to supply the bulk of the workforce, had to be fed. Buyers toured the villages to purchase maize, groundnuts, and cassava. There were also substantial populations in the towns who purchased their food, most of which was supplied by peasant growers. Thus, at a time of growing labour migrancy, land shortage, soil erosion and exhaustion, the people remaining in the villages were required not only to feed themselves but also to feed an increasing proportion of the local population which did not produce its own food.^[144]

To the aggregation of these problems the government offered one major and several minor solutions. The minor solutions included the gradual purchase of unused estate land for resettlement and a legal requirement that the big tea and tobacco estates grow their own food for their workers. A 'yeoman farmers' programme was also created that involved special allocation of land, distribution of free seed, fertilizer and advice, and the payment of cash bonuses for work well done to a select few, a plan that was generally unpopular because of its perceived unfairness. And in the Central Province, where policies of the Native Tobacco Board had already alienated thousands, the government took yet another step against African producers. By the late 1940s many producers had begun to produce large quantities of maize for the market to earn the money they needed. The state reacted to this initiative as it had towards tobacco growing in the 1930s, especially after a severe famine in 1949 had underscored the fragility of the country's agricultural systems.^[145] Because agricultural experts asserted that mono-cropping maize was injurious both to soil fertility and soil structure, the state intervened in the 1950s, reducing prices paid for maize, abolishing many marketing facilities for it, and even uprooting growing maize, all in the hope of forcing a reduction in maize production.^[146] This further fuelled African discontent.

The government's major solution to the perceived ecological threat, energetically pursued throughout the country from the mid-1940s onwards, was that the villagers themselves should bring soil erosion under control by

constructing thousands of miles of contour ridges in their fields. Three facts about this decision are plain. First, it was clearly illogical to impose highly labour-intensive contour ridging in those parts of the country where the land's declining quality was caused by labour shortages. Second, it was equally misguided to expect peasant producers to take measures to increase their yields without a pricing policy to make the extra work worthwhile. Villagers well understood that land shortages were caused partially by land alienation, and they could see that the European estates were underutilized. They also were aware that the state's marketing regulations were requiring them to feed the towns and the labour compounds at prices lower than a free market would have secured.^[147]

Third, and most important of all, it was over the issue of contour ridging that truly national politics finally came to Nyasaland. To the Africans, the fact that ridging was compulsory was another example of colonial brutality. To the administrators and agricultural experts, African resistance to ridging was another example of peasant conservatism and irrationality which had to be overcome with force if necessary. No other issue—not even the political question of the creation of the Federation of the Rhodesias and Nyasaland itself—generated such united mass protest at the village level as it became mixed with the political turmoil that surrounded the creation of the new Federation. This protest provided ample grounds for the Malawi Congress Party to mobilize an anti-colonial nationalism throughout the entire country, regardless of the presence or absence of local ethnic ideologies.^[148]

Malawian Politics and the Rise of Chewa Ethnicity

In 1943, a group of anti-colonial intellectuals from the Southern and Northern Provinces formed the Nyasaland Educated African Council to press for concessions from the government that would open up avenues for African advancement. By 1944 the name had been changed to the Nyasaland African Congress, but the aims were still the same, with a great stress upon access to more education and the appropriate rewards for such education. Internal divisions and a lack of appeal at the village level, however, kept the Congress weak and ineffectual into the 1950s.^[149]

In the 1950s the twin issues of the Federation of Rhodesia and Nyasaland and the colonial government's agricultural and soil conservation policies galvanized opposition to the state at all levels of Nyasaland's society and throughout the country. The result was the creation of a coalition of interests that revivified the moribund Nyasaland African Congress. After the return of Kamuzu Banda to the country in 1958 after a long absence overseas, organizational work aimed at exploiting village-level discontent was accelerated and marshalled widespread popular support—notably in the Central Province, hitherto outside the currents of African politics in Nyasaland. In 1960 the Congress changed its name to the Malawi Congress Party (MCP).

The coalition that comprised the MCP proved strong enough to undermine the Federation and win self-government in 1964. These were the aims of all its members and the extent of their vision. Yet Congress was nonetheless a coalition of widely differing interests. Although these interests were essentially economic in nature, through the limitations of Congress's own analyses they came to be expressed largely in terms of ethnicity and regionalism. For the Livingstonia-educated Tumbuka-speakers of the north, politics was still fundamentally about the possibilities of African advancement. The language campaigns of the 1930s had been fought to protect the interests of those Tumbuka-speakers educated in

Mission schools. The strongest objection of northern intellectuals to Federation was that it blocked promotions for which their education had prepared them. With independence, the key issue for these northerners would be the rapid Africanization of the bureaucracy, the hoped-for pay-off for all the years of educational investment.

In this desire they were supported by southerners—both Mang'anja and Yao—who had acquired their education largely at the Blantyre Mission and who also coveted positions in the civil service and in commerce. For the majority of people in the south, however, politics was about an end to the hated thangata system and access to land. By the 1940s, southerners were the most uniformly poor people in the country. Those living on the estates still paid rent in kind, while those on Crown Land were, years in advance of others in the country, being driven on to the labour market by the growing shortage of land. Politics in the south looked for the freedom to extend the villages on to empty estates and for the ability to enjoy whatever benefits cash-cropping might offer.

Politics in the Central Province, where mission education had had far less impact and Chewa cultural institutions had largely endured, were, by further contrast, essentially about agricultural policy. From the days of the tobacco boom of the 1920s and the confrontations with the state about pricing and buying policies for tobacco and maize in subsequent decades, spokesmen for the local Chewa-speakers had sought to secure for Africans all the opportunities of the European planter and the Asian trader. One of the most consistent claims of the new President of Malawi, Kamuzu Banda, himself a Chewa from the Central Province, was that Africans should be allowed to grow anything and to engage freely in business.

Immediately after independence in 1964, the underlying cleavages in the MCP coalition surfaced in the so-called 'Cabinet Crisis' that occurred at the end of the year, in which President Banda found himself challenged by a group of young, well-educated cabinet ministers from the Northern and Southern Regions. One of the central issues that provoked it was President Banda's unwillingness to improve civil servants' pay or to proceed with the Africanization of the bureaucracy that the politicians who represented the educated petty bourgeoisie from both the north and the south so deeply wanted.^[150] As Rotburg has observed, the African civil servants were 'the group within Malawi that, compared to its own expectations, had benefited least by independence'.^[151]

Although the Cabinet Crisis has usually been interpreted in terms of ideological or generational conflict, the regional and ethnic dimensions were clearly evident. Comments made in Parliament by Richard Chidzanja, an important Member of Parliament from the Central Region, illustrate this point nicely. He declared the support of the Chewa people for President Banda and then went on to complain that the educated young politicians from the north and south had long despised the Chewa and their culture and had denied them a fair share of what politics had been meant to achieve.^[152] As the Crisis developed, none of the cabinet ministers who resigned or who were dismissed were Chewa. Chiefs from both the Northern and Southern Regions, however, were deposed, while three out of five district councils in the north and six out of ten in the south were dissolved. But no chief or district council was touched in the Central Region, where support for Banda was unwavering.^[153]

Part of the reason why this pattern appeared lay in the history of the Congress Party itself and the way its political coalition was built up. Tensions within it, dating from as early as the 1940s and springing from the different regional perceptions of the main goals of the struggle against Federation, had acquired a

strong congruence with ethnic distinctions. One aspect encompassed the division between the Yao and Makololo chiefly elites and the subordinated Nguru/ Lomwe and Mang'anja peoples. In Chiradzulu district, to cite but one example, the 'anti-Federation' disturbances of July and August 1953 took the form of attacks on those Yao chiefs who had manoeuvred themselves into positions of power between 1915 and 1930. Meetings called by Congress supporters successfully demanded the removal of several Yao chiefs. The meetings ended in violence, and one chiefs house was burned down. The official enquiry, in true post-Chilembwe tradition, blamed 'immigrants from Mozambique' for the disturbance and pointed to Congress's involvement. Congress had consciously become the supporter of the oppressed Nguru/Lomwe in this ethnically defined local battle as part of its strategy to build an effective mass base.^[154] Not surprisingly, the Nguru/Lomwe became among the firmest supporters of Congress, which demanded an end to thangata, transferring this loyalty to President Banda after the Cabinet Crisis. In the Census of 1966, therefore, after years in which many Nguru had passed themselves off as 'Yao', large numbers of them identified themselves as 'Chewa', cementing thereby their alliance with the victorious President Banda.^[155] This strain of anti-chief, anti-Yao sentiment eventually encompassed such popular figures as Henry Chipembere, an educated Yao cabinet minister, at the time of the Cabinet Crisis.

Similarly, in the lower Shire area, President Banda, in his pursuit of a 'Chewa' political base for himself, was able to build on anti-Makololo sentiment that was present because Makololo chiefs had been prominent in enforcing contour ridging during the period of Federation. Reinstating Paramount Chief Lundu a century after the last Lundu had been dispossessed by the Makololo, Dr Banda made one his most blatantly ethnic speeches:

And I am happy that because of my harping on the fact that all the people here are, in fact, Chewa, not Mang'anja, the people themselves have realised and admitted the truth, this pleases me.

I am happy because this is why I have done this [resurrected the paramountcy], because the people themselves have recognised the truth, have admitted the truth they are, in fact, Achewa, although for the past one or two hundred years they have been calling themselves Anyanja or Amang'anja.^[156]

By such alliance-building among people who had been subordinate to the Yao and Makololo political elites, President Banda was able to lump togther the various dialect groups of the Southern Region—Chipeta, Nyanja, Mag'anja, even 'Lomwe'—to produce a national population that was, at least on paper, more than 50 per cent 'Chewa'.^[157]

The reasons why the Cabinet Crisis took on an anti-northern aspect were rather different, for northern leaders had long been at the very heart of the nationalist struggle and had supplied many of the movement's foremost intellectuals. The difficulty with the northerners lay in the conflict between their education and President Banda's vision of the nature of the future Malawi. As the planters' spokesman, R.S. Hynde, had expressed it to the commissioners enquiring into the Chilembwe Rising in 1915, 'This country is agricultural and there is no room for the highly educated native.'^[158] If it seems excessive to suggest that President Banda, so far from being a reincarnated John Chilembwe, is closer to being a reincarnated Governor Sir George Smith, the fact nonetheless remains that the philosophy and the practice of government evolved by the British in response to the Chilembwe Rising show marked similarities to those espoused by President Banda himself.

Both the colonial rulers and the post-colonial ruler have assumed that paternalism—the landowner dealing with his peasants, the chief with his subjects, the master with his servants, the President with his people—constitutes the form of government best suited both to the economy and the general temperament of Malawians. President Banda, in the process of acquiring a plethora of plantations for himself, appropriated the planters' ideology as well.^[159] Bizarrely, but entirely logically, President Banda made speeches boasting that the clinics and schools on his own estates were superior to those available from the state. In this situation, the aspirations of northern and southern intellectuals for African advancement on the basis of education and merit posed a double threat in 1964: the threat of a bureaucracy dominated by those who were not members of President Banda's own broadly defined 'Chewa' political base, and the threat of the evolution and entrenchment of an educated middle class that might challenge the personalized patron-client relationship of the President with 'his' people.

President Banda turned to patronage to cement ties with his loyal political base. The most profitable segment of the agricultural economy at independence was the European tea and tobacco estate sector. After independence, the state began acquiring European tobacco estates and reselling them to members of the petty bourgeoisie who had remained loyal to Banda. Necessary loans came from the state's agricultural marketing board, which in effect transferred capital obtained from the peasantry to the new bourgeoisie through its control of prices paid for peasant-produced commodities.^[160] Furthermore, the Malawi Land Act of 1965 gave the President the power to grant private estates on leasehold over huge areas of the country, including even areas developed for peasant agriculture under the aegis of international agencies.^[161] Through these two means, President Banda was able to consolidate his ties of patronage with his chosen political allies and, by retaining the power to revoke leases, could at the same time ensure that no independent bourgeoisie would emerge to challenge him.^[162]

This dispensation of specific patronage to loyal followers was paralleled by the extension of general patronage to the most loyal area of the country, the Chewa-dominated Central Region. Many government institutions have been transferred from the Southern Region to the Central Region, most notably the capital itself. While the Central Region obtained only 11 per cent of development funds expended in 1967, it received 40 per cent by 1972–3.^[163] Agricultural loans are largely restricted to farmers in the Central Region, and it was this area which marketed crops that provided fully 86 per cent of the surplus that the Farmers Marketing Board made in 1971–2.^[164] Since the early 1970s governmental preference for investment in the fertile Central Region has been consistent.

President Banda, however, was himself sufficiently a culture broker to realize that something more was needed. In the aftermath of the Cabinet Crisis he moved swiftly to seek in Chewa institutions the basis of a new Chewa ethnic ideology which was to be projected on a national scale through state-controlled schools. This approach came naturally to Banda, who had been self-consciously a Chewa throughout his life. While studying in Chicago in the 1930s, he had been 'a very excellent informant' to Mark Hanna Watkins, the first grammarian of the language, and to this day Banda retains a fanatical devotion to the language and to what he conceives as its purity, ardently attempting to stop any adoptions from English into the language.^[165] In addition to his abiding interest in Chewa language, Banda has long been interested in Chewa culture, and, with Cullen Young, he wrote Our African Way of Life in 1946.

As culture broker for the Chewa, Banda had a broader vision, however, than formulating an ideological statement for his ethnic group alone. He has instead

equated 'Malawian-ness' with Chewa-ness, and he has depicted the Chewa as the very soul of the country, often going as far as maintaining that many Yao and Lomwe people were actually Chewa people who did not realize it. During the years he was in Nyasaland after 1958 campaigning against the Federation, he frequently denounced signs of 'tribalism' in others, yet he also frequently emphasized that he was himself a Chewa.^[166] In his speeches he has always been at pains to paint a picture of a glorious Chewa past. Using the alleged territorial extent of the ancient Maravi state of the sixteenth and seventeenth centuries, for example, he has frequently asserted publicly that large parts of eastern Zambia and northern Mozambique should be under his rule.^[167]

In short, Banda emphasized uniquely Chewa cultural attributes, not some sort of secular Malawian 'nationalism', and in his preoccupation with language, history, and culture, he was very much a latter-day Chewa version of Edward Manda and Yesaya Chibambo, like him, graduates of the Livingstonia Mission. Long defunct chiefdoms were resurrected. 'Authentic' clan names were revived. The 'correct' form of the Chewa language—the dialect understood by Banda—was stressed and the paramilitary Young Pioneers, made up of failed school-leavers, drilled it by rote into the school-children of non-Chewa areas.^[168] The Chewa nyau societies, and especially their dances, had been long despised by the educated as a symbol of Chewa backwardness in a modern world. After the nyau societies demonstrated their loyalty to President Banda in the Cabinet Crisis by actively intimidating his opponents with physical threats, they became perhaps the heart of what is depicted as a national culture, even though they are the hallmark of Chewa culture and only that.^[169]

The transformation of the nyau from the sign of backwardness to the symbol of authenticity was achieved with the aid of expatriate Africanist scholars, analogues of the Tumbuka's Cullen Young and the Ngoni's Margaret Read. Matthew Schoffeleers and Ian Linden, in their work, found great favour among Chewa-speaking intellectuals.^[170] Schoffeleers was especially important because of his depiction of nyau as not only being at the very centre of Chewa art and culture, but, even more importantly, at the very root of Chewa resistance to colonialism and western cultural imperialism. With his timely writings, Schoffeleers became the source of a usable past for the developing ideology of the present.

By the same token, work at the University of Malawi on the history of the Yao has been highly revisionistic, concerned with demolishing the myths of Yao nationhood established in the inter-war period and with providing a history in terms of which the Yao are shown to have been defeated, scattered and politically disunited peoples with but limited territorial claims. This work has also stressed that many 'Yao' and 'Lomwe' people really belong to the Chewa group.^[171] The message from the late 1960s into the 1980s was clear: the Chewa people and Chewa culture was the core of modern Malawi by right of being most ancient and least compromised by colonialism, and Malawi culture would be considered synonymous with Chewa culture.

But it was not enough to dispense patronage to supporters and develop a new ideology of Chewa ethnicity. Government policies also attacked the vested interests of the northern petty bourgeoisie. Because of their long history of having access to an educational system far superior to that existing elsewhere in the country, northerners had come to be 'over-represented' in important positions at independence and in the expansionist years that followed. At the University of Malawi, the seedbed for new technocrats, a disproportionately high number of places went to northerners.^[172] In the civil service in 1969, out of the 113 highest-level Malawian civil servants, the Northern Region, with but 12 per cent of the

population, held over 50 per cent of the places, while southerners comprised most of the rest.^[173] The established influence of educated northerners and southerners had to be lessened directly.

The first, appropriately symbolic signal of this attack upon northern influence occurred in mid-1968, when Tumbuka, the symbol of northern regionalism since the early 1930s, was abolished as an official language and Chewa was made the sole national language. No longer could Tumbuka be used in the press or on the radio, a situation which resulted in bitter resentment throughout the Northern Region, a resentment made worse by Chewa-speakers' triumphal assertions that other peoples of the country were cultureless because 'they had no language'. Soon afterwards, the Parliamentary Secretary for Education announced that all school-children who failed their required examination in the now-required courses in Chewa would have to re-sit all their examinations.^[174] Soon afterwards, the establishment of the Malawi Examinations Board to replace the Cambridge Overseas Examination was followed by a change in examination grading policy which required both northerners and southerners to obtain considerably higher grades in their school-leaving examinations than those in the Central Region if they were to qualify for places in the secondary schools. While economic opportunities were channelled to the people of the Central Region, then, school fees throughout the country were raised considerably, making access to education in the north and the south comparatively more difficult. The University of Malawi, the source of future bureaucrats and teachers, was systematically purged of its non-Chewa administrators and faculty in the early 1970s as part of an attempt to make it a secure seat for the elaboration of a Chewa ethos by loyal Chewa-speakers. Finally, to remove non-Chewa officers from the civil service, a mandatory retirement age of fifty was imposed and large numbers of northerners and southerners thereby retired. Non-Chewa-speaking northerners and southerners were also removed from other positions of authority through widespread and arbitrary detentions, especially between 1973 and 1976.^[175]

Such measures are defended within Malawi as attempts to repair regional discrimination that occurred during the colonial period, through which northerners were given unfair advantages. The fact that they are equally directed against the south disproves this point. The south, as has been seen, and especially those parts most afflicted by the thangata system and by over-population, suffered equal or greater deprivation in colonial times. The official marketing bureaucracy followed pricing policies throughout the 1970s and early 1980s that were barely distinguishable from those of the former Native Tobacco Board and this fact, coupled with the curtailing of opportunities for raising capital through labour migrancy abroad, locked southerners firmly into continuing abject poverty. School fees are beyond the means of most villagers, and they have abandoned the belief in education as a route from poverty. The intensifying land shortage, to which the abolition of thangata produced only a temporary solution, is driving more and more young men to work as migrants on the estates and plantations of the Central and Northern Regions, often at rates of pay below the official minimum wage, and they are abandoning behind them their wives and children.^[176] Others find a career in the army, which, as part of the alliance with Dr Banda, continues to be Lomwe-dominated.

The plain fact is that, however inappropriate their dream of solving Malawi's problems through control by an educated bureaucracy might have been, the general clamp-down on non-Chewa intellectuals in independent Malawi has been profoundly damaging. There has been no serious public discussion in President Banda's Malawi of the problems the country has encountered since independence:

the impact of the oil crisis; South Africa's policy of regional destabilization; the collapse of the country's economy in the late 1970s and its takeover by the International Monetary Fund; or the exceedingly high rate of population growth in a country of limited resources. The Cabinet Crisis, despite its central significance to Malawi's history and despite the fact that it occurred well over twenty years ago, remains a wholly embargoed topic. The names of those who contributed to the rise of Malawian nationalism in the 1950s can be mentioned only in secret. During the show trial of Orton and Vera Chirwa in 1984, when Vera Chirwa began her testimony with the statement, 'When I founded the Malawi Women's League . . .', a tremor of excitement ran through the spectators. The simplest historical fact has become subversive.

In this closed atmosphere, rumours of ethnic conspiracies abound. Such rumours serve as explanations for people who lack explanations: they arise when an intellectually closed society turns in on itself in search of scapegoats. The language of this discourse in rumour largely takes the form of crude ethnic stereotypes that derive from the colonial period's experience of uneven development. With meaningful analysis by Malawians of the problems that face Malawi remaining proscribed, there is the real danger that the ethnic explanations that are now current will be the only ones available for future discussion, a legacy of the past that will increase the likelihood of communal violence in the country at times of political transition.