4—
Missionaries, Migrants and the Manyika: The Invention of Ethnicity in Zimbabwe

Terence Ranger

Introduction: Politics Through a Tribal Lens

Over the last twenty years there have been all too many conflicts in Zimbabwean African politics—conflicts between and within African parties and guerrilla movements, divisions between voters, strains within the cabinet and government of Zimbabwe. There have also arisen a number of schools of interpretation of such divisions. Some see them in terms of class conflict; others, however, see them in terms of ethnicity. They invoke not only the allegedly 'traditional' hostility between the 'Ndebele' and the 'Shona', but also an asserted conflict between Shona sub-ethnicities, the so-called 'Korekore', 'Zezuru', 'Karanga', 'Kalanga', and 'Manyika'.

Thus in March 1976 an international commission of inquiry, appointed by the Zambian government and composed of representatives from a dozen or so African states, reported on the murder of Herbert Chitepo, then National Chairman of the Zimbabwe African National Union (ZANU), in March 1975. The report found that Chitepo had not been killed by a Rhodesian agent. He had instead been a victim of the 'mutual hatred and suspicion among the tribal groups' within ZANU. The commission found 'abundant evidence of tribal and regional manifestations in ZANU'. In particular, it found that Chitepo's death was the climax of a struggle for power between the Manyika and the Karanga. The victorious Karanga, now supreme in the party's command, had eliminated the Manyika Chitepo.^[1]

In May 1976 Ndabaningi Sithole, who was then struggling to regain the leadership of ZANU, endorsed these findings. In an open letter, calling on all Zimbabweans 'to do away with tribalism and regionalism', he analyzed the successive elections to ZANU's central committee in terms of tribal rivalry:

The main thesis of my letter is that ZANU as we had first formed it became constantly subject to a process of tribalization or regionalization so that it completely lost the nationalist perspective . . . When we formed ZANU in 1963 it was called the Zimbabwe National Union, but by 1974 and at the beginning of 1975 it had become in practice Zimbabwe African Tribal Union . . . If the death of Herbert Chitepo is to be associated with any 'ism', it cannot be directly or immediately with colonialism or capitalism, but rather with tribalism or regionalism . . . If it is to be associated with any race, it can

only be the African race . . . I want everyone to know that this tribalism did not originate from the people at home but from the people outside Zimbabwe. The Karanga, Zezuru, Korekore, Ndau, Ndebele, Kalanga and other tribes in Zimbabwe are solidly united and determined to become a nation.^[2]

In March 1977, by which time Sithole had set up his own splinter ZANU group, he sought to carry this analysis into practice:

In appointing the central committee, I have been guided by definite objective criteria . . . [but] in the allocation of the various offices I have been guided by the principle of ethnic balance, in order that we may forestall any ethnic manipulation of any of our political structures. In ZANU we have had tragic experience resulting from ethnic imbalances in the central committee. The allocative principle of ethnic balance will be applied to the national executive committee, the central committee, the provincial committee, the district committee, the branch executive, the high command, the general staff, all committees and to any other organs of ZANU.^[3]

Sithole appointed two Ndau, two Karanga, two Zezuru, two Ndebele, two Manyika and two Korekore to his executive. Political scientists also began to adopt this style of analysis. In 1979 Ndabaningi Sithole's brother, Masipula, published his Zimbabwe: Struggles Within the Struggle . With the aid of ethnic triangle diagrams, he argued that prior to 1971 ZANU's party leadership was shared between Zezuru, Karanga and Manyika, an arrangement that had prevented straightforward 'tribal' hostility:

Competition, whether perceived in personal, tribal or regional terms, was in three directions. An individual from the Manyika group could not only rely on regional/tribal support. . . . One had to appeal at least to one other tribal/regional grouping.

After 1971, however, 'the Zezuru group pulled out', and 'bi-polarity sharpened tribal/regional competition in the party'. A deadly struggle between the Manyika and Karanga developed.^[4]

The analysis spread to British commentators. In July 1979, for example, Professor Claire Palley gave an ethnic interpretation of the recent Internal Settlement elections which had been won by the 'Manyika' Muzorewa:

If the tribal basis of [Muzorewa's] support is analyzed, it will be seen that by the time of the election, his party (the UANC) largely consisted of a coalition between the Manyika, Korekore and Zezuru tribes, totalling 43% of the population. The UANC received only about 40% of the total potential vote. There is here a remarkable tribal correlation. Since Mr Chikerema led Zezuru dissidents out of the bishop's party, much of the bishop's Zezuru support has been lost. The bishop is now left with basically 25% of the African population supporting him. The Ndebele speakers support Mr Nkomo . . . The Ndau and some Rozwi support the Rev. Sithole. The Karanga by and large support Mr Mugabe.^[5]

The British press and others began to reproduce and refer to maps produced by the Southern Rhodesian government which purported to show the exact boundaries of these sub-ethnicities and the exact proportion of the country's African population which each contained. Such maps often carried beneath a note to the effect that 'the above divisions are based on historical fact'. Even the Minority Rights Group, a body highly critical of the Rhodesian government and

suspicious of its propaganda, devoted a page of its Inequalities in Zimbabwe, published in December 1979, to the 'tribal background', in which it informed its readers that the Korekore constitute 12 per cent, the Zezuru 18 per cent and the Karanga 22 per cent of the African population, and that tribal rivalry 'looks set to grow'.^[6]

Analyses such as these raise two main questions in a historian's mind. The first question, to which I shall return briefly at the end of this essay, is whether they provide an accurate explanation for recent conflicts. The second question, to which most space will be devoted, is from where the idea of such entities as the 'Manyika', the 'Zezuru', and the rest has come. These entities certainly do not represent pre-colonial 'historical fact', nor can they in the present be properly described as 'tribes' or 'clans', no matter that both African and European commentators employ these terms. Yet they evidently have come to possess a subjective reality in the minds not only of commentators but of participants. How has this come to pass?

Ethnicity and History

Carolyn Hamilton and John Wright have recently urged that we seek the roots of twentieth century ethnic divisions in the pre-colonial past:

The processes by which rulers and ruled in a specific region of an emerging pre-capitalist state manipulated, or sought to manipulate, concepts of group identity in pursuit of what they perceived as their particular material interests. . . . Our argument that the emergence of pre-capitalist states in south-east Africa was grounded in the development of ethnic categorization would lead us to dispute the common view that (in the words of Colin Leys) '. . . modern tribalism is a creation of colonialism. It has little or nothing to do with pre-capitalist relations between tribes'. More likely, we would argue, is that modern 'tribalism' is a creation of the impact of colonialism on forms of ethnic consciousness whose roots lie deep in the pre-colonial past.^[7]

However convincing this proposition may be for the Zulu state, it does not apply in the case of Zimbabwe.

David Beach's admirable history of the Shona-speaking peoples shows clearly that before 1890 there were two 'historical realities'. One was that all speakers of Shona possessed not only a language but also many other cultural traits in common. Scattered over a large area, in contrasting environments, and pulled in different directions by trading links and military alliances, however, these Shona-speakers were not conscious of a cultural identity, still less a political one. The second reality, and the one of which Shona-speakers were conscious, was the local chieftaincy group. These numerous groups were not and never had been clustered together in self-conscious ethnicities such as are today implied by the terms 'Manyika', 'Zezuru' and the rest. At times in the past, powerful states had emerged which had exacted tribute from chiefly groups in a system of over-rule. But these states had never pulled all their subjects together into self-conscious identities, nor had they manipulated concepts of group identity in a manner which left a lasting ethnic legacy. Between the Shona culture as a whole and the local chiefly group there existed no intermediate concept of ethnicity.

Beach does show, however, that the terms Korekore, Zezuru, Manyika and the rest did have a pre-colonial currency. Each arose in a different way and had different connotations and each was available to be pressed into distorting service by the classifiers of the twentieth century. Two of the terms had long significance,

two were topographical, one was a slang term invented by an enemy.

The terms which had a long recorded history were 'Karanga/Kalanga' and 'Manyika'. When the Portuguese came into contact with the Shona-speaking peoples in the sixteenth century, they recorded that the chiefly lineages which ruled over the commoners were known as 'Karanga'. They also reported the existence of a chiefly territory which was called 'Manyika'. European usage came to transform the significance of these terms. As we shall see, the Portuguese called a large region around the Manyika chieftaincy by the name of Manicaland, and the name was picked up by the British in the late nineteenth century. Most of the peoples of this region, however, did not think of themselves as related in any way to the Manyika chieftaincy. As for the term 'Karanga', it suffered a shift both of location and of meaning. The Portuguese had used it to refer to the ruling lineages of the northern and eastern Shona-speakers. The incoming British at the end of the nineteenth century picked up this 'historic' term to describe the first Shona-speakers they encountered, naming the total populations of the southwest area 'Kalanga' and those of the southern plateau 'Karanga'.

The terms which had a topographical connection were 'Korekore' and 'Zezuru'. Beach tells us that 'Korekore gradually appeared in the north. . . . It generally meant the people of the north and northwest.' He also tells us that 'by the middle of the eighteenth century the Portuguese were beginning to refer to the people around the head of the Mazoe Valley as "Zezuru". The term meant "people who live in a high area".  The two words were, then, the equivalent of 'northerner' and 'highlander' rather than ethnic or tribal categorizations. As for the term 'Ndau', it was a derogatory nick-name given to the peoples of the eastern frontier by the raiding Gaza Nguni of the mid-nineteenth century. Beach concludes by regretting the projection backwards into 'tradition' of what have become modern 'tribal' names since, as he says, 'most of these terms were originally used in a much more restricted sense'.^[8]

In this way terms which certainly did not mean to convey the idea of ethnic homogeneity in pre-colonial times were picked up in the colonial period precisely to convey that idea. Even then they did not convey the idea very successfully to rigorous observers. Thus the anthropologist Hans Holleman, writing of a number of Shona chiefships, remarked:

All these are commonly regarded as belonging to the Zezuru cluster of the 'Shona' tribes, but it is doubtful if such a classification can be justified on ethnological grounds as no detailed comparative study has yet been attempted. From a native point of view this affiliation is meaningless, as it is not supported by any special ties of a political or other nature. Few, if any, have an intelligent conception of a 'Zezuru cluster' as distinct from, say, a Karanga or Manyika cluster.^[9]

The terms have no 'traditional' validity, then, nor did they correspond to any clearly perceived twentieth century 'tribal' realities. The administrative units created by the colonial government were territorially defined districts which bore no intended nor actual relationship to 'tribal' or 'ethnic' identities. Later, when the Rhodesian government began to take the idea of 'tribal politics' more seriously, it demarcated and recognized literally hundreds of 'tribes', each under its own chief. Thus, there was no incentive for Africans to invent regional ethnicities in the hope of tapping the flow of administrative patronage.

And yet, despite all this, the terms have come to possess at least that degree of reality illustrated by this chapter's opening quotations. I believe that this has happened as a result of the agency of both 'unofficial' Europeans and of

'unofficial' Africans—of missionaries and their converts and of African labour migrants. Later, when these 'unofficials' had achieved a diffused sense of 'Manyikahood', 'Zezuruness' and so on, the concepts were belatedly taken up by 'officials' and by chiefs, by the administration and by the propaganda agencies of the Rhodesian regime. I shall illustrate this generalization through the presentation of a case study of twentieth century ethnicity in parts of eastern Zimbabwe. In particular, I shall examine the changing sense of self-identification in the old kingdom of Maungwe under Chief Makoni, an area in the twentieth century constituting the greater part of the Makoni administrative district. In the 1890s no one in Makoni thought of themselves as 'Manyika'; by the 1930s most of them had come to accept that they were members of a wide Manyika identity. I shall try now to explain how this happened.

Ethnicity, History and the Early Colonial Administration

Before I can turn to the task of tracing the growth of an extended Manyika identity in the twentieth century, however, I must deal with at least three other uses of the concept 'Manyika'. The first three of these I derive from Dr H.H.K. Bhila's Trade and Politics in a Shona Kingdom: The Manyika and their Portuguese and African Neighbours, 1575–1902 .^[10] For Bhila the essential historical meaning of 'Manyika' is the territory and people of chief, or king, Mutasa. 'My research was focused on the area where the Mutasa dynasty has settled since the late seventeenth century', Bhila writes, and he shows this area on a map as a relatively small zone, north and northwest of modern Mutare, comparable in size to one of the later Rhodesian administrative districts and bounded by other African 'kingdoms' which were not in any sense parts of 'Manyika' so defined. Bordering Mutasa's kingdom to the west lay the kingdom of Maungwe under Makoni, which in the nineteenth century was Mutasa's main enemy and competitor for land, cattle, women and slaves. As Bhila explains, 'the boundaries of these kingdoms were often shifted following the vicissitudes of wars'; hence the best definition of traditional Manyika was political, rather than geographical. Manyika comprehended all those who at any one time ackowledged the authority of Mutasa—and nobody else.

A second use of the concept 'Manyika', however, was developed by the Portuguese and propagated by them especially in the later nineteenth century. Claiming that the then reigning Mutasa had made a 'voluntary submission' to them in 1876, they expanded the area of 'Manyika' on their maps to cover an enormous territory to which they laid claim. As Bhila writes:

On a Portuguese map of 1887 . . . its boundaries extended along the Zambezi from Shupanga to near Tete, then south-west along the Mazoe and south by the Sabi river valley to its junction with the Odzi river, then east along the Musapa and Buzi rivers to the mouth of the Pungwe. This enormous size of Manyika was evidently fixed by political and commercial considerations. The Mazoe river valley was included because of rumours of abundant alluvial gold. The Kingdom of Manyika over which the Manyika rulers . . . exercised authority . . . was a much smaller area.^[11]

This greatly inflated Portuguese Manyika did include, among much else, the territory of Maungwe but, as a merely notional and paper definition, it did nothing to affect sense of identity. In other moments, moreover, the Portuguese treated the Makoni chiefs of Maungwe as independent sovereigns and made treaties with them.

A third use of the concept 'Manyika' was that made by the British as a counter to these Portuguese claims. In their attempt to gain control of 'the Pungwe River route, which was the main water way to and from Beira', the British South Africa Company imposed 'a treaty on Mutasa on 14 September 1890'. This treaty 'provided that no one could possess land in Manyika except with the consent of the BSA Company', and once it was signed, the Company invented its own 'Greater Manyika', the western boundaries of which lay deep inside Portuguese territory, though areas like Mazoe and Maungwe, to which the Company made quite differently based claims, were excluded.^[12]

Once the Company's frontiers had been fixed by means of war and arbitration, there was no longer any need to inflate the power and territory of Mutasa; rather the reverse. The old kingdom of Manyika was broken up between the two administrative districts of Umtali and Inyanga; much of its land was alienated to white farmers; and the administration was very concerned to advance a minimalist definition of Manyika-hood. 'Umtassa's country and people are called Manyika', wrote the Native Commissioner, Umtali, in January 1904. "They do not speak the same dialect as the other Mashonas.'^[13] The same desire to separate Mutasa off from neighbouring peoples can be seen in the early district reports from Umtali in which Native Commissioner Hulley spelt out that the three chiefs in the district, Mutasa, Maranke and Zimunya, were of quite distinct origins, even if there was a popular tendency to refer to his district as 'Manicaland'.

So far as the administrative district of Makoni was concerned, the Native Department was concerned to emphasize the distinction between its people and the Manyika. In 1910 there was a boundary dispute between the Native Commissioners of Makoni and Inyanga districts. The Native Commissioner, Inyanga, wrote to the Superintendent of Natives, Umtali, to explain why he was collecting tax from Africans on farms which lay just within the western border of Makoni district:

There are no Makoni (Shonga) natives on any of these farms. I have always acted on your suggestion—that is I have dealt with Manyikas only . . . [Let] the Native Commissioner Rusapi deal with Makoni natives and I with Manyika. . . . No dispute should arise.^[14]

The matter was so decided and the Chief Native Commissioner determined that 'the N.C., Inyanga deal with all Manyika natives and the N.C., Rusapi with all the Makoni'.^[15]

It is clear from this that the Native Department firmly separated the Ungwe of Makoni from the Manyika in a political sense. The separation was also insisted upon in the cultural sphere. Thus, in 1915 a debate arose within the Native Department about the significance of the term mayiaini in relation to Manyika marriage customs. Llewellyn Meredith, who had been Native Commissioner in both Melsetter and Makoni districts which today are held to fall within 'Manicaland' and whose inhabitants are included in the percentage of the population allocated to the 'Manyika', ventured to advance his opinion about 'Manica customs and language'. He was crushed by the scorn of the Manyika specialists. The Superintendent of Natives, Umtali, mocked Meredith's ' 18 years experience of Manyika customs gathered in other districts', and invoked the authority of Archdeacon Etheridge, the leading missionary expert on Mutasa's chiefdom. 'I do not of course know,' wrote Etheridge, 'what word may be used in Chindau, or Chirungwe, the dialects spoken in Melsetter and Rusape [Makoni] districts, but as regards Chimanyika there is no question at all.'^[16]

In short, nothing in the pre-colonial history of eastern Zimbabwe had

predisposed the people of Makoni to think of themselves as 'Manyika', and the colonial administration—by means of its decisions on district boundaries and by its issuance of registration certificates which included an official allocation of 'tribal' status—actively told them that they were definitely not Manyika. Posselt's Survey of the Native Tribes of Southern Rhodesia, which was issued by the Government Printer, showed the Ungwe of Makoni as one of the sixteen 'separate tribes' of Southern Rhodesia, although his map of the 'Approximate Distribution of Tribes and Languages' showed them as part of the Zezuru language zone. Posselt's classifications were confused, but the one thing of which he was certain was that the people of Makoni were tribally, culturally and linguistically distinct from the people Mutasa's Manyika.^[17]

At the beginning of the colonial period the people of Makoni themselves certainly also believed that they were distinct from the Manyika. The unity of the pre-colonial Ungwe had been, in fact, a unity of political allegiance rather than a unity of a shared ethnicity. Within the Ungwe area there lived not only the related chiefly dynasties of Makoni and Chipunza, but also headmen and their followers who were the acknowledged descendants of the original occupants of the land. During the nineteenth century, and no doubt long before, small groups from the north, east, and west entered the Ungwe area as hunters, specialists, refugees and adventurers and were allowed to settle there in return for their services as smiths, elephant-hunters, diviners, rain-callers or fighters. Many of these incomers were used by the Makoni chiefs in their wars with the Mutasas of Manyika and in hostility to Manyika they forged their new political identity.

Maurice Nyagumbo's unpublished account of the arrival in Maungwe of his grandfather, Nyagumbo from 'the Nyashanu country', who had fled with his two brothers as a political refugee, is very revealing of this process:

When the three brothers and their sister arrived in Makoni's country they settled in a cave. . . . Here they were found by Makoni soldiers who suspected them of being spies. They were taken to the chief, where they were kindly received . . . Chief Makoni had become a terror to all those who surrounded his chiefdom, especially the Wamanyika of chief Mutasa. . . . It was at this time that the Nyashanu brothers and their sister arrived in Makoni's country. The sister later became Makoni's wife; the brothers joined one of the regiments of the chief. After a few raids, Nyagumbo, the youngest of the Nyashanu brothers, proved to be an excellent fighter, and he was later promoted to lead his own regiment.

Nyagumbo's regiment had a special task to perform. This was to scout and waylay enemy soldiers. After a few years as a soldier, Nyagumbo became one of the main councillors of the chief and as a reward for his outstanding achievements in the raids, he was given a slave girl of the Wamanyika tribe as his wife. . . . At that time Mutasa of the Wamanyika prepared to attack Makoni with a large army. . . . Makoni had been informed of these preparations and since he had only a limited number of soldiers he decided to move his army to a fortified hill at Mhanda. Makoni had large stores of food for his army on the hill. . . . When Mutasa arrived with his unwieldy army he found the villages deserted. . . . Then very early in the morning, as the war-drum beat and the women ululated thunderously, Makoni's army descended to meet Mutasa's army. This historic battle started early in the morning and was fought throughout the day. . . . Makoni's army, although suffering heavy losses, inflicted even greater losses on the enemy. Mutasa fled with his few remaining soldiers. . . . This was known as Pakafa dende remukaka by the Wawunge people—the calabash containing milk was smashed—because both Mutasa

and Makoni lost their most important warriors. The two Nyashanu brothers who fought in the battle were both wounded but they survived.^[18]

Memories of these wars remained vivid in the early twentieth century, reminding the people of Makoni of their distinction from the Manyika. In December 1902, Archdeacon Upcher, itinerating around kraal schools in Makoni, 'passed a wooded kopje, the scene of a massacre as horrible as that of Glencoe. M'tasa surrounded the village and set fire to the houses. . . . People do not like to pass that way by night as they hear, so they say, people talking'.^[19] Ungwe women, captured by the Manyika during the wars, took the opportunity of the colonial 'pacification' to escape from their masters and to return home. When, in October 1898, the Anglican missionaries, Douglas and Vera Pelly, travelled to the farm of their Zulu catechist in Makoni, they tried to persuade Mutwa, widow of the martyred Bernard Mkezi, to come back with them to Umtali so that her little son could 'be brought up as a Christian for Bernard's sake'. Mutwa's response revealed all the local dislike of and sense of distance from the territory of the recent enemy—'She said that she did not want to come and live with us as she did not like the Manyika country.'^[20] And yet, by the 1930s, very many people in Makoni were defining themselves as 'Manyika'. It is precisely in the context of the missions that we must begin to search for an explanation.

Language and Ethnicity: the American Methodist Episcopal Church

There has recently been a good deal of attention given to the importance of language in defining ethnicity—what Isabel Hofmeyr has called 'Building a nation from words'.^[21] Increasing attention has also been given to the key role played by missionaries in specifying African languages.^[22] Three mission churches dominated the Christian history of Makoni and Umtali districts—the Anglicans, the Trappist/Mariannhill Catholic fathers, and the American Methodist Episcopal Church (AMEC). All three developed a particular vested interest in these eastern territories. The Anglicans developed their main Southern Rhodesian intellectual and educational centre at St Augustine's near Umtali, and they had by far their greatest early evangelical successes in Umtali and Makoni districts. The American Methodists radiated out from their main station at Old Umtali: unlike the Anglicans they did not have missionary bases in any other parts of the country. Their first and deepest penetrations were in Umtali and Makoni districts where they, too, scored considerable evangelical success. Although Catholicism aspired to and increasingly achieved a territory-wide presence, the Trappist/Mariannhill mission represented an odd enclave in an otherwise Jesuit domain. The Mariannhill fathers operated out of their main mission at Triashill on the borders of Inyanga and Makoni, and, from there, they spread into Inyanga, Makoni and Umtali districts. Beyond these they could not go without invading the Jesuits' sphere of influence. It was not surprising, then, that all three of these churches came to develop a strong doctrine regarding alleged special qualities of the people of these eastern districts—of what they all came to regard as 'Manyikaland'.

This doctrine was founded intellectually on the language work of the three churches. Missionary language work was of central interest and importance, both to the missionaries and to the Africans among whom they worked. Oral communication and rhetorical skills generated amongst many Africans a great

linguistic enthusiasm as well as a fascination with linguistic variations and with linguistic techniques. The missionaries possessed the linguistic skills and tools to give the people what they were interested in and curious about. The missionaries, of course, began with the attempt to understand the spoken vernacular but it was their intention to develop a written form of the language as soon as they could.

The American Methodists made their linguistic assumptions explicit. They knew that they must acquire the vernacular and they knew that to do so they had to learn from their first converts. On the other hand, they believed that Africans were longing for something which only they could give—namely, the reduction of the language into writing, with a formal orthography and regular grammar. At first, they concentrated especially on giving God's word—'We never forget', wrote Helen Springer in 1905, 'that the primary object object of our work here is to give the native the Bible and enable him to read it.'^[23] Soon, however, they broadened their ambition. 'We need literature—TRANSLATIONS—for these people,' wrote J.R. Gates in 1911, 'they cannot grow without it. They are dependent upon it as every other race has been dependent upon it.'^[24] 'We ourselves have hundreds of devotional and other volumes on our shelves,' wrote H.I. James, 'while they have almost nothing.' 'The situation with us as far as vernacular literature is concerned is nothing short of appalling', added Gates. 'What a supply other peoples of the world have of national life, of literature, of history. . . . But all these are absent in Africa.'^[25]

These ideas attained their most impassioned expression in 1918 with two articles in the AMEC journal, African Advance . Under the heading 'And He Said Unto Me Write', one of these described the linguistic and printing work at Old Umtali. 'What would the World be today without . . . books?' it asked, 'Without a written language?' In Africa,

without a history and a literature, or the knowledge of how to make them, the shadows of the primeval forest have deepened into darkness. . . . That there has never been a written language may in some measure explain why there are nearly a thousand different spoken languages and dialects. . . . The highway to constantly rising levels of human life and living is paved with good books. Africa must have books. . . . and educational literature in the vernacular.'^[26]

From the beginning, then, American Methodist missionaries worked with their converts to create a written language. On the missionary side, the great pioneer was E. H. Greeley. 'My duties have been largely with the boys', he wrote in 1901, 'and it has been delightful to talk with them and to learn their languages, manners and customs.'^[27] On the African side many of the first converts and pupils from the area around Umtali worked closely with Greeley as they progressed on their way towards becoming a Christian elite. They provided the vocabulary; Greeley provided the orthography and rules of grammar. 'I am learning grammar and helping Mr Greeley in translation work', wrote Mark Kanogoiwa in 1909. 'I teach the Chimanyika primer, the New Testament in Chimanyika also.'^[28] Another African pioneer was Enoch Sanehwe, a product of Old Umtali School and a subject of Mutasa, through whom, wrote Greeley in 1910, 'much of our vernacular work has reached its present stage of perfection'.^[29] Others of the founding generation of African Methodism who were deeply involved in language work were David Mandisodza, Joseph Nyamurowa, Paul Mariyanga and Jason Machiwanyika, all of whom were 'Manyika' in the narrowest sense.

Between them, these converts and the missionaries had created something new—a literary instrument which in various ways differed from speech; an instrument to which the mission-educated had unique and privileged access; an instrument

which African teachers and clergy went on to use to write the histories, moral fables and collections of literature which justified their claim to be the leaders of their people. Machiwanyika, for example, produced 'hundreds of pages of history and folklore . . . a volume of material concerning the kings and their wars, and the native customs and a number of hymns of merit. . . . No native I know', wrote Greeley on Machiwanyika's death in 1924, 'spent more of his spare time for the good of his people.'^[30]

Moreover, this written language in which the mission elite came to have such a vested interest was innovative in another significant way. It created rather than merely reflected one specific dialect of Shona—Manyika. In pre-colonial Zimbabwe there did not exist bounded dialect zones within the overall Shona-speaking territory. Each village spoke the 'same' language as its neighbour, across the whole territory, but there was nevertheless gradual lexical and idiomatic change so that, by the time a man from the extreme western edge of the Shona-speaking area reached the extreme eastern edge, he encountered significant differences. Missionary linguists created discrete dialect zones by developing written languages centred upon a number of widely scattered bases. The American Methodists at Old Umtali, the Anglicans at St Augustine's and the Mariannhill fathers at Triashill together produced Manyika; the Jesuits at Chishawasha, near Salisbury, produced Zezuru; the Dutch Reformed Church at Morgenster produced Karanga. Differences were exaggerated, obscuring the actual gradualism and homogeneity of the real situation. And once these new forms had been codified, they then expanded out from these missionary centres by means of the mission out-school networks until specific dialect zones had been defined. As Clement Martyn Doke, missionary and noted linguist, put it in 1931 in his report on Shona language and dialect:

Owing to the way in which Missionary work (and hence language study and literary production) has been developed in districts isolated one from the other, and Missionary Societies working independently, four distinct dialects have been pushed into prominence, viz., Karanga in the 'Victoria Circle', Zezuru in the 'Salisbury Circle', Manyika in Manicaland, and Ndau in Melsetter District. . . . The difference between the dialects has been grossly exaggerated by these artificial means . . .

For the sake of argument, let us suppose England to be a heathen country. Four distinct Missionary Societies commence work, one among the Cockneys, one among the University class, one in Yorkshire, and one in Devonshire. Each produces a translation into the 'local' vernacular, each further uses a different orthography and some split up their words into their component parts. What an enormous difference there would be between the four literary efforts; they would not be mutually understood.^[31]

It was not at all the original intention of the American Methodists to create a dialect. At first they thought they were reducing the language of all the 'Mashona' into written form. When Helen Springer produced the first vocabulary and handbook in 1906 she called the language 'chiKaranga' to emphasize its links with the recorded past of all the Shona. But inevitably the language work of the mission reflected the particular forms of the language spoken among the Manyika of Mutasa. Thus, wherever they worked in the American Methodist zone of influence, they produced Manyika. In 1909 Greeley was based at Mount Makomwe among the people of Chief Maranke, whom the Native Commissioner insisted was not Manyika. In his school, he had '10 sons of the King [Maranke] and all are praying. I am diligently seeking after every possible heir to the throne.

Greeley was hoping to indigenize and localize the gospel among a people who were quite distinct from the Manyika of Mutasa. 'A hundred boys and girls are reading the Gospels in their own tongue', he exulted, 'and going out to talk it and sing it and live it.' But the African teacher with whom he worked in Mount Makomwe and who helped 'in translation work' was Mark Kanogoiwa from Umtali.^[32] In 1910 Greeley remarked that four out of the seven school boys who were going on to higher training at Old Umtali from his school in Maranke were themselves 'from Mutasa's kingdom . . . part of the fruit of our labours there'.^[33] Indeed, the great majority of the 'Native Pastor Teachers. The Apostles to the Manyika People. The output of our Training School', whose group photograph was published in 1916, had been born in Mutasa's chiefdom. Hence, in its formative stages, the 'language of the American Methodists' was profoundly shaped by the experience and contribution of a particular group of converts.

Fairly soon the AMEC missionaries came to accept that what they had produced was 'chiManyika', the language of the 'Manyika nation'. Greeley, who had begun his work by finding 'the language so limited as to vocabulary and especially so in words expressing religious truth', ended by admiring and loving its richness:

The Lord has helped me to say some things in the native language which I have never said in English. . . . There are greater depths in Chimanyika than you dream and while in some ways it is inferior to our world conquering English there are wonderful possibilities in using a vernacular. . . . Even Paul Mariyanga, who has been doing little else for years than translating, is constantly discovering new words and new meanings for old ones.^[34]

Armed with this new literary language and with instructional and devotional materials in Manyika, teachers and evangelists poured out into all the area of the American Methodist outreach. The AMEC paid no attention to colonial administrative boundaries. So far as they were concerned, their thriving stations in the Gandanzara, Chiduku and Headlands circuits, situated in the Makoni Reserve, the Chiduku Reserve and the white farmland areas of Makoni district respectively, were an integral part of their core territory and hence 'Manyika'. Converts in Makoni were offered the full American Methodist progressive package—plough production for the market and literacy in Manyika going hand in hand. A scene recorded in Gandanzara, which rapidly became the nucleus of American Methodist entrepreneurial endeavour in Makoni district, vividly illustrates the interconnections:

In November I carried a trunk full of books and some grain bags to the quarterly meeting at Gandanzara. When the announcement was made that books would be given in exchange for grain, the children scattered in every direction. Presently they returned, each laden with a basket of grain. . . . After the purchase little groups could be seen here and there about the village, the proud owner of a new Catechism asking questions and the other children chiming back the answers.^[35]

In this way a Methodist Manyika 'language' entered Makoni as one of the marks of commitment to modernization. At the same time, boys were recruited out of Makoni district to be added to the elite of potential teachers and catechists being trained at Old Umtali. In 1910 Herbert Howard, propagandist for both plough and pen, toured Makoni to hold up 'to the boys there their need for such a training as Old Umtali affords, and of the need of our church for the best boys'.^[36]

Reluctance to go to live in the feared 'Manyika country' was overcome: soon teachers and evangelists for Makoni constituted a second wave of AMEC 'progressives'. They married into the families of the Methodist leaders in Umtali district and were themselves posted to yet other areas of the AMEC's outreach. As they responded to the great popular clamour for literacy which raged through these eastern regions in the 1910s, such teachers did so through the medium of the literary Manyika which was their particular and prized possession. These men from Makoni had come to see themselves as among 'the Apostles to the Manyika People'.

Such a process of assimilation and incorporation could not go on forever. Over the years the hierarchies of status and authority among the African agents of the church hardened and it was no longer possible for the bright young men of a newly evangelized district to add themselves rapidly and easily to their ranks. So, when the established teachers and evangelists from Umtali and Makoni districts spread out into Mrewa and Mtoko, they encountered a new hostility and resentment from local converts. The differences perceptible between the Manyika texts from Old Umtali and the speech of Mrewa and Mtoko were magnified by this rivalry into specifically ethnic notions. Mrewa American Methodist converts sought to repudiate foreign 'Manyika' teachers and to obtain local 'Zezuru' ones. 'It has always been recognized that there has existed a strong and at times unfriendly feeling between the Africans in the Eastern sections of Southern Rhodesia and those of the Mrewa and Mtoko Areas', noted an AMEC memorandum. 'Although both sections speak different dialects of the same Chishona language, yet the language barrier has always been exaggerated as a reason for these feelings.'^[37] Nevertheless, the idiom of the confrontations in Mrewa and Mtoko served to confirm and sharpen the sense of men from Makoni district as well as men from Umtali that they belonged to a superior Manyika culture.

Language and Ethnicity: the Anglican Church

The influence of the American Methodists alone would certainly not have sufficed to make most people in Makoni district come to think of themselves as 'Manyika'. The AMEC influence was most marked in those areas of Makoni district in which relatively large-scale production of maize for the market was possible—in the Gandanzara area of the Makoni Reserve, which had unimpeded access to the Umtali urban and mining market, or in those areas of the Chiduku Reserve which were close to the district centre at Rusape or to the railway line. It was much less felt among the peasant families who produced a small surplus for sale or among those who lived so far from markets or communications that they had virtually no opportunity to sell produce: families who between them were, of course, in the large majority.^[38] I have argued elsewhere, however, that missionary influence did reach these other two groups; that Anglican influence was especially strong amongst the smaller surplus producers and Catholic influence especially effective in the subsistence production areas in the east of the Makoni district.^[39]

Different though their spheres of influence—and their theologies—were from those of the American Methodists, one thing the Anglicans and Catholics did share with them. They were equally committed to language work and ultimately equally committed to the production of a specifically Manyika language. The Anglicans began with a rather delightful series of speculative assertions about language. Archdeacon Upcher concluded in 1893 that 'the Mashona have two languages, High Mashona and Low, they use the former while clapping their

hands'.^[40] Douglas Pelly told his parents in July 1892 that he was 'learning the language rather quickly or rather 4 tongues, i.e. all the languages between Beira and Matabele, i.e. Shangaan, Mashuna, Makone and Matabele'; which suggests that at that time he was thinking of the eastern dialects in terms of Makoni's rather than of Mutasa's people.^[41] But even as they speculated, and long before they had mastered the language, the Anglicans began linguistic work.

From the beginning Bishop Knight Bruce thought of the eastern districts as the natural site for the main base of Anglican missionary work amongst Rhodesian Africans. This was partly because they were the part of Southern Rhodesia closest to the sea and to international communications; partly because he had been impressed with the evangelical potential of the region. During the early 1890s the bishop pondered where best to establish this eastern headquarters. The two possibilities were Mutasa's country or Makoni's country. In May 1893 Pelly told his parents that he had been chosen 'to start the work at Umtasa's kraal' and to build a station 'which in time the Bishop wishes to be the big station of the diocese. I shall have lots of boys, cart, oxen, etc and everything I want and am to build dwelling houses, church, native hospital, etc and also to begin teaching there. . . . Fancy being chosen to build . . . the first church and station for natives in Manica.'^[42] In January 1894 the bishop himself wrote from the Mission House, Umtali, that:

The House is being dedicated by our beginning the first methodical translation into Seshona of parts of the Bible, the Lord's Prayer, the Ten Commandments and the Creed. Our party consists of our two leading Catechists—Frank and Bernard—who have been learning the language now for 2 1/2 years, Mr Walker, who has been studying it with Frank, and Kapuiya, a Headman's son, and myself. . . . Every word in the grammar and its pronunciation has to be passed by [Kapuiya] before it is allowed to exist. The peculiarities of grammar are extraordinary.^[43]

Language work thus began for the Anglicans where it had begun for the American Methodists—in Umtali district. But Mutasa's kraal turned out to be hotly contested territory; rival parties formed there, one backing the Methodists, another the Anglicans, and a third under the chief himself seeking to repudiate both of them. Knight Bruce's successor began to think of Makoni's territory as a more attractive base. After Pelly's ordination he was posted to:

Maconi's, a Mission Station to the East of Salisbury, founded by my predecessor, and a very important centre which, I hope, will become a strong 'base of operations' for work throughout the district, and, in time, develop into a 'school of the prophets', with its influence radiating through the whole of Manicaland. . . . Much, I am sure, depends upon our having one strong centre of Native work to begin with, where two or three clergy can live in community, and associated with them a number of Catechists, who shall go forth to various stations around.

The bishop planned a hospital and industrial training centre at Makoni's:

I am not eager for a mere literary education of natives. . . . English manhood and womanhood had acquired its distinctive character and nobility long before School boards were dreamt of.^[44]

Despite these strictures on literate education, Pelly's work at Lesapi Mission, Makoni, began with teaching and translation. In October 1895 he told his parents that some fifteen boys 'had learnt the catechism which I have been translating and

also could sing well some hymn tunes which Bernard has taught them. . . . I am going to translate for the singing at once.' A letter in December 1895 gives a vivid glimpse of the linguistic complexity of this little station, where Pelly, the Englishman, Frank, the Zulu, Bernard, the Mozambican, and Kapuya, the local convert, worked daily:

Frank comes to my hut and I give his Bible lesson. . . . Then I read and work and copy translation. . . . Frank meanwhile gardens and helps Kapuia to read S. Mark in Zulu. . . . At half past six Evensong with Kapuia present. The hymns, creed and Lord's Prayer and Advent Collect being in Chino [Pelly's name for Shona]. Then tea followed by school, the first half taught by Frank, while I taught Kapuia English and he taught me Chino.^[45]

It is intriguing for a historian of Makoni district like myself to imagine what would have been the consequence of the development here, rather than at Umtali, of the major Anglican school and language centre. Perhaps the 'Anglican language' would have come to be called 'chiUngwe' rather than 'chiManyika'. This did not happen. Chief Makoni and most of his people came out in arms during the Chimurenga Rising of 1896, sacking the mission station, while Mutasa and his people maintained a somewhat ambiguous neutrality. Thus it was determined that Anglicanism's chief centre should be established in Umtali district.^[46]

The cumbrous machinery of Anglican linguistics was now focused once more on Mutasa. In August 1896 the Zulu priest, H.M. M'tobi, wrote an account of his visit to Mutasa's kraal in which he revealed both his ignorance of the vernacular and of long-established chiefly linguistic ceremony:

I am only just beginning to learn the language of this country, for although I am a native, Mashonas speak quite a different language from any that I have ever heard. . . . I have visited [Mutasa] several times, and he has spoken with me through three interpreters, although I have always taken with me a young Mashona who fairly speaks Zulu, and through whom we might speak to each other. . . . He must have seen white people being interpreted for whenever they go to see him, and he must have thought it expressive of dignity to have so many interpreters. His plan is to sit among the councillors, and then speak very lowly to the one nearest him, which passes what has been said loudly to another and then this one passes it on to the young man I always take with me.^[47]

But Mutasa and his people were to have the last linguistic laugh, as their speech became the model for the Anglican Manyika 'language'.

In March 1897 the decision was taken to set up the Anglican educational and missionary base at St Augustine's College, near Mutasa, whose prestige and the antiquity of whose line began to be built up in missionary reports. Bishop Gaul described him as 'His Majesty occupying the throne of the Monomotopo dynasty, dating certainly from King Solomon's time, 3000 years ago, and who knows how much longer'.^[48]

The missionary brotherhood at St Augustine's set out to learn the local language from their first pupils. Ronald Alexander, writing in December 1898, described the chaotic state of collegiate language learning at that time. 'We want this place to be a college for natives to come from all parts of the country to learn to read and write', he noted. But at that time there was only one full-time pupil, a Mozambican, upon whom they were completely dependent linguistically, together 'with three or four little boys from a kraal close by', upon whom, in turn, the Mozambican depended for his Manyika vocabulary. 'Talking is rather a slow

business', confessed Alexander, 'as one of us speaks first in English, then John [a Zulu teacher] puts it into Zulu, and then our pupil . . . puts it into Chino.'^[49] But less than two years later, in August 1900, Alexander was himself engaged in the business of translation: 'Mr M'tobi and I have been translating Collects today and have done several. We appeal to the boys for some doubtful word, and then a great and often heated argument takes place.'^[50] By 1904 this method, so dependent upon local pupils, had achieved translations of 'all the most useful parts of the Prayer Book . . . together with St. Mark, St. Luke and St. John.'^[51] As with the American Methodists, the missionaries contributed orthography and word-division; their pupils contributed vocabulary.

Meanwhile, Pelly's first catechumens in Makoni had been scattered by the Rising and only slowly came together again. Nevertheless, when the Anglican mission in Makoni reopened at its twin centres of Epiphany and St Faith's, its clergy, and particularly Edgar Lloyd, at once resumed translation work. What happened was that Etheridge at St Augustine's and Lloyd at St Faith's collectively produced an Anglican version of Manyika.^[52] 'Our educational aims will not appear very high', wrote Lloyd in 1905. 'To enable our people to read and write in their own tongue, and to understand what they read, seems the only ideal practicable and perhaps desirable at this stage.'^[53] So he and John Kapuya continued language work. 'After breakfast John Kapuya arrives', wrote Lloyd. "Thanks to the labours of the priest in charge of Penhalonga [St Augustine's] . . . we no longer lack so woefully translations in the vernacular. Yet there remains much to be done and our mornings are made busy in copying and correcting translations.'^[54]

Gradually Anglican language work intensified and hardened into what was no longer called 'chiNo', but 'chiManyika'. Gradually, too, teachers from Makoni and Umtali came to play the same role as their peers were playing in American Methodist work. In 1914, for example, a committee appointed by the bishop at the Native Conference was hard at work at St Augustine's; it consisted of Etheridge and Buck of St Augustine's and Christelow from St Faith's:

When they meet, the discussion of points of grammar . . . is long and learned and often loud; till two vote one way and the third is silenced. . . . What a laborious work it is, and what an immense deal there is to do. . . . Cyril, the eldest of the three native teachers living here, is relied upon most of all, I think, in the translation work. And he is constantly to be found in the afternoons, sitting by the side of Mr Buck; who must first make sure that he really understands the English, and then must get the English turned into the Chimanyika.^[55]

The mission press at St Augustine's had 'three great works on hand . . . the Provincial Catechism, which was originally translated at Rusape and published by them, and now has been entrusted to the St Augustine's Mission Press'; a shorter catechism; and a reader compiled by Buck. 'Rusape have a "Lives of the Saints" in hand.'^[56]

Thus, the Makoni mission remained a major centre of the production of Manyika, ultimately producing a set of 'Rusape Readers', in which the history of Makoni district was presented in the new language. Moreover, as with American Methodist Manyika, but even more so, the teachers and catechists of Makoni and Umtali, who had come to possess this new language, found themselves carrying it into other regions of Southern Rhodesia. The material published at St Augustine's was used in Anglican churches and schools throughout the territory.^[57] And as white missionaries and their African agents encountered rural populations in

other parts of the country, so they came to think of the Manyika language as one aspect of a superior Manyika ethnic culture.

Anglicanism in the east was buoyed up in the late 1900s and 1910s by a grass-roots demand for learning in the new language. In Makoni district schools and churches sprang up, often under the leadership of young men who themselves took the founding initiative, setting up their own Christian villages and only later being appointed as agents of the Anglican church. The area around Umtali clamoured for teachers 'due to the general desire for the New Learning'.^[58] 'Here in Manicaland', wrote Canon Etheridge in 1909, 'it is no longer a question of the conversion of individuals—it is a question of the conversion of a district—practically of a people—the Manyika.'^[59] But it was very different in other parts of Southern Rhodesia, and disillusioned missionaries contrasted these with 'Manicaland' in terms which assumed an identity between language, ethnicity and culture.

Thus S.J. Christelow, earlier based at St Faith's, Rusape, was posted to Selukwe in the Midlands. In December 1914 he drew the contrast between them:

My earlier labour . . . had been at St Faith's, Rusape, where I was privileged to see the wave of enthusiasm for teaching, etc, then passing over that part of the Diocese. . . . Here it seems that many years of hard work must precede the harvest. . . . Instead of about thirty schools as at Rusape, here I have only two in my care. Perhaps you wonder why such opposite conditions exist in our diocese. Geographical position seems to have much bearing on the answer—the Mashona in the north and the Matabele in the south. The Matabele has all along been less responsive to missionary work, whereas the Mashona has, after much labour, responded to an abnormal degree. Here at Selukwe, between the two tribes, we have the Wakaranga. . . . They are bi-linguists. The old men speak fluently both the language of the Mashona and the Matabele. Further, it would seem that in habits they conform more to the Matabele and are slow in embracing the Gospel.^[60]

In the same year another missionary made the same contrast: 'The Wakaranya, as the people here are called, are not nearly so keen or enthusiastic as the Manica.'^[61] 'It is clear', wrote Etheridge from Selukwe in 1916, 'that we must not expect quick results amongst these Makaranga folk; they differ in many ways from those with whom we have to deal in our larger stations. . . . Much more intent upon ploughing than learning, they need different treatment from our Manica people.'^[62] Nor had the situation changed by 1920. 'The Wakaranga', wrote a missionary in that year, 'are very slowly turning to God and his Church; there is no enthusiasm as there is among the Manyika people.'^[63]

In this way the Anglicans gave more and more reality to notional entities such as the 'Karanga' and the 'Manyika'. And wherever else they went in the territory, they longed for the network of mission stations which gave Anglicanism in 'Manyikaland' its almost 'national' character:

It is a weird country, is Matabeleland, dry—physically and spiritually . . . with a proud native race, that shows almost no desire for education, and almost less than no desire for our religion. . . . I want to show you a few contrasts in our work between this district and Manyikaland. Manyikaland appears to me to be a garden watered from above, and with springs and rivers of water rising within it. . . . There growth came from within by native initiative; here it must be organized, and fed, and fostered. We have here no large centre humming with life and devotion, to which the out-stations look for encouragement and

with pride as their birth-place. Thus here we have no esprit de corps, we are split into tiny parishes, and there is almost no inter-communication between stations, congregations or teachers. . . . Again up north there is practically a homogeneous people, and not more than three types of Missions at work. Here the tribes are mixed up, and every imaginable type of Christianity is at work.^[64]

It was, indeed, this esprit de corps, this 'intercommunication between stations, congregations . . . [and] teachers', which brought the African Anglicans in Makoni, Umtali and Inyanga districts into one Manyika identity. Thus Maurice Nyagumbo, whose Nyashanu ancestors had been integrated into Makoni by means of a shared hostility to the Manyika, but who himself was a pupil at St Faith's, regards himself and his family as 'Manyika'. In 1934 the Anglican missionary B. H. Barnes illustrated the situation with reference to language and ethnicity by taking the examples of of the peoples of chief Makoni of the Ungwe and chief Zimunya of Jindwe:

In our Mashonaland there is no single race and language which is definitely representative of the whole area. There are, instead, a large number of small tribes or clans, and almost as many divergent dialects. Any individual will declare himself to be of this or that clan and a speaker of this or that dialect. None calls himself Mashona and there is at present none who will say that his speech is Shona. You will find that he speaks Ciungwe or Cijindwi and belongs to the the Ungwe or the Wajindwi. Either of these may claim to be Manyika and to speak CiManyika, but they know nothing of a Shona race or a Shona language. . . . In the various districts you will find that the lesser divisions are already able to recognize themselves as included under one or other of four or five principal groups, such as Karanga, Zezuru, Manyika, Ndau, Korekore. The process of unification has, in fact begun . . . [uniting] neighbouring sub-dialects under the main dialects.^[65]

Language and Ethnicity: the Roman Catholic Church

The third main missionary influence in Makoni, Inyanga and Umtali was the Trappist/Mariannhill mission. This established itself at Triashill a good ten years after the introduction of American Methodism and Anglicanism and at a time when the main outlines of missionary Manyika had already been laid down. Nevertheless, it made its own peculiar contribution to the spread of Manyika identity, and this for three main reasons.

The first was that Triashill and its associated station of St Barbara's was not near Umtali, like St Augustine's and Old Umtali, but rather on the border between Makoni and Inyanga districts. As its out-stations spread out eastwards into Inyanga and westwards into Makoni, Triashill served to bring together its converts across what had been a military frontier between chiefs Makoni and Mutasa. Secondly, Triashill and St Barbara's were cut off from good communications and markets by a great range of hills. The people who lived on the mission farm and the areas around it were subsistence producers and labour exporters, very different from either the entrepreneurs of the American Methodist Gandanzara or the surplus-producing peasantry of Anglican Chiduku. The Mariannhill fathers and brothers had a good deal to do with the Manyika identity in Makoni becoming more than merely an elite self-identification. Thirdly, the Mariannhill missionaries were perhaps more self-consciously concerned to stress the past glories of the 'Manyika' and to foster a sense of local pride than were either the American Methodists or the Anglicans, perhaps

because they felt themselves constantly under the scrutiny of the disdainful and Zezuru-propagating Jesuits.

The main Triashill linguist was Fr. Mayr—a man 'so simple, so thoroughly earnest about all the work he undertook, so modest and unsparing of himself . . . a highly talented linguist, a thorough master of Zulu and Chimanyika languages'.^[66] According to a later note prepared for the Jesuits by Fr. Withnell:

Father Mayr was given the work of preparing books in Shona for the new Mission. Fr. Mayr . . . said he wished to visit St Augustine's and asked me to accompany him. I was present with an attentive ear at the discussion which Fr. Mayr had; Canon Etheridge was the chief spokesman. Fr. Mayr seemed very keen on forming his native language on the prevailing St Augustine's language. What he seemed to insist upon was that the language of his projected books should not slavishly follow Chishawasha [i.e. the Zezuru literature of the main Jesuit centre], that in fact it should be different, as far as was seemly, from Chishawasha. With a smile between them the two learned linguists were at one on this point. . . . I know that prayers, as arranged by Fr. Mayr, differed much from those of Chishawasha. . . . So far as I have heard, these versions appeared without any approval. . . . The books were rushed through the press to supply a pressing need. . . . What struck me in their books was a determination of the Trappists to differ from Chishawasha language.^[67]

Once Mayr's Manyika material was ready, the Triashill fathers wished to use it throughout Makoni district. They also ran a station at Monte Cassino in western Makoni, as far from Mutasa's Manyika as anywhere in the district. In August 1911 Fr. Fleischer wrote from Triashill to the Jesuit Apostolic Prefect asking permission to 'introduce the Chimanyika language spoken by the natives in and around Triashill at our mission of Monte Cassino. At present we make use of different catechisms on our two stations there, i. e., a Chiswina one at Monte Cassino and Chimanyika one at Triashill but in the future time we think that the catechism which Fr. Mayr compiled in Chimanyika for Triashill to take also for Monte Cassino . . . [to] make uniform our work for the black people.'^[68]

This proposal to bring all Makoni district into the Manyika language zone triggered off a great language controversy between the Jesuits and the Mariannhill fathers and among the Jesuits themselves. Many Jesuits reacted with hostility to the Triashill proposals:

There is but one language in Mashonaland [wrote the Jesuit Fr. Bert in October 1911]. It may be called Chiswina around Salisbury, Chimanika near Inyanga, Shona, Shuna generally, Chikaranga . . . about Victoria, but the language is essentially the same. . . . There is and should be only one language for our Mashonaland stations. . . . I would boldly stand by the Chishawasha catechism. Not because it comes from Chishawasha—although the first or oldest . . . station might reasonably urge its claim on that score—not merely because it is the one actually 'in possession', but chiefly because it is the translation of the one followed in England.^[69]

Fr. Withnell was also an advocate of the dominance of Zezuru. He hoped that 'the Salisbury [Chishawasha] dialect may be kept to; in my experience it is the most widely understood; just as in my experience, too, this dialect is richer and, as far as I can judge, more correct than others. . . . When I went to Chilimanzi. . . . I was agreeably surprised to find that Chishwasha had everything that Chilimanzi had, and more besides, Fr. Hornig had previously told me the same of Chimanyika.' 'I think that the central Rhodesian dialect ought to be

preserved', wrote Withnell again. 'I think it would be very unsafe to follow Chimanyika . . . with its wretchedly poor vocabulary.'^[70]

As the controversy developed it extended beyond vocabulary into theology and desirability of missionary 'adaptation'. The Mariannhill fathers had followed St Augustine's and Old Umtali, for example, in translating 'God' as 'Mwari', while the Jesuits in Chishawasha used 'Jave'. The priests at Triashill chose Mwari after taking the advice of their teachers, drawn from both Inyanga and Makoni. The record of the teachers' discussions illustrates very well the compromises that were drawing them together within missionary Manyika:

Four of us said that 'in olden times the word Mwari was never used, it only came to us by the first missionaries of the Church of England. The old people always used the word 'Nyadenga' . . . the rest of the teachers said the word Mwari was always used in some districts and Nyadenga in others. . . . Then all of them agreed that the word Nyadenga mean only the owner of heaven, then Mwari should be the really name of God.^[71]

Jesuit critics of Triashill objected precisely to such consultations and to the indigenous input into the formulation of Manyika. Old Fr. Richartz wrote in very different terms in describing the evolution of Chishawasha Zezuru:

We found that the heathen word Mwari had no definite meaning besides 'Rainmaker' and we were in this special important case like in general, very much impressed by the secrecy, want of clearness and straightforwardness with which all actions of the Natives were surrounded, especially everything that could be supposed to be in connection with 'religious' ideas as prayer, sacrifice, etc at occasions like child-birth, marriage, field work, war, etc, etc. Most of all was our teaching hampered by using existing words. . . . This very doubtful meaning of Mwari forced us to avoid this dangerous word and replace it with another word which surely meant the true God and fitted well in the native language (Chizezuru and Chikalanga), Yave.^[72]

The persistent Withnell extended these criticisms to Triashill—and St Augustine's—usage in general, instancing the case of the translation of the word 'prophet':

You are perfectly right in coining the word 'anoprofeta' [he wrote to the Jesuit Fr. Johanny]. . . . You are simply applying the principle which has guided us all along in rendering ideas of the religious or supernatural order, for which Chizezuru could not possibly have an exact equivalent. The Trappists went on a different tack, and tried to pour the new wine of Christian ideas in the old, rotten skins of pagan words, creating confusion where there should have been unity.^[73]

The critics drew a picture of sound Catholic usage everywhere else in Southern Rhodesia contrasting with an infernal alliance between Triashill and the 'sects' in the use of 'wretchedly poor' and 'rotten' Manyika.

There were no obstacles to the spread of Old Umtali Manyika out into Mrewa and Mtoko or to the spread of St Augustine's Manyika into almost the whole territory. Triashill's case was very different. Jesuit opposition prevented them from extending Manyika even to Monte Cassino and threw them very much on the defensive. Soon Jesuit opponents followed up this negative success. They persuaded the Jesuit Apostolic Prefect, Msgr. Brown, that the use of 'Mwari' was intolerable. In August 1923 Brown ordered that the word must be replaced everywhere by 'Jave' forthwith.^[74] When the Triashill congregations continued to

use 'Mwari', Brown threatened to close down all Trappist/Mariannhill stations immediately.^[75] As Brown later wrote:

Although our Jesuit Fathers had been in the territory for forty years and knew the native language well and had finally decided on the word to be used for God, the Mariannhill fathers without consultation with us began to use a word which although used by Protestantism had a very evil connotation. . . . I gave a formal order that Jave was to be used. In my next visit to Triashill, six months after my order had been given, I found no notice had been taken of my decision and that the objectionable word was still used. I then gave an order on the subject which had to be obeyed under pains of censure. This was obeyed.^[76]

This prohibition was deeply resented by both the missionaries and the Triashill Christians and was seen as auguring a more general attack on Manyika. Together with disagreements over other questions of evangelical policy, the question led the Mariannhill fathers to approach Rome with a request that Manicaland be made a separate prefecture so that they could create their own Manicaland Christianity. When this failed, the Mariannhill superiors began negotiations with the Jesuits for an exchange of territory which would give the Manyika stations over to Jesuit control and give the Mariannhill fathers complete control over a mission area elsewhere. But the priests and brothers actually on the Manyika stations profoundly disliked abandoning them to the Jesuits, so that when Jesuit priests moved in to take over Triashill, St Barbara's, Monte Cassino and the rest of the Mariannhill stations in September 1929, they found the Mariannhill fathers still in definite possession and refusing to move.^[77] In the end they did move out, leaving behind them an apprehensive and disgruntled 'Manyika' church to confront its new missionary masters.

No love was lost on either side. In March 1930 the Jesuit Jerome O'Hea wrote to Brown bewailing 'the dearth of teachers who are fit for the work. . . . The Manyikas want and seem to expect incessant propulsion a terga . They are an awfully slack crowd and they find good medicine in Fr. Schmitz who won't stand any nonsense and tells them often and clearly what he thinks.'^[78]

Manyika Identity and the Migrant Labour Factor

I have argued above that the expansion of missionary Manyika into other areas, carried as it was by teachers and evangelists from Makoni and Umtali districts who enjoyed privileged access to it, went a long way to build a common sense of Manyika identity among them. But it was the Jesuit takeover of Triashill in 1929, with the accompanying introduction of the Chishawasha church and school books in the Zezuru dialect, that provoked by far the most articulate expression of a Christian sense of Manyika self-identity.

The Triashill teachers tried to defend their leadership of a Manyika church. 'Trouble is brewing among the teachers', wrote O'Hea in September 1930, 'as may be clearly seen from the set we have here doing their course. One or two have openly declared that they will never teach children from a Chizezuru book.' One of these teachers announced in class that if Msgr. Brown 'doesn't give us everything we want' and restore the Manyika church, he would 'give up the whole thing'. 'To my way of looking at it,' exploded O'Hea, 'they are treated FAR too softly. I wouldn't give that chap a chance of giving up—I'D FIRE HIM AT THE END OF MY BOOT. Close his school and give a jaw to the Christians who were any good around about it, letting them see that such bolshies brought nothing but trouble and unhappiness.'^[79]

But the most active pan of the protest was taken by migrant labourers from the Triashill zone of influence. Very many men had taken their mission education with them to the towns of Southern Rhodesia and South Africa. Now they intervened in the crisis at home. "The boys of Jo-burg', wrote O'Hea, 'have sent unsigned letters to boys about here to stir up protests at the church on Sunday and to get the teachers on their side against the use of Chizezuru books in Manyikaland. "You are taking away our language . . . our King Mtasa . . . the religion of our forefathers" . . . all this high-sounding rot is being pumped out at various kraals.'^[80] Triashill, wrote the Johannesburg migrants, 'belongs to us black people, not to the whites'. The Jesuits had come 'to spread divisions, because they have come like intruders to despise us, as we have made no agreement with them. . . . See how these Jesuits are stepping into our possessions and snatching them from us! We, the owners of these things, how are we being treated by them? We are being looked upon as nobodies.' The teachers were urged to write 'many and countless letters' to Fleischer, now bishop in Natal, 'for he knows Chimanyika', and through him their case would reach 'the Holy Father'. 'Announce this to all the people, assemble together at all times.'^[81] Patrick Kwesha, the leader of the Johannesburg migrants, wrote to Brown himself to complain that people at Triashill 'are forbidden their own tongue and forced to use a tongue which we never knew nor our fathers'; that the Jesuits were closing down schools and were 'not the means of teaching us or serving us or raising us, but the means of destroying us and killing us.'^[82]

Meanwhile, the economic depression was forcing many of the migrants back home. These men joined in the protest against the Jesuits:

There is a certain amount of trouble, of a Bolshi sort, brewing and coming to the surface among the Manyikas [wrote O'Hea]. . . . The Young Ethiopians, the ICU and a few local societies are talking a great deal of hot air in the Reserves. . . . Boys back from Capetown and Johannesburg are particularly active and it is really astonishing to hear the clear echoes of Moscow out here in the wilds. The missionaries, of course, are merely for the sake of drugging the blacks to make them the slaves of the whites. . . . A fairly good grasp of the theory and practice of 'class war' being in the hands of the black . . . it is easy to be understood how the slightest grievance is seized upon and ventilated to the full. One of these grievances is the language question.^[83]

The combination of teachers, labour migrants, returned labour migrants and the general congregation confronted O'Hea with a populist Manyika movement, expressing itself both in Makoni and Inyanga. O'Hea called down curses on the whole idea of the Manyika identity:

Nothing but a rod of iron is any use for these people. . . . They are utterly blinded by the most foolish vanity. It is a poison that has its roots in history—they the most despised of the despised Mashona are given a chance at last, owing to the coming of the white man, and they now openly declare that they are the cream of the black race!!! Trouble that has its roots deep down in history is big trouble. As far as I can gather Bro. Aegidius fostered this idea, if he did not hold it himself, among them. . . . We had a great discussion the other day in the Makoni Reserve and there were a number of pagans present. . . . One fellow envisaged the rising of the kingdom of Mtasa, claiming independence.^[84]

All this was the most spectacular manifestation of migrant 'Manyika' identity: after all, in addition to the Jesuit intervention, the Triashill missionaries had made

more contact with the zones of subsistence production and labour migration than had the Anglicans or the Methodists. Still, both these missions had also prepared young people to carry their skills of literacy into labour migration. Because of these skills, Anglican and Methodist migrants had access to much desired jobs in domestic service and in hotels. The churches kept in touch with them in ways which heightened their sense of Manyika identity. 'The little intercession paper printed here in their own tongue every quarter', wrote Buck from St Augustine's in September 1914, 'has reached a circulation of 2000 and goes to Manyika boys and girls as far away as Johannesburg and Kimberley.'^[85] The American Methodists found in the mid-1930s that they were 'every year drawing from the Reserves hundreds of the best of the boys and girls, training them, and seeing most of them return not to their homes to enrich the life there, but to the town', going into 'cooking, waiting table, carrying messages or wheeling babies in the park'.^[86]

So many Old Umtali pupils were to be found in Salisbury that the AMEC sent one of their ministers there to preach to them in Manyika. Rather than having to join 'Zezuru' Wesleyan Methodist congregations, 'to conduct our own evangelistic services, to sing our own hymns, to be free to develop in our own way, would make our preacher and our people rejoice'.^[87]

An engaging and unusual insight into the network of the American Methodist Manyika migration is given in Katie Hendrick's little-known The Bend in the Road . Katie was a coloured woman, daughter of a Manyika labour migrant to Cape Town, Mandisodza. Mandisodza and his elder brother, John, both attended Old Umtali school. But:

In the dry dusty reserve near Umtali . . . the people went naked and struggled with the land to yield a mealie crop . . . Necessity forced on father and his people the first lesson in economics; to buy food . . . they had to have money. To get money they must go to work for the white man.

Brother John went to work as house-boy to a settler family in Rhodesia, whose service he left after sixteen years to return to Old Umtali and eventually to become an American Methodist minister based in Maranke. The younger brother had more wide-ranging ambitions: 'The many Manyika natives in Cape Town were earning more pounds than they had dreamed of in Mashonaland and they wrote home glowing reports. Father, like many of his contemporaries, became obsessed with the idea of reaching the Cape.' So he worked his way slowly south, taking hotel jobs in Salisbury, then jumping trains to reach South Africa: working for a time for a Constantia farmer and eventually getting a domestic service job in Cape Town by means of his contacts among the Manyika there. Some time in the late 1920s Katie was born to a 'Cape Coloured' woman with whom Mandisodza had established a liaison. Thereafter Katie's story is obsessed with the Manyika and with the idea that Manicaland should be her home. At eighteen years old she has an affair with a Manyika house-boy—'his high-sounding English I found irresistible . . . my weakness for English and Manicas got me into trouble'. Later she visits her uncle John at his mission in Maranke, feeling instantly at home. Uncle John gives her a letter of introduction to an American Methodist Manyika in Cape Town, Willie Marangha from Makoni district; she meets Willie; they marry; and the book ends with their return to Makoni in the early 1950s, where Willie begins a transport business and Katie learns to fit into 'Manyika' society.^[88]

By no means all migrants from the eastern region were mission Christians, of course. But it was to the advantage of such migrants to attach themselves to the growing 'myth of the Manyika' in the towns of southern Africa, with their

irresistible 'high-sounding English'. In the urban and industrial ethnic hierarchy the 'Manyika' came to be thought of as 'natural' domestic servants. "The Wamanyika are in great demand as domestic servants', noted the Native Commissioner, Inyanga, in 1929, 'and they obtain good wages for this class of labour. . . . The wage for this unproductive labour is generally higher than for other productive work.'^[89] Chief Native Commissioner Bullock gave official backing to the myth: 'These Wa Manyika . . . now supply the best houseboys to South Africa.'^[90] During my field research in Makoni district I found that very many informants had worked as waiters in South Africa and retained photograph albums to prove it. They also retained memories of the 'Manyika' networks along which they were able to move from job to job and town to town.^[91]

All this meant that the realities of the migrant labour market made an impact even on the most stubborn 'traditionalist' groups in Makoni district. Thus the Rozvi chief, Tandi, and his people were very reluctant to accept even a Makoni identity, let alone a Manyika one. They were concerned to argue that they had always been autonomous of Makoni, and they were regularly involved in schemes to re-establish the Rozvi paramountcy. But as chief Tandi and his councillors told me in 1981, they ceased to speak Rozvi and came to speak what they now call Manyika 'because we had to marry foreign women. And when our young men went to town they had to accept that they were Manyika.'^[92]

Similarly, at one level of politics within Makoni district the old distinctions between 'Waungwe' and 'Wamanyika' retained their importance in the 1930s. Thus in 1936 there erupted a great controversy between chief Makoni and chiefs Mandeya and Mutasa. Makoni claimed that Chikumbu, a headman in Inyanga district, was his nephew and should be recognized as subordinate to him:

This matter is wholly inspired by head-of-kraal Chikumbu [wrote the Native Commissioner, Inyanga], who for some years, has been proselytising amongst the Va-Manyika in his vicinity. . . . All this people will contend they are Va-Hongwe if asked, although their names and dialect are of the Manyika.^[93]

This micro-level campaign to change Manyika identity into an Ungwe one infuriated chief Mandeya, whose men assaulted headman Chikumbu and his supporters. It also ran quite counter, of course, to the irresistible pressures at the macro-level which were compelling all men from Makoni to define themselves as Manyika in the towns.

In September 1913, for example, a cluster of self-described 'Manyika workers' at the Argus Printing Company in Salisbury, which included men from both Umtali and Makoni, wrote in protest to the administration:

There are three of Manica girls here and they have no thing to do. They are not working. Their work is to commit all sorts of evil going from one man to another. We wish you could report it to Native Commissioner in Salisbury and have them sent back to Umtali; or sir if it please you send us a letter that we may carry it over to Native Commissioner in Salisbury. Also we wish if there is a law that shall never allow a Manica girl to come here and not let one of them ride the train from Umtali, or from any of the stations and sidings, such Odzi, Rusapi and etc save she that has her husband.^[94]

In Salisbury boxing and other recreations were 'organized around a number of "tribal" clubs, comprising the WaKorekore, the MaBlantyre, the WaManyika on the some side, the WaZezuru and WaBudjga on the other'.^[95] More radical groupings existed also. 'There are societies established in the following towns', wrote the Native Commissioner at Inyanga, in November 1930, 'Cape Town,

Bulawayo and Salisbury known as the Young Ethiopian Manica Society. The objects are political and many of the younger Natives of the district belong to one or other of the branches in the towns.'^[96]

There was little point in maintaining a distinct Makoni or Ungwe identity in the towns, and little possibility of doing so. Certainly other African workers lumped everyone from Umtali, Makoni and Inyanga districts together. In Bulawayo, for example, 'Manyika' migrants had entered domestic service and other employment from the turn of the century. Their access to good jobs had caused them to be disliked by Ndebele workers. Thus the Superintendent of Natives, Bulawayo, explained the great 'faction fights' in Bulawayo in 1929 as 'an attempt to oust the Manyikas and the Natives of Victoria District, large numbers of whom are employed in the town, and because they are good servants, were preferred by employers at good wages. Also because these men attracted Matabele girls because they were in a position to do so financially.'^[97] In 1930 rumours flew that there would be a renewal of fighting at Christmas time. It was rumoured that a meeting of Ndebele had agreed 'that the following tribes must be wiped out, or so terrorized that they would leave Bulawayo and the general labour field to the Ndebele, viz., Abanyika, Makorokori, Blantyre, Amabemba.'^[98] Such reports made many migrant labourers from Makoni district flee Bulawayo and return home to avoid any such attack on the 'Manyika', an identity which had become for a time disadvantageous. But in rumour at least, 'a number of Manyikas are coming down from Salisbury at Christmas time to assist the members of their tribe in Bulawayo against the Matabele'. 'On the part of the Mashonas,' it was reported, 'the daring and defiant ones hail from Victoria and Umtali Circles who consider they are a match for the Matibili and are quite ready to meet them.'^[99] It is clear, then, that by the 1930s the idea of an extended Manyika identity, which included the men of Makoni district, was very much alive, not only in 'Manicaland' itself, but throughout Southern Rhodesia.

It is also clear that for some men this 'Manyika' identity was not just a convenient reference group, but an ideal which sustained them during their migration. There was, for example, Patrick Kwesha, spokesman of the Manyika workers of Johannesburg in their protest against the Jesuits. Kwesha used 'to convert people in Johannesburg. . . . Many, many were baptized. These people were nearly all labour migrants from Manyikaland. There may have been some South Africans too but what he was really interested in was the Manyika.'^[100] Kwesha's 'only wish which led my whole life', was to go back to Manyikaland and to set up an all-African missionary order there:

I wanted God to be fulfilled . . . and his great servants to rise up among them, and that some of them as apostles may swarm over the whole of Manyikaland, as the bees, and banishing all heathenism, protestantism and superstition, and establish a pure and holy kingdom of God and glory of Mary so that from that our God may be called God of Manyikaland as He was called God of Israel, and Our Lady be called Queen of Manyikaland as She is called queen of angels and saints.^[101]

The Movement Towards a Shona Identity

It seems plain, then, that by the 1930s the Manyika identity was a reality in Makoni, Umtali, and Inyanga districts and in the migrant diaspora. It had arisen as a result of the operation of the main forces which transformed Makoni district under colonialism—participation in peasant agriculture, labour migration and

education. It had arisen as a result of the emergence of local entrepreneurial and church elites, but it also allowed for wider protest against the manifestly failing colonial political economy in the 1930s. Whites and especially missionaries had played a key role in the definition of the Manyika identity but in such a way that the idea was open for all sorts of use by Africans.

It is plain, too, that by the 1930s there was hardly any sense of a wider Shona identity. Commenting on the visit of the Prince of Wales in 1925, the Chief Native Commissioner contrasted the impact of his appearance before Ndebele and Shona chiefs respectively. To the Ndebele the event was 'a natural outcome of their unforgotten traditions of personal fealty', but 'to the disunited Mashona tribes a message of different import seemed to be conveyed. To them the occasion was charged with unity under one head as a fact which they have been slow to realize.'^[102] But during the 1930s a number of forces, some African and some European, combined to work for the creation of a single Shona language and hence the sense of a single Shona identity.

On the African side, the main impulses came from men who found employment in government service and could be posted anywhere in the territory, and from young teachers and trainees for the ministry, who wanted to contest the prestige of the older generation of teachers and evangelists by mastering and controlling a new language of their own. On the European side the chief impulse arose out of the need to communicate effectively beyond the limits of one of the dialect provinces.

Benedict Anderson, in his brilliantly suggestive Imagined Communities, has discussed the interactions of language development and national feeling. 'Print-languages', he writes, 'laid the bases for national consciousness'. Yet my own data up to this point has been very different from his. In Anderson's case, what he calls 'print-capitalism' created large unified languages where there had been an oral multiplicity:

In pre-print Europe, and, of course, elsewhere in the world, the diversity of spoken languages, those languages that for their speakers were (and are) the warp and woof of their lives, was immense; so immense, indeed, that had print-capitalism sought to exploit each potential oral vernacular market, it would have remained a capitalism of petty proportions. But these varied idiolects were capable of being assembled . . . into print languages far fewer in number. . . . Nothing served to 'assemble' related vernaculars more than capitalism which, within the limits imposed by grammars and syntax created mechanically-reproduced print-languages, capable of dissemination throughout the market.^[103]

In the case I have been describing, printing had actually broken up a common language into distinct and competitive dialects. Anderson tells us that at the root of ethnic nationalism in Europe lay the demotion of 'the old sacred languages—Latin, Greek and Hebrew', now forced 'to mingle on equal ontological footing with a motley plebeian crowd of vernacular rivals'. In the Southern Rhodesian case, the written or printed vernacular functioned in the first place as a sacred language in itself. The mission printing presses, constrained by lack of finance, were not true representatives of 'print-capitalism'. Initially, the various churches were more concerned to Christianize intensively in their own particular zones than to create a territory-wide market for missionary literature. It was the use of literacy by labour migrants which linked this process with the industrial political economy.

But after the 1930s Anderson's ideas of the direct impact of 'print-capitalism'

and of the secularizing character of language development become more and more relevant to Southern Rhodesia. First the state, and then capitalist enterprises began to wish to reach the widest possible market with vernacular publications and in these publications the printed vernacular was widely employed for the first time to communicate a whole variety of secular messages. Hence a drive towards 'Standard Shona' began.

Yet, of course, the mission churches themselves were also parties to this movement. As the Anglican and Catholic churches became more truly 'national', so they pressed for a language more generally useful. Take, for instance, the change that had come over the school at St Augustine's. As we have seen, this had begun with pupils drawn only from the 'Manyika' zone. By the 1930s the situation was very different. St Augustine's had become an elite school for boys from all over Southern Rhodesia:

Three languages are spoken in the Diocese, viz. Chishona in one or other of its dialects . . . Sindebele . . . and Sechuana. In addition to these we have boys in the school from Nyasaland and two different districts of Northern Rhodesia. Many educationalists nowadays stress the value of the teaching being given in the vernacular. This is entirely impossible in the higher of our standards in the schools at Penhalonga, owing to the varied linguistic groups from which the pupils are drawn, so English has perforce to be the medium of instruction . . . St Augustine's, Penhalonga, is the only Church school of its type between the Zambezi and the Limpopo, and between the Indian and Atlantic Oceans. The problem of race and language is also seen among the African clergy of the Diocese. Not only have we Mashona, Matabele and Bechuana, but we also have a Nyasaland native and a Mundao . . . yet they all get on amicably together, and with a little tact in placing them the language difficulty is overcome . . .

It must be recognized that in this part of Africa, at any rate, the difficulty of language does not arise from the difficulty of learning another speech, but from racial pride and tribal animosities; a centre where Africans of different races are collected may easily prove a place where faction fights break out. . . . It says much for the Christian character and influence of our Mission schools that these people who comparatively recently were members of tribes at war with one another are now amicably meeting together under the aegis of the Christian church.^[104]

The solution for St Augustine's was to use English in teaching rather than to employ a 'Standard Shona'—and one must recognize that command of English had a great deal to do with the emergence of a nationalist elite. But it was also significant that Anglicanism had ceased to see its role as propagating Manyika. And many Anglican missionaries took a leading part in the movement towards unifying Shona that commenced in the 1930s. These developments coincided with the generalization of ethnic identification in the towns, and particularly in Bulawayo, where 'Ndebele/Shona' rivalries became established.

All these factors—the aspirations of a second-generation African elite, 'print-capitalism', the enlargement of urban ethnicities, and so on—meant that the movement towards a unified Shona language and a sense of wider Shona identity became irresistible. If in the 1930s no one in Makoni would have described themselves as 'Shona', by the late 1950s, when the nationalist movement came to the district, very many people thought in such terms. Yet if the movement was irresistible, this is not to say that it was not resisted. African vested interests in Manyika were strong and the new language policy was much resented.

Jeffrey Peires, in his account of the Lovedale Press, has shown the devastating impact upon the first generation of mission converts of the 'great orthographic upheaval which struck Africa in the 1930s' as a 'result of the work of the International Institute of African Languages and Cultures founded in 1926 . . . [which] sparked off a new quest for language standardization throughout the continent':

By comparison with other languages, which were severely mauled by ill considered attempts at dialectical unification, Xhosa was fortunate [writes Peires]. . . . Nevertheless, problems with regard to matters such as tone and aspiration had led to inconsistencies in spelling. . . . Far-reaching changes were adopted. . . . However satisfactory these new arrangements may have been from a purely linguistic standpoint, their social implications were disastrous. The manner in which the regulations were imposed awakened deep resentment among educated Africans. . . . The awesome effect of the 'New Orthography' was to turn every literate African into a functional illiterate.^[105]

No one has yet done the research in Southern Rhodesia to determine whether the language changes of the 1930s constituted 'ill-considered attempts at dialectical unification'. But there is no doubt that the changes were deeply resented. There is plenty of evidence that both local missionaries and local African Christians clung to Manyika as best they could.

In March 1934, for example, a visitor from Triashill—even under its Jesuit dispensation—came to talk with Edgar Lloyd at St Faith's. Both men were 'dead keen to stop the objectionable "new language" stunt. . . . It is silly to force it, by printing all sorts of special papers in it, which the natives don't understand and want to read. The Diocesan Intercession forms, the Mothers' Union prayers, etc besides school things.'^[106] American Methodists, whose influence remained focused in the eastern districts, not surprisingly, especially resisted language standardization. In 1940 their Commission of Worship and Music under Amos Kapenzi demanded that 'Choirs should make it their practice to sing Chimanyika.' And in 1947 one of the Old Umtali missionaries wrote to the Scripture Gift Mission, thanking them for sending illustrated lessons in standard 'Chi-Shona':

I find that our educated people can use them, but the common people, common because there are so many of them, cannot understand these lessons as written. The vocabulary is often beyond their understanding, beside the foreign spelling of words they might know. All our people consulted are as puzzled as I about what dialect they may be. They feel that it is highly seasoned with Sindebele, and must come from the extreme South or West border of Rhodesia. . . . Unfortunately, it is so far removed from the whole Eastern section in its composition that it is not understood by the peoples of this section. It is supposed to be Unified Shona, worked out by a committee from the different sections, but some were not called in for the committee meetings, and so these sections were not represented.

The missionary sent a text in Manyika, confident that it 'will meet the needs of a large section, not only of Southern Rhodesia but also of Portuguese East Africa.'^[107] Thus the sense of loyalty to Manyika and to a Manyika identity continued underneath the rise of Shona cultural ethnic awareness and Zimbabwean political nationalism.

Conclusion

I have taken my discussion up to the late 1940s but I cannot leave this chapter without some attempt to connect up my story with the questions raised at the beginning. What were the connections between the development of the sense of a wider Manyika ethnicity and the assassination of Chitepo, the faction fights within ZANU, or the support given to Muzorewa in the 1979 elections? Did the surviving loyalty to a Manyika identity surface destructively in these ways?

There are various possible answers to this. One, which is certainly true, is that the ethnic component in nationalist disputes and in the voting support for Muzorewa has been greatly exaggerated. But other answers speak to what ethnic component does exist. Just as migrant labourers in the past found a principle of mutual assistance in dialect ethnicities, so exiled nationalists in the 1970s did begin to group together on the same basis. The regional politics of development and patronage of today's Zimbabwe also make these identities pertinent—no one doubts that Edgar Tekere, as Chairman of the Manicaland regional committee of ZANU/PF, speaks for 'Manyika' interests. But there is another dimension.

The developments that I have been describing were in no sense a monolithic process. A Manyika identity emerged, but it emerged as the distinct work of three different missionary churches, each appealing most to different areas and to different classes or proto-classes. The African groups interested in developing the Manyika identity were also distinct from each other, both in their experiences and their interests. Hence 'the Manyika' never existed as a bloc, entirely at the disposal of any leader or party. Indeed, it seems to me that where Manyika-hood did play a role in recent politics it was in this particular rather than general way.

Thus in December 1975 Joshua Nkomo's Zimbabwe Star attacked Muzorewa for 'appealing to the African of the Mushona tribe using him or her as a cover while in reality the group leadership is interested in . . . the Manyika tribe. . . . The death of Herbert Chitepo and Bishop Muzorewa's United Methodist Church are cleverly used as baits to woo the Vamanyika.'^[108] It was not in fact true that Muzorewa cared little for the support of other groups, but it was true that in Manicaland his support was founded upon his control of the old American Methodist Manyika network. The stalwarts of the church, who had previously been unwilling to commit themselves to mass nationalism, came solidly behind a leader who had himself been brought up in Makoni district and whose family represented the quintessential ambitions of the Methodist entrepreneur.^[109] As I have argued elsewhere, Muzorewa's essential support came from this particular strand in the 'Manyika' tradition.^[110]

In February 1980 James MacManus of The Guardian brought out clearly the distinct strands in the 'Manyika vote':

There lies St Augustine's, the most famous mission school in the country, whose old boys, including the late nationalist leader, Herbert Chitepo, have risen to positions of influence in every sphere of African life in Rhodesia. The Anglican school has strong nationalist connections which . . . explains the loyalty among the staff and students to Robert Mugabe's party and army. . . . The arguments advanced at St Augustine's can easily be reversed at Old Umtali Mission a few miles away which counted Bishop Muzorewa among its old boys. Here one hears lurid tales of ZANLA activity.^[111]

I have tried to show elsewhere that the Catholic strand of Manyika-hood played the most radical role of all during the guerrilla war.^[112]

Nevertheless, though certainly not 'traditional' or 'natural' or monolithic, the

continued existence in this multi-layered way of the idea of an extended Manyika identity does offer, along with the continuance of the other sub-Shona ethnicities, a potential danger to the Zimbabwean state. During the war the Rhodesian regime belatedly picked up and exploited the notion of this 'traditional' ethnic division—hence the maps setting out their exact percentages. In January 1976 the Zimbabwe Star alleged that the Rhodesian Front had a secret plan. Operation Shaka. 'According to this document the first task of the R.F. is to divide the Ndebeles and Shonas by inciting tribalism and magnify their tribal differences. The second task is to divide the Shonas themselves and break them into the constituent clans—the Manyikas, Makaranga, Mazezuru, Makorekore, etc, etc, etc.'^[113] The Zimbabwe Star feared 'good-for-nothing megalomaniac so-called leaders who cannot impress anybody and resort to tribalism as an instrument for manufacturing support'. It might be that strands in the Manyika identity could again be manipulated. For this reason it seems useful to set out, as this study has tried to do, how and why the notion of the extended Manyika identity arose in the first place; not indeed as the fruit of megalomaniac tribalism but as a very human and often constructive response to socio-economic change, a response however which now needs to be replaced by the development of other kinds of consciousness in a period of even sharper transformation and contradiction.