Minority language and the state in Zambia and Botswana

by
Wim van Binsbergen

Minority language and the state in two African situations

The Nkoya of Zambia and the Kalange of Botswana

van Binsbergen, W.M.J., 1994, ‘Minority language, ethnicity and the state in two African situations: the Nkoya of Zambia and the Kalanga of Botswana’, in: Fardon, R. & Furniss, G., red., African languages, development and the state, Londen etc.: Routledge, pp. 142-188.

Language differences often provide an anchorage for ethnic identity. Ethnic self-articulation tends to have a linguistic component: propagation of the language spoken by a national minority, in the face of lack of recognition of that language in a nation-state’s language policy covering such domains as formal education, the judiciary, contacts between the state and its citizens in general, political discourse, freedom of expression and the media. Language policy — even if appealing to ‘objective’ considerations of linguistic analysis, constitutional equity and socio-economic development — is often formulated and implemented in a political and ideological context partly defined by ethnic parameters. In the present paper I shall briefly trace, and contrast, the ethnic aspects of the language situation in two contemporary African communities: the Nkoya of central western Zambia, and the Kalanga of northeastern Botswana. The choice of these two cases is inspired by more systematic considerations than personal preference alone: while my own current anthropological and historical research happens to concentrate on these two communities, their choice here is strategic in that — looking at their linguistic, ethnic and political situation within their respective nation-state — they are rather comparable, yet display striking differences with regard to the role language has played in their respective processes of ethnicisation in the twentieth century. Thus the comparison is hoped to have heuristic value in that it highlights some of the crucial variables informing the interplay between language, ethnicity, the state and development, even though, of course, a two-case comparison can never in itself yield viable generalizations.

The comparative empirical data concerning the two languages, and the ethnic groups of the same name which focus on these languages, are compiled in a elaborate matrix (Appendix). It has the disadvantage of being unreadable but the advantage of accommodating more information than could be presented in a short article. Against this descriptive background, I may be forgiven if my discursive argument turns out to be somewhat selective.

My argument is set, implicitly, against the background of studies of ethnicisation and inter-ethnic relations in Zambia and Botswana. While the relevant literature on Zambia is considerable (including classic studies in this field by Mitchell and Epstein),[2] the multi-ethnic dimension of contemporary Botswana society has been very much ignored by scholarship. Researchers have themselves internalized the image of a peaceful, ethnically and linguistically homogeneous, thoroughly Tswana country — an image propagated by the national elite under conditions of Tswana linguistic and cultural hegemony. The notable exception, of course, is a considerable attention for the plight of the Khoi-San (locally called Sarwa), under conditions of social humiliation and economic exploitation at the hands of the Tswana.[3] The claim of Botswana’s homogeneity has also been accepted by linguists (e.g. Alexandre 1972: 89) and can still be found expressed today:

‘The rest of the Southern African states are very much less interesting because they are typical/normal anglophone African countries. Perhaps the only noteworthy feature is that (without talking English into account) all three of them (i.e. Botswana, Lesotho and Swaziland) are virtually monolingual.’ (Mkangawi 1991: 10)

The language we call Nkoya today (with its constituent dialectical variants such as Nkoya-proper, Mashasha, Lushange, Lukolwe, Mbwela), now a scattered minority language in central western Zambia with c. 30,000 speakers, is generally accepted to have been a language of people who were relatively early to trickle down — like so many others in the past half millenium — from southern Zaire into the savanna of South Central Africa as from c. 1500 AD.[5] On the strength of political and cosmological notions deriving from the Zairean homeland (Kola), some of these immigrants soon began to involve the local population (partly mere earlier immigrants) in a process of state formation, which — along lines which shape up quite distinctly in recent research — from the late eighteenth century, if not earlier, led to a number of small polities in which Nkoya was the court language. Most probably, the language and the people identifying by reference to it were early on known not as Nkoya but as Mbwela. The origin of the name Nkoya itself remains somewhat obscure: it is associated with a forested area near the Kabompo/Zambezi confluence, later became the toponym for the entire region (roughly coinciding with today’s Kaoma district) where Nkoya is spoken by the majority of the population, and may well be a dialectal variant of the magical name of Kola itself. At any rate, our first record of its use for the political elite of one of these polities dates back to c. 1840: in the praise-name under which a female ruler, Mwene [6] Komoka, acceded to the major Mutondo royal title. Only a few years later these polities in the eastern fringes of what later (e.g. in Max Gluckman’s famous anthropological studies) became known as Barotseland, were made tributary to the Kololo state, by which immigrants from what is today South Africa had supplanted the earlier Luyana administration. The original Luyana ousted the Kololo immigrants again in 1864 but largely retained the latter’s southern Sotho language, amalgamating it with their original Luyana to form today’s ‘Lozi’ language. It was in the context of political incorporation into the Lozi state that ‘Nkoya’ (in its Lozi form Mankoya, which was also extended to become the name of a district capital, to be renamed Kaoma in 1969) became the name of one particular Lozi ‘subject tribe’ and of the latter’s language — myopically uniting, under this Lozi-imposed label, not only a certain dialectical variation but also several incapsulated polities who has never before identified as ‘Nkoya’. Favoured by the colonial state which was imposed in 1900, Lozi administrative and judicial subjugation, social humiliation and economic exploitation of the people in the eastern Barotseland fringe even increased during the colonial period. While the Luvale (another ‘subject tribe’, to the north of the Lozi core area) were allowed to secede from Barotseland and form a district of their own, Lozi colonisation of Nkoyaland went on through the creation of a Lozi court at Naliele near Kaoma in the 1930s, where the son of the Lozi Litunga (King) was put in charge of the newly created Mankoya Native Authority; Mwene Mutondo Muchayila who opposed these developments, was ousted from office and for ten years (1948-1958) exiled to a remote part of Barotseland — only to return to office in the years 1981-1990. Under the unifying impact of this shared negative experience within an overall administrative and political framework, it was in the period around World War II that the name ‘Nkoya’, now reflexively used by the people themselves, became a rallying cry of an increasingly comprehensive ethnic identity facing a common perceived ethnic enemy, the Lozi, whose language as used in the Lozi indigenous administration including the courts had become a main instrument of control and humiliation.

Nkoya was a minority language in the Barotseland Protectorate, whose indigenous administration retained considerable autonomy under colonial rule. Meanwhile, throughout Northern Rhodesia (now Zambia) of which Barotseland formed part, seven languages had come to be recognized by the state as vehicles of formal education, broadcasting, the judiciary, and state/subject interaction: Bemba, Tonga, Nyanja, Lozi (throughout Barotseland and in the region of Livingstone, the early colonial capital in the south), Lunda, Luvale and Kaonde. At Independence the colonizer’s language became the country’s official language. For fear of ‘tribalism’ and in the service of ‘nation building’, in the first decades after Independence strictly no other language but English was used in state-citizen communication — which went to the extent of President Kaunda addressing crowds in his native Chinsali district not in the local Bemba language but in English. Nkoya found itself among the sixty-odd languages or dialects in Zambia to which no official status was accorded. Thus the Nkoya language had become doubly peripheral: a minority vis-a-vis Lozi which remained dominant in Barotse (later Western) Province in most formal situations including education, local government and the courts; and vis-a-vis English. Peripheral also in terms of participation in the modern economy, minimum access to national markets of labour, produce and power (while the regional ones were totally dominated by Lozi speakers), the Nkoya ethnic identity (defined by speaking the Nkoya language, and by allegiance to local chiefs — the encapsulated heirs to the independent polities of the 18th and 19th centuries) in the 1960s was characterized by great resentment of continued Lozi domination and by rejection of the independent nation-state of Zambia which (from the parochial perspective of Kaoma district) had allowed itself to be captured by the Lozi aristocracy.

Circulation of people of great distances has been a constant feature of Nkoya rural society. It is an essential feature of the social organization that young men and women continually move from village to village in search of kinship-based patrons and spouses, until they become less mobile by middle age. The geographical scope of this intra-rural migration has extended beyond the areas where Nkoya is spoken by the majority of the population, and as a result many Nkoya were and are bilingual or trilingual in the languages of western Zambia. Meanwhile, the local language and ethnic situation has considerably diversified since the beginning of the twentieth century. Lozi domination had facilitated the immigration of Lozi speakers to the fertile and well-watered, scarcely populated lands of Nkoya. Angolan immigrants (speaking such languages as Luvale, Luchazi, Chokwe and Mbundu and ethnically identifying by these same names) likewise flooded the region as from the late 1910s. As a result Nkoya soon became a minority expression even at the newly-created district capital. The influx of immigrants (whose agricultural and hunting methods tended to be more modern and aggressive) for the first time in history created pressure on the local land. Incapsulated within the Lozi indigenous administration (which moreover controlled part of Nkoyaland directly, through Lozi indunas[7]), Nkoya chiefs were unable to curb this invasion. After Independence (when their power was further eroded by the institution of Local Courts over which the chiefs had no longer formal control)[8] the selective granting of land to ethnic strangers was adopted by the chiefs as a means towards prestige and additional income. Appointed as members of the new Rural Council which after Independence supplanted the Lozi-controlled Mankoya Native Authority, the chiefs facilitated a major development project, which in the eastern fringe of Kaoma district led to a massive agricultural scheme attracting thousands of ethnic strangers from all over western and southern Zambia in what was to become the new rural town of Nkeyema. Not only did this further diversify the local language situation; it also confirmed the Nkoya as linguistic and economic underdogs, serving — usually in a language other than their own — immigrant farmers on their own lands as casual labour; or pursuing, in their nearby villages (and then in their own language), the meagre yields of an eroded historical agricultural production system which because of the depletion of the forests could less and less be supplemented by the time-honoured techniques of hunting and collecting.

In this linguistic, ethnic, political and economic desolation the local Nkoya-speaking groups, as from the 1920s, found an ally in Christian missions and (since the local Catholic Mission was rather Lozi-orientated) especially in the fundamentalist evangelical South African General Mission, which shifted to Nkoyaland from an increasingly hostile Angola. In its wake, the mission brought Mbundu immigrants to the district, thus contributing to its ethnic and linguistic differentiation; however, in the context of this mission, ethnic strangers would adopt Nkoya as a second language. Establishing excellent relations with the Nkoya chiefs, the mission pioneered in Nkoya literacy, published school primers, had hymns and part of the Bible translated into Nkoya,[9] and was largely responsible for the creation of a climate in which peasants would go about their Christian activities and their social contacts (especially in the form of letters to the many relatives who were temporarily absent as labour migrants) in their native language. A remarkable form of ethnico-religious discourse emerged, in which local Christian leaders would also be the articulators of the budding Nkoya ethnic identity, and improvised prayers in Nkoya would mix pious and political elements in fervent evocations (full of predictable biblical parallels) of their ethnic plight at the hands of the Lozi. In this context the first Nkoya pastor, Rev. Johasaphat Shimunika (1899-1981), nephew and son-in-law of the first Mwene Mutondo to be baptized, not only was largely responsible for Bible translation but also collected Nkoya oral traditions, collating them into passionate statements of Nkoya ethnic identity and anti-Lozi manifests, which circulated among the Nkoya since the late 1950s. His main work, Likota lya Bankoya (The history of the Nkoya people), was recently published by me (van Binsbergen 1988, 1992a: parts II and III).

Away from their rural homeland, Nkoya labour migrants had no choice but to reproduce the peripherality which was their fate at home. A few years of mission education in Nkoya hardly compared with the splendid educational facilities, in the empowering languages of Lozi and English, which the Lozi aristocracy had managed to attract and develop in the centre of Barotseland. The Nkoya’s small numbers and lack of specialized skills made it impossible for most of them to capture substantial portions of the urban labour markets, which were controlled by the Lozi and other dominant ethnic groups. Occasional urban success often involved ethnic and linguistic passing, dropping the (still only emergent) Nkoya ethnic identity for that of Lozi or Bemba, and loosening the home ties. The majority of the many Nkoya labour migrants however remained insecure strangers in town, and continued to rely heavily (in times of unemployment, illness, bereavement and personal conflict) on such security as the intensive (and costly) cultivation of rural ties would accord them. Regrettable as this state of affairs may be judged from a perspective as personal achievement in modern, capitalist relations of production, this state of affairs was largely responsible for the continued vitality of Nkoya rural society. With the persisting investment, on the basis of migrants’ access to cash, in rural-based institutions (kinship, marriage, chieftainship, old and new cults of affliction), Nkoya rural society remained the relational, symbolic and therapeutic power-house of dispersed Nkoya-speaking individuals, and thus a viable basis for an increasingly vital Nkoya ethnic, linguistic and cultural identity.

Twenty years ago, when I started research among the Nkoya, the lack of recognition of their language was still felt by them as the major sign of their powerlessness at the national and regional level, which they interpreted exclusively in terms of Lozi oppression. Primary school teaching was no longer in the hands of the mission but had become the responsibility of government, and as a result it took place in the recognized language of Western Province and Livingstone, Lozi. Very few Nkoya primary school graduates found their way to secondary school and fewer still matriculated: largely because educational success depended on the mastery of two languages (Lozi and English) hardly used in the Nkoya rural milieu, but also since schoolchildren’s attendance was poor both in numbers and in frequency — due to the Nkoya’s emphasis on boys’ hunting and musical skills, and girls’ domestic chores and puberty ceremonies. In the collective Nkoya consciousness a large and sinister place is occupied by a district educational officer, inevitably (like the majority of local teachers) of Lozi identity, who allegedly rounded up and burned virtually all Nkoya school primers. Neither was the Nkoya language used in any of the provincial or national media. Since the early-19th century the Nkoya royal orchestra has been a standard element in court culture all over western Zambia, and as a result the folklore programmes of Zambia Broadcasting Corporation often featured Nkoya songs. Requests of specifically Nkoya-language programmes, however, were systematically turned down with reference to the country’s formal language policy. Nkoya-speakers only occupied the most lowly jobs at the district headquarters and UNIP party office, and any dealings between Nkoya villagers and the outside world, for administrative, medical or judicial purposes, would have to take place in Lozi, of which only half the adult men and very few women had more than a smattering. Political meetings, too, had to be conducted in Lozi or English. At one, held in preparation of the 1973 national elections which for the first time brought a Nkoya into Parliament,[10] the District Governor (of eastern Zambian extraction), when challenged why the meeting could not have been held in Nkoya, spoke out in anger:

‘This nonsense has to stop. Chief, you must control your people. There is no Nkoya. Nkoya does not exist!’

Various processes combined to change this situation substantially in the course of two decades, even if Zambia’s language policy formally remained the same.

The integration of the Barotseland Protectorate, its traditional ruler the Litunga, and the Lozi aristocracy, into the independent state of Zambia had been difficult, and had had to be bought at the onerous conditions of the Barotse Agreement. One section of the Lozi aristocracy had promoted UNIP in Barotseland, which had been a reason for many Nkoya to side with UNIP’s rival, the African National Congress (ANC),[11] in addition to short-lived political organizations of a specific Nkoya signature. The commitment to the struggle for independence had been massive among the Nkoya, not so much for disgust with the colonial state (whose blessings were to form a standard topic of conversation among the Nkoya in the post-independence period — despite the fact that Lozi domination had been greatly reinforced by the colonial state) but for the hope that independence would bring the end of Lozi domination. Until the late 1960s the Lozi played a major role in the successive factional coalitions around which the Zambian national politics revolved. The prohibition of the habitual labour migration from Barotseland to Rhodesia and South Africa increased ANC sympathies among the Lozi, at a moment that UNIP was already contemplating one-party rule. The Lozi were outwitted of their political power at the national level, and UNIP found in the Nkoya welcome allies in an otherwise hostile province.

The Nkoya’s ethnic claims of access to regional and national representative bodies, restoration of the prestige of their traditional leaders, and increased development efforts in their area, were met to a considerable extent, and such few Nkoya as could be considered to constitute a traditional and modern elite (e.g. the royal chiefs, and Mr Kalaluka) soon found themselves in a position where, as brokers between the modern world and the local villagers, they could combine ethnic mobilization with personal economic and political advancement. UNIP branches, and ward and village development committees mushroomed, and for the first time the repertoire of UNIP political songs was translated and sung in the Nkoya language. Political meetings in favour of the ruling party were locally held in the same language. The increased economic opportunities in Nkoyaland increasingly contrasted with the bleak situation of many Nkoya migrants in the declining economy of Zambia’s towns, and people began to re-migrate home.

Moreover, at the national political and ideological level, the earlier universalist insistence on English and fear of ‘tribalism’ gradually gave way to considerations of authenticity and pluralism, and the more the impoverished and disintegrating Zambian state proved unable to mobilize popular support on the basis of services and benefits extended to the citizens, the more passionate and desperate the appeal to a composite cultural heritage to which each ethnic and language group was now agreed to contribute, even outside the established happy few of the seven state recognized languages.

While the political acceptability of the Nkoya language increased, at the major Christian mission establishment in Kaoma district the work on the translation of the Bible continued steadily. Largely under the supervision of Rev. Shimunika until his death, and carried on by his former associated since, draft translations were made of th entire Old Testament, and these have been discussed at general conferences which the church organized in Kaoma and Lusaka in the late 1980s. The text has been ready for publication for some years now. Although a subscription campaign has been launched, funds are still lacking to place a print order.

While this translation work, and the enthusiasm it generated over the years, certainly testifies to the vitality of the Nkoya language, the organizational framework for the text consultations was no longer exclusively that of the mission and of the Evangelical Church of Zambia which it has engendered. In fact, the editorial process of the Bible translation in recent years, as well as similar consultations in the context of my edition of Rev. Shimunika’s Likota lya Bankoya, took place within the context of a new Nkoya ethnic association.

Ethnic associations had thrived in Northern Rhodesia but had been discouraged after Independence. After 1980 they became viable again, and with the restored ethnic pride, the return of educated manpower to the rural homeland, and the fruition of the ethno-historical seeds which Rev. Shimunika had sought to plant for so many years, the time was ripe for the Kazanga cultural society to be initiated in the early 1980s. The society derived its name from an ancient Nkoya institution, the king’s first-fruits festival, which (partly because of the connotations of ritual murder which it shares with all royal ceremonies in the Nkoya context — among others) had hardly been held in the twentieth century. While continuing (in vain, so far) the campaign for the Nkoya language in the media and schools, while joining hands with Nkoya politicians in their attempts to further the cause of Nkoya chieftaincy, and formalizing an economic and social support structure for urban-rural migrants on a modest scale, the society’s main project was to develop a newly ‘bricolaged’ form of kazanga as an annual festival, bringing together all Nkoya chiefs (especially the four royal ones, who under historical conditions would scarcely meet but observe a strict avoidance, each in his own area), and presenting to the crowds of urban and rural Nkoya, other locals, government officials and hopefully tourists, a densely packed programme encompassing the entire (if slightly orchestrated, folklorized, and electronified) repertoire of Nkoya music and dance (van Binsbergen 1992b).

Thus the festival was to form the Nkoya answer to the famous Lozi Kuomboka ceremony, which has attracted large crowds since the beginning of the twentieth century. At the second Kazanga festival, in 1989, the triumph of the Nkoya language could hardly have been more complete: not only did the junior Minister of Culture, Lazarus Tembo (of eastern-Zambian background, once Zambia’s most popular folk singer, and a blind man), attend in his official capacity, but he seized the opportunity to be the first-ever high-ranking state official to address a local crowd in Nkoya — mispronounced and apparently off the cuff, but in reality touch-read from the braille notes hidden in the Minister’s pocket. The previous night the state had declared a 100% devaluation of the Zambian Kwacha, and villagers who later that week went shopping at the district capital had to return to their homesteads empty-handed since their money could no longer buy even what little was available in the shops. But the state could not have chosen a more effective way to impress the Nkoya with, in Mr Tembo’s words, ‘how much we have to be thankful of.’

In October 1991 the Kaunda era came to an end when UNIP lost the national election to the new MMD coalition party, and Mr. F. Chiluba became state president (Baylies & Szeftel 1992). The Nkoya of Kaoma district were divided. In his last year of office, President Kaunda had successfully intervened to protect Nkoya chieftaincies against the Litunga’s mounting aspirations, and this is a major reason why UNIP is still a remarkable presence in the area. However, especially among the peasants there was and still is considerable support for MMD. The new administration offered new national-level opportunities to politicians from the area, including some who are full Nkoya, and those who are not make a point of expressing themselves in Nkoya and to further the cultural and traditional-political aspirations as articulated by the Kazanga society. Gradually shedding their underdog image, the Nkoya are becoming increasingly apt at the situational manipulation of their ethnic identity at the regional and national level, and begin to command considerable political resources. At the regional level, ethnic antagonism now occasionally gives way to a more comprehensive ideology of ethnic solidarity between the groups of western and northwestern Zambia — as against the ethnically dominant centre and especially north (Bemba, Aushi), on which the Chiluba administration leans heavily. Although it is still too early to make predictions, it does look as if the upward movement of the Nkoya identity in the 1980s will continue under the new regime.

Like the Nkoya language, the western Shona dialect cluster known as Kalanga and today extending from northwestern Zimbabwe all the way into the North Central and North East districts of Botswana (where it mainly exists in the form of the Lilima dialect) boasts a considerable local presence. While much of the history of this language and of the ethnic group who identifies by it remains to be written,[12] it is a well-established fact that Kalanga, already called by that name, was the state language of the Changamire state which in the late seventeenth century succeeded the Torwa state; the latter produced the archaeological complex known as the Khami culture, and was historically closely associated with the earlier extensive state system centring on the famous site of Great Zimbabwe.

When as an aspect of the Zulu expansion the Changamire state was supplanted by the Ndebele state in the early nineteenth century, Kalanga speakers lost their association with dominant political power. The southern part of the Kalanga area then found itself in the overlapping and competing spheres of influence of the Ndebele state in the northeast and one particular expanding Tswana polity, to be known as Ngwato, to the south. While these powers were more or less in balance, the relative no-man’s land on the Tati river became a major area for White prospecting and mining, agricultural enterprise and urban settlement: the Tati district, later known as the North East district, focussing on the new town of Francistown. Land alienation and in general the implantation of the capitalist mode of production went on here on a scale equalled nowhere else in the Bechuanaland Protectorate during the colonial period. Attempts to annex the Protectorate as a whole for South Africa failed as much as attempts to incorporate the Tati district into the Southern Rhodesia of which it was so reminiscent. After Botswana’s Independence (1966) the Botswana/Rhodesia boundary from an administrative formality became rather more impenetrable, and under UDI, during the Zimbabwe war of liberation, and during its violent aftermath in southwestern Zimbabwe (when local Kalanga suffered along with the Ndebele under the ZANU state’s aggression) the experiences, and political and cultural concerns, of Kalanga on either side of the border more and more diverged. Yet massive emigration of war and post-war refugees, dispelled by violence in Zimbabwe as much as attracted by the post-Independence economic boom of hitherto tranquil and rustic Botswana, kept the lines of contact open.

In at least one respect the Zimbabwean Kalanga immigrants found an unpleasantly familiar situation in Botswana: their ethnic and linguistic identity made them, along with the original Botswana Kalanga, stand out as politically and socially suspect in a country which for fear of appearing disunited, emphatically proclaimed to be a monolithic Tswana state — by the adoption of Tswana as its national language, by its ruling party’s (BDP — Botswana Democratic Party) populist imagery centring on the Ngwato royal family (whose one-time heir apparent, Sir Seretse Khama, was to be BDP’s leader and the country’s first president), and by its very name of Botswana, i.e. ‘Tswanaland’. In Botswana, Kalanga is very much a minority language, in which no formal education is offered, which is not used in the media, practically not admissible for use in courts of law except in outlying villages, and with hardly any published material circulating in that language.

The Kalanga (comprising c. 120,000 speakers or 13% of the population (Picard 1987: 5) constitute the largest non-Tswana-speaking group in the country, but by no means the only one: e.g. in the north, northwest and west Mbukushu, Yei, Koba, Ndebele, Subiya, Herero etc. defied ethnic and linguistic classification as Tswana, and so did the Khoi-San (called by their Tswana name ‘Sarwa’) scattered all over the country. The Kgalagadi are a borderline case in that their language is similar to standard Tswana but as a separate branch of the Sotho-Tswana peoples they are not counted among the eight constitutionally recognized Tswana groups,[13] and they share with the Sarwa a history of serfdom and humiliation at Tswana hands (Gadibolae 1985; Mautle 1986).

Under the Protectorate the Tswana had formalized a model according to which the entire country’s territory was neatly parcelled up among themselves, each ‘tribal’ area administered by a hereditary chief. Consolidating the realities of Ngwato expansion in the second half of the nineteenth century, the area where the Kalanga lived (with the exception of most of the North East district, which had become freehold land of the Tati Company) fell under the kgotla (‘tribal’ headquarters, court) of the Ngwato chief (kgosi). Kalanga traditional authorities were incorporated in the Ngwato indigenous administration as mere village headmen (sing.: kgosana, ‘little chief’). At Independence the Tswana chiefs’ constitutional and juridical status was re-defined as complementary to the modern central state and its democratic institutions. A House of Chiefs was instituted as the apical structure of tribal administrative and judicial organization, and in terms of the Constitution (Republic of Botswana 1983) only senior members of the Tswana tribal administrations qualify for membership. Kalanga activists read into this section of the Constitution a denial of groups and languages within the national territory, other than the eight Tswana groups explicitly mentioned by name.

In that part of southern Kalangaland which lies in present-day Botswana, the influx of relatively small offshoots of other Tswana groups (primarily the Khurutshe, since the late eighteenth century, and the Rolong in the early twentieth century)[14] and of non-Tswana (especially Kalanga, Ndebele and Tonga) recent immigrants from the north and east, had turned the ethnic and linguistic situation of northeastern Botswana into a complicated mosaic. The Kalanga ethnic identity and language, which had such a long local history, had a considerable but not total attraction for these immigrant groups: Khurutshe in the village of Ramokgwebane, and Rolong in the nearby Moroka, soon adopted Kalanga identity, whereas the offshoots of the same groups in Makaleng, Tonota, Matseloje and Borolong retained their original ethnic identity and their Tswana tongue (Schapera 1952; van Waarden 1988; Malikongwa & Ford 1979). The Khurutshe kgosana of Makaleng came to represent the local population, including the Kalanga, in the Ngwato indigenous administration and in the House of Chiefs. This means that the Kalanga were and are not represented, in their own right, in the far from nominal traditional political structures of the country (cf. Gillett 1973; Silitshena 1979).

Especially in the second quarter of the twentieth century, under the rule of the regent Tshekedi Khama (Seretse Khama’s paternal uncle) Ngwato overlordship in northeastern Botswana was resented and often challenged, especially over church matters (Benson 1960; Chirenje 1977; Wylie 1991). Not unusual in Protectorate Botswana, the Ngwato administration did not permit any Christian diversification and upheld the monopoly of, in this case, the London Missionary Society. African Independent churches, therefore, which were already flourishing in South Africa where thousands of Botswana labour migrants became acquainted with them, in this part of Protectorate Botswana inevitably acquired overtones of ethnic and tribal defiance of Ngwato dominance. The Tati concession, however miserable in other respects, offered a White-controlled sanctuary from Ngwato rule, and it is here that Christian Independency first flourished in the country. In the historical consciousness of contemporary Kalanga in Botswana much is made of the high-handed way in which a particular immigrant Kalanga group around John Nswazwi, defying Ngwato overlordship both in religious and in tributary matters, was beaten into submission by Tshekedi’s regiment in 1947.[15]

The Kalanga’s reliance on agriculture rather than animal husbandry made their children more easily available for such schooling than, for instance, the Tswana, whose school attendance had to be balanced against the need to herd cattle. In the Protectorate period, ideas and people moved freely between Bechuanaland and Rhodesia, and while educational services (or any other services to be provided by the colonial state and the indigenous administrations it upheld) were kept at a minimum in the Protectorate, Christian missions in nearby Rhodesia were flourishing, translated the Bible into Kalanga, and offered a great many Kalanga both the formal education and the ideological outlook on the basis of which to advance in colonial society, while increasingly challenging the premises of inequality on which that society was based (Bhebe 1973). The great Zimbabwean politician Joshua Nkomo, for instance, is very much a product of this situation (Nkomo 1985). But so are others (e.g. Mssrs K. Maripe, T. Mongwa, P. Matante, D. Kwele) who later, as commercial entrepreneurs, Kalanga ethnic activists and national-level politicians, were to play a prominent role in the modernizing and highly proletarianized situation of Botswana’s northeast, with its rapidly growing town of Francistown. After Christian Independency, Francistown became the cradle of the first major independence party, the Botswana People’s Party, which from the start was highly critical of Tswana ethnic, administrative and linguistic hegemony (cf. Nengwekhulu 1979; Murray et al. 1987).

It testifies to the complexity and situationality of ethnic identity, that most of these leaders could also, and did, adopt other, non-Kalanga, idioms of mobilization. Nkomo could identify as Ndebele as much as as Kalanga, and it is in the latter identity that he gained world-wide renown. Maripe completed a doctorate in industrial relations in Belgium and, long before gaining local prominence as a Kalanga novelist and as BPP president stood out as a trade unionist active not in Botswana but in the Federation of Rhodesia and Nyasaland (Meebelo 1986: passim); in other words he could have identified, and probably did at one time, as Zimbabwean and even Zambian. Mongwa is Pedi as much as he is Kalanga. Matante was prompted to form the BPP on the inspiration of his membership of the South African African National Congress (ANC), as was moreover active as the leader of an independent African church in Botswana; so we might have heard from him as a South African black politician or as a minister of religion. All this reminds us of the fact that appeal to an ethnic idiom in the context of formal, national-level politics is not the expression of the primordial attachments ingrained through socialization in early childhood — as first-generation studies of ethnicity in Africa and the Third World in general tended to stress (cf. Geertz 1963) — but is often a deliberate and strategic choice of a particular political instrument, identity and career from among alternatives.

After Independence, Botswana rather unexpectedly saw an economic boom, largely based on the diamond industry (in which South African capital and expertise was wisely matched with Botswana state control) and the beef export industry, against the background of open economic relations with South Africa which were guaranteed by a Customs Union putting Botswana (along with Swaziland and Lesotho) in an awkward but economically favourable position among the Southern African front-line states. The BDP government, which gained power democratically in the drive for Independence and retained it ever since, therefore had plenty to offer to the Botswana state elite and to the population at large, and prudently but consistently delivered enough to ensure stability, economic progress and popular support. In the process, the multi-party system was nominally encouraged and gained the country international esteem and donor support. In reality however, with every national election which was held at the constitutionally stipulated times, the impotent opposition parties — including the BPP and the BNF (Botswana National Front) increasingly became an ornamental fringe to a de facto one-party, populist and rather authoritarian political regime (cf. Picard 1987; Holm & Molutsi 1989). Repeatedly, when the outcome of democratic elections had led to opposition majority at the district and town-council level, the dilution of representative bodies by state-appointed BDP representatives, and the persuasion of elected opposition representatives to cross the board to BDP while retaining their seats, proved standard tactics to retain or regain BDP control.

This situation was not entirely unlike Ngwato/Kalanga relations in the nineteenth century and under the Protectorate: occasional and dramatic Kalanga challenge of Ngwato hegemony did not preclude that the ordinary, and widely accepted, situation was one of peaceful accommodation, where the Kalanga, as ‘Northerners’, had their assigned place in the Ngwato polity, not only in distant homogeneous Kalanga villages at a distance from the Ngwato capital, but also in the Ngwato heart-land, even in specifically Kalanga wards at the capital (Schapera 1952, 1984, 1988). A remarkable contradiction can be observed on this point. Challenge of Tswana hegemony and explicit proclamation of Kalanga identity were retained and became more and more bitter as standard expressions of political opposition to the BDP. That political opposition in the struggle over state control had to be phrased in an ethnic and linguistic idiom was also due to the fact that such religious and class oppositions as had unmistakably arisen, at the level of people’s consciousness were still not sufficiently articulate to serve as basis for mass mobilization. In independent Botswana, a class idiom is mainly propagated by the BNF, in intellectualist Marxist terms which fail to attract mass support. Of course, the unsettled nature of class contradiction as a basis for mass mobilization has, until quite recently, been a general theme in post-independence politics throughout Africa.

In fact, however, the Kalanga’s relative educational and entrepreneurial success had led to a situation where a disproportionately large percentage of BDP politicians at all levels (including Cabinet Ministers and MPs) happened to be Kalanga, who as a condition of political eligibility and respectability played down their Kalanga identity and allowed Tswana ethnic and linguistic hegemony in the country to go unchallenged.

Thus the very people who, being affluent and relatively well-educated, might have been involved in the production and consumption of Kalanga symbolic culture (in the form of literature, drama and ethno-history) inside Botswana, tended to have vested interests not to do so. Maripe’s Kalanga novels are nowhere to be bought in Botswana. Copies of the 1929 Bible translation in Kalanga for years could be seen to rot on the shelves of the Francistown Bookshop along one of the town’s main shopping streets. The collecting of Kalanga oral-historical traditions, folklore and proverbs was largely left to foreign researchers and had no market inside Botswana. It is commonly believed that it would even be a legal offence to publish books in Botswana in any language other than English and Tswana; not being a jurist, I have however no information on jurisprudence actually limiting down, in this specific sense, the section of the Constitution dealing with the freedom of expression. And the insights which modern scholarship, mainly on the basis of Zimbabwean material, gained in the splendour and historical depths of Kalanga history, highlighting its intimate link with the glorious Zimbabwe state and the widespread Mwali cult which is among Southern Africa’s major religious expressions,[16] so far never managed to be fed back into the publicly articulated ethnic consciousness of the Botswana Kalanga. There is an amazing contrast between the objective riches of Kalanga history, and the poverty of the Botswana Kalanga collective historical consciousness which seldom reaches beyond the Nswazwi episode, never taps the sources of ethnic pride history has so abundantly to offer, and even reproduces the erroneous Tswana view [17] that the Kalanga in Botswana are mere recent immigrants enjoying, but dishonouring, Ngwato hospitality! Such inspiration as could have been derived from ethnic identification with the Zimbabwean Kalanga across the border [18] seems scarcely to have been tapped after Independence. While along personal kin networks assistance was offered to Kalanga victims of the Zimbabwean war of liberation and its atrocious aftermath (mainly in the form of accommodating illegal immigrants into Botswana), at the public and national level the border communities went out of their way to dissociate themselves from such violence as spilled across the border, stressing — not always spontaneously — that their first allegiance was to the Botswana state and not to an international Kalanga ethnic identity.

The increasing entrenchment of Botswana Kalanga within the national territory of Botswana was one of the resons that prompted, in the mid-1980s, the Kalanga Bible Translation Project, headed by White Lutheran missionaries recruited from Germany and the USA, and with strong organizational backing from the Lutheran mission in South Africa. Justifications of the project included the incomplete nature and linguistic defectiveness of the existing translation from 1929, whose orthography, moreover, was judged inadequate. Draft translations are undertaken in the Project Office variously based in the town of Francistown (1987-88), the village of Zwenshambe (1988-89), and Francistown again (1989-). The actual translation work is mainly in the hands of Kalanga native speakers of post-secundary education, with advisory committees throughout the Botswana Kalanga area. Attempts to expand the project into a revitalisation of Kalanga language and culture in general have so far not taken wing, and the project has been severely hampered by conflicts over external, White control, over conditions of service, and over the conflicting national cultures and management style of the various missionaries involved. So far the project’s main achievement has been the development of a new standardized Lilima orthography in consultation with Botswana native speakers.

In the course of the 1980s Kalanga ethnic and linguistic identity developed into a major issue, but more so within the Kalanga community than between Kalanga and Tswana. Students at the University of Botswana founded the Society for the Propagation of the Ikalanga Language (SPIL), with amazingly little government opposition: the required registration with the Registrar of Societies was almost a matter of routine, even though this was to be the only overtly ethnic association in the country and one obviously not propounding Tswana hegemony.[19] On various occasions SPIL branches and individual members (never the entire association as such) challenged the Kalanga members of the government to speak out in favour of the teaching of other African languages than Tswana, and to identify publicly as Kalanga instead of submitting to Tswana hegemony. This led to major outcries, in which the official position with regard to the exclusive use of Tswana for the sake of national unity and efficiency was repeatedly expounded and defended, and local BDP politicians clamoured — in vain, so far — for the prohibition of SPIL.

However, so far the SPIL initiative has failed to engender massive ethnic support. It has remained a pastime of middle-class people whose command of Tswana usually equals that of Kalanga, and whose children often have been raised to have Tswana or English as their first language. With the exception of the few and far between issues of the mimeographed journal Tjedza [Light], the society can boast little literary or cultural production. Its annual meetings are enlivened by performances of dance troupes but expert and exciting as their repertoire is, their very presence brings out the dilemmas of ethnic mobilization along Kalanga lines in Botswana: traditional dancing is part of the general primary and secondary school curriculum throughout the country, the movements and songs hardly stand out as specifically Kalanga, and to the extent to which elements are borrowed from territorial and possession cults these, too, combine Kalanga references with Ndebele, Venda and even Tswana ones.

My reading of SPIL is that it primarily reflects a struggle, within the Kalanga middle-class community, between those whose acceptance of Tswana hegemony has paid off in terms of political and economic power — in other words has made them share in state power —, and those (typically younger, perhaps slightly better educated and perhaps with slightly stronger roots in their rural home communities and the latter’s traditional leadership) whose access to political and economic power so far has been frustrated and who through insistence on a Kalanga ethnic idiom seek to either capture their own share of state power, or at least to discredit the state, proving it to be less universalist, and more ethnically particularistic, than its constitutional pronouncements would suggest. But these ideological expressions mobilise less support in a context where de facto numerous individual Kalanga have had more than their fair share in the capturing of the Botswana state — even if they could not publicly claim to have done so under the ethnic label of Kalanga.

SPIL emphatically declares itself to be a non-political society. Churches find themselves in a similar position, and it is noteworthy that — even though my Botswana research has come to concentrate on the religious domain — I have no evidence of linguistic and ethnic antagonism playing a dominant role in churches and non-Christian cults in northeastern Botswana today. Multiple and situational ethnic and linguistic identity, code switching, mixture of songs and texts from Kalanga, Ndebele, Tswana (and even Shona and English) in the course of one religious event, and the accommodation of potential ethnic opposition within an encompassing idiom of religious transcendence of disunity, — these are the catchwords to describe the local religious situation.

Having concentrated on urban research, I do not really know to what extent this pattern in the religious domain reflects any relative absence of group conflict along ethnic lines in other spheres of life in the rural communities of northeastern Botswana today. My impression — not supported by extensive field-work — is that there ethnic and linguistic diversity, particularly the Kalanga/Khurutshe distinction of long local standing, has no negative impact on social relations; bilingualism, intermarriage, patterns of residence, and the frequent passing of Khurutshe (of the Mpofu totem) as Kalanga, support this view. However, my extensive participant observation in the urban setting of Francistown shows the religious situation there, as described above, to be a remarkable departure from the ubiquitous, petty confrontations on language and ethnic issues, in relationships between neighbours and between friends, on the work floor, in access to the informal sector of the economy, in amorous matters, in drinking and nightlife, in the conceptualization of social relationships in terms of sorcery. These frictions are clearly reflected in the cases tried at the urban customary courts. As an urban society Francistown is saturated not only with African/White but also with Tswana/Kalanga conflict, where ethnic stereotypes and the failure or refusal to learn and understand each other’s language often adds an awkward dimension to casual interaction between strangers.

In the face of this conspicuousness of inter-ethnic conflict in everyday urban life, it is remarkable that the political effectiveness of Kalanga ethnic mobilization has remained so slight. While closely linked to the Kalanga cause and always vocal on minority language rights (National Executive Council 1984, 1988; Maripe 1987), the BPP’s nationwide aspirations have prevented this party from identifying too narrowly as a one-‘tribe’ affair. Daniel Kwele’s Botswana Progressive Union (BPU) was specifically founded in 19... when his being Kalanga prevented him from assuming the national leadership of the BNF (which has a strong regional backing in the south of Botswana), and among Botswana politicians today his pronouncements were the most militant and unashamedly pro-Kalanga — until his death in 1990. Never really successful, the performance of BPP and BPU in the 1989 national elections was extremely disappointing. There is little to suggest the imminent failure of the BDP strategy of Tswana hegemony, populism, and co-optation of potential opposition, as long as the state elite remains in a position to ‘deliver’. The country’s language policy is likely to remain as it is. In the long term, however, the diminishing diamond resources, the impact of continued drought on the problematic Botswana agriculture, the paradoxically negative effects for the Botswana economy of the dismantling of apartheid in South Africa, increasing rates of inflation and unemployment, the increasing public arrogance of the Botswana military, and the wave of democratic change and popular participation in many other (far less liberal) African countries since 1990, may bring Botswana to a critical point where language and ethnicity may yet be turned into effective political capital (Bernard 1989).

When comparing the trajectory of Nkoya with that of Kalanga as minority languages, underneath the many superficial correspondences considerable differences can be detected.

First the correspondences. Due to the mfecane upheaval which has affected the entire Southern African subcontinent in the nineteenth century, in two regions a Bantu language of considerable local antiquity and both structural and lexical continuity with adjacent languages,[20] is confronted by a Sotho language (of a very different group of Bantu languages, and unintelligible for speakers of Kalanga and Nkoya) originating from the south and carried by a group of such power that it relegates the other local language to minority status;[21] in the colonial and post-colonial period this minority status is formalized in a national language policy and implies exclusion of the language involved from the state’s political and administrative practice; the dynamics of state formation and hegemony, and the introduction of Christianity as a literate world religion, then engender an ethnic consciousness largely focussing on language; the recent political and economic history of the people who carry the minority language casts light on the extent, direction, degree of organization, degree of cultural and ethno-historical elaboration, and success, of the language-centred ethnic strategies of each group.

At this level of generality a similar story could be told for scores if not hundreds of languages and ethnicities in the modern world. Only on closer scrutiny, systematic differences between the Nkoya and the Kalanga trajectory are highlighted, which suggest crucial underlying variables.

Difference in scale is very manifest: the 30,000 Nkoya-speakers constitute less than 1% of the Zambian national population, while the 120,000 Kalanga-speakers constitute 13% of the Botswana population. That the Kalanga national percentage is almost twenty times higher than the Nkoya one cannot be ignored if we want to understand the difference between Kalanga and Nkoya in terms of access to the modern state and economy, education, and privileged class positions. The absolute numbers also point to a marked difference in size in terms as a potential market for missionary efforts, book production and marketing, creative talents to be mobilized in literary production, although it can hardly be said that the Botswana Kalanga have done better than the Nkoya in these respects.

One important aspect relates to the number of hierarchical politico-administrative levels between minority speakers on the ground and the nation-state, and the historical changes this set-up has undergone in recent constitutional history — in other words the definition and historical restructuring of the national political space.[22] Nkoya-speakers in the course of the last two centuries have had to accommodate to two states relegating their language to minority status: first the Kololo/Luyana state, which constituted the highest level of political organization until 1900 (while much of this situation was retained by the Barotse indigenous administration under colonial rule), and whose majority language (in terms not of number s but of power relations) was that of the Lozi elite; and subsequently the colonial and post-colonial state, whose majority language was and has remained English. The Lozi did not manage to perpetuate their Protectorate in the form of a seceded post-colonial state of Barotseland — an option seriously contemplated by many at the time (cf. Mulford 1967; Caplan 1970), and still lurking around the corner today — but had to accept integration in the Zambian state on terms which were increasingly similar to those applying to other regions in the Zambian territory. This meant that the Lozi, after having increased their ethnic and political domination over the Nkoya during most of the colonial period, yet failed to capture the wider post-colonial state; their own Lozi language was relegated, in political and administrative status, from the supreme level as state language to an intermediate level as one of the seven state-recognized regional languages under the hegemony of the official national language, English. It is the decline of Lozi power at the national level that offered the Nkoya, after 1964, room to enhance their linguistic status, their chieftaincies, engage in ethnic organization and cultural and ethno-historical revitalisation, and start, as an ethnic group, on a centripetal movement towards the nation-state and its development initiatives.

From the point of view of the re-definition of the national political space, the Botswana Kalanga trajectory has been fundamentally different: if Tswana domination in southwest Kalangaland (i.e. northeast Botswana) in the nineteenth century and under colonial Protectorate conditions was rather similar to Lozi domination in Nkoyaland in the same period, the subsequent difference between the Nkoya and Kalanga linguistic and ethnic trajectories owes much to the fact that (as a result of international political en economic relations prevailing in the subcontinent since the late nineteenth century, but also because of the undeniable presence of a Tswana language majority over a huge part of what today is Botswana and the Republic of South Africa) no intermediate level emerged: the Tswana did capture the post-colonial state of Botswana, managed to impose their language as the national language for use in its state institutions along with English, retained their hold on the state throughout the post-colonial period by a dextrous utilization of democratic institutions and international esteem, and thus ended up in a position incomparably more powerful than that of the Lozi in Zambia today.

(b) Tswana as just another regional language, at the sub-national level of northeastern Botswana, spoken by Khurutshe locals and southern urban immigrants, on the other.

This contradiction explains some of the inconsistencies in the patterns of ethnic animosity (in everyday urban life), ethnic accommodation (in everyday rural life, if my reading of the scanty evidence at my disposal can be supported), transcendence of ethnic opposition (in the religious domain), antagonism over language in the formal political domain of party politics, and the failure of mass mobilization over minority language issues and Kalanga ethnicity in general. Tswana national political hegemony empowers speakers in certain state-defined situations in northeastern Botswana: BDP politicians addressing a political rally, judges in the urban customary court, civil servants in offices, school teachers, broadcasters; but (because of the effectively constitutional functioning of the Botswana state, which makes it difficult to openly mobilize state power for petty particularistic interests of a personal nature) such empowering is far less relevant in the day-to-day contacts between urban neighbours who are ethnic strangers, and between rural fellow-villagers who know that despite their different tongues they have shared a local history for a century or more. We must resist the temptation to consider ethnic phenomena in the national political domain and those in the local domain as necessarily converging to the same category of ethnicity.

Traditional rulers have played a very different role in the ethnic and linguistic trajectories of the Nkoya and the Kalanga. No matter how much, and for whatever good reasons, the Nkoya turned the Lozi into their main perceived ethnic enemies, the fact remains that some Nkoya royal chieftainships survived the incorporation process, and not with the status of mere village headmanships (as among the Kalanga), but as senior members of the Lozi indigenous administration, sharing (albeit to a lesser extent than their Lozi counterparts, and at the cost of considerable contestation) in the financial proceeds of the 1900 Barotse Treaty and the 1964 Barotse Agreement. They have largely retained their regalia, royal enclosure, palace, subsidized orchestra, paid councillors throughout the colonial period and until today.[23] The creation of the Naliele Lozi court in the 1930s as a tangible expression of Lozi internal colonization was greatly resented, but it provided an administrative continuity when its Mankoya Native Authority became Kaoma Rural Council, with the chiefs (or their Prime Ministers) as appointed members — a situation again persisting today. Thus the chiefs could continue to function as foci of cultural, ethnic, linguistic, and historical identity, and at the same time act (in collusion with their junior relatives: Nkoya modern politicians such as Kalaluka) as political brokers bringing together their people and the state. Among the Botswana Kalanga Tswana domination effectively and early rooted up everything of this nature, so local ethnic consciousness remained without a traditional political focus, which might also have served as a tangible connexion with the splendid political history of the Kalanga over the past 500 years or more.

An important lesson is to be learned here in terms of the appreciation, from a point of view of national integration and development, of the potential role of ethnicity and of traditional leaders as foci of ethnic consciousness. One cannot blindly generalize that all sub-national identity is divisive and leads to centrifugal tendencies away from the state — as has often been maintained both by African politicians and by political scientists. Under specific circumstances, which empirical sociological and historical research has to identify, the road to increased participation in the state and its development efforts leads via ethnicity and traditional leaders — as in the case of the Nkoya. One can only speculate how different Botswana Kalanga ethnicity would have been today if Ngwato hegemony had not eclipsed Kalanga traditional leadership, or if the latter had been restored when the opportunity to do so arose at Independence. My guess would be that in that case Kalanga ethnicity in Botswana might be more vocal and vital today and less of a backroom middle-class pastime, that the status of the Kalanga language would have been higher so that a Kalanga elite could be more loyal to it, and that the Kalanga identity and the Kalanga language would have developed into respectable sub-national expressions with their own recognizably loyal place in the Botswana nation-state, vehicles for qualified integration rather than (as seen today by the Tswana) for centrifugal disruption. It is for profound reasons that the position of Kalanga traditional rulers in Botswana today is analogous to the position of the Kalanga language today: powerless, peripheral, without organic place at the national scene. And just as the 1980s have rediscovered the importance of traditional rulers for an understanding of the contemporary African state and its defects, it is likely that the near future will see the same for minority languages.[24]

Given the three-tiered complexity of the Nkoya ethnic situation (local level/Barotseland/national level) Nkoya ethnicity could grow to become centripetal vis-a-vis the state; given the two-tiered make-up of the Kalanga situation (local level/national level), the Tswana having captured the state could only mean that any Kalanga ethnicity opposing the Tswana, would have to be centrifugal vis-a-vis the state. This is an important reason why language policy in Botswana is inflexible to an extent it did not prove to be in Zambia. Speaking Nkoya does not in the least directly threaten the constitutional premises of the Zambian state, but refusing to speak Tswana can very well be interpreted as a subversive act. Of course, there is plenty of room for semantic mystification here: it is only by sleight of hand that BDP politicians manage to convince their audiences that

(a) whoever is constitutionally a Motswana [25] (which includes all Botswana Kalanga except the most recent immigrants), is

(b) linguistically and ethnically a Motswana (which leaves out thirty percent of the citizens of Botswana), and therefore

Here we are confronted with the highly significant fact, often overlooked in discussions of language policy and ethnicity, that the state has a discourse not only on language, but also in language — so that the premises of elite power are subjectively implied in the very social constructs that, in terms of language policy, are being negotiated under the pretence of disinterested objectivity. Language policy in Botswana may be formally enacted in English documents, but it is largely prepared, thought out, discussed in Tswana.

It is probably due more to naivety than to cunning that semantically twisted syllogisms of the above kind are propounded time and again in the national political discourse in Botswana: the identification of the Botswana state with Tswana ethnicity and the Tswana language is generally perceived by the actors as an social reality, and whoever challenges that reality threatens to destroy all the undeniable blessings which make Botswana stand out as a stable and affluent country in a lost continent (cf. Good 1992). Whoever is loyal to these underlying premises, and shows such loyalty by submitting to Tswana hegemony in public behaviour, is apparently welcome to use whatever language in the privacy of his home. Of course, sophisticated Kalanga activist (like Dr Maripe, or the lawyer Mosojane, member of a Kalanga royal family, BPP presidential candidate in the 1984 national elections, BPP national secretary, SPIL member, and daily confronting the Botswana state on behalf of his clients, in court, in correspondence with the Registrar of Societies, etc.) have on numerous occasions pointed out the constitutional flaws in this widespread political reasoning (cf. Maripe 1987; National Executive Council 1984, 1988), but to no avail. The fact that most of their fellow-Kalanga in the same educational and income bracket have chosen to submit to Tswana hegemony, is a major reason why the non-Kalanga state elite can afford to ignore the principled constitutional argument and perpetuate the existing language policy.

Speaking of the national political space, we also realize that the difference in the Nkoya and Kalanga trajectory has to do with the concrete geographical location of their speakers’ homelands on the map. For the purpose of the present argument we have slightly simplified the Nkoya situation so as to assume that all Nkoya speakers fall within the Lozi territory — ignoring, that is the less prominent royal chieftainchies of Mwene Momba and Mwene Kabulwebulwe outside Zambia’s Western Province; it requires no simplification however to maintain that the entire Nkoya-speaking rural population is found in the heart-land of western Zambia. People speaking languages close to Nkoya are found in southern Zaire and in eastern Angola (Mbwela, Ganguela), but present-day Nkoya speakers, even their ethno-historical specialists, are virtually unaware of these connexions, and — as I have argued in detail elsewhere (van Binsbergen 1992a) — they certainly do not inform their ethnic consciousness. Kalanga, by contrast, is spoken in western and southern Zimbabwe as well as in the adjacent northeastern Botswana. Without going into a discussion of Zimbabwe Kalanga ethnic and linguistic accommodation to the other Shona languages of that country, to Ndebele expansion since the nineteenth century, to Tonga and other trans-Zambesian languages, it is clear that the trajectory of Kalanga ethnicisation, since the creation of the Bechuanaland Protectorate and especially after Independence in 1966, must be seen in the light of their incorporation into an entirely different national political space. Botswana Kalanga define themselves not so much — through inclusion — by reference to Zimbabwean Kalanga (with whom they continue to entertain kinship, marital and ritual ties), but — through opposition — to Botswana Tswana, particularly the Ngwato. This must be part of the explanation why cultural and linguistic production among the Zimbabwe Kalanga hardly seems to filter across the border; why the exciting insights of Zimbabwe history (but relating to a time when the Botswana/Zimbabwe border did not exist, and much of the Kalanga-centred state system lay in northeastern Botswana) fail to be incorporated in Botswana Kalanga consciousness. There is an additional reason on the Zimbabwe side: although Kalanga is a western variant of Shona, and the latter is the majority language (numerically and politically) of Zimbabwe, the Kalanga language’s status in Zimbabwe (and hence its organization and resources) is weak and problematic since it is identified, by the political centre, with anti-ZANU tendencies in the (largely Ndebele) southwest of the country. The Botswana/Zimbabwe border has increasingly hardened into a real boundary, even to the extent of the expansion and growth to considerable local autonomy of Botswana branches of the Mwali cult, which continues to have its centre in the Matopos hills southeast of Bulawayo; the influence of this cult on trans-border linguistic continuity may be limited since its personnel is multi-ethnic and its archaic ritual language is not today’s Kalanga.

The Botswana Kalanga mainly speak the Lilima dialect, which is the standard for SPIL and for the Kalanga Bible Translation Project of the Lutheran church in Botswana. The insistence on a new Lilima orthography which (however arbitrary as all orthography) is emphatically different from the Zimbabwe convention, and the enormous investment in the translation of the Bible into this dialect following this new orthography (even when the New Testament is largely available in a 1929 translation; Ndebo 1985), must be understood as an attempt at manifest localization or Botswana-ization: to claim and define a place for Botswana Kalanga within the national political space of Botswana, with as little reference as possible to a Zimbabwe which, in Botswana eyes, is poor, conflict-ridden, criminal, violent, non-Tswana, in short subject to negative stereotypes. There is a strong element of artificiality and uprootedness in this accommodation to the Botswana political space, and one cannot really be surprised if the exercise has not, so far, led to the general cultural revitalization the White expatriate Lutheran missionaries had expected, Instead the Kalanga Bible Translation Project (with close links with SPIL in terms of personnel) is ridden with conflict over White control and African initiative; as if the opportunity to put one’s ‘own’ language on the Christian and publishing map, condescendingly offered by the missionaries at great expense and effort, is not sufficiently redeeming in its own right.

Bible translation, among the Nkoya and the Botswana Kalanga, is an example of the convergence between ideological expansion of a world religion, and the linguistic and ethnic processes at the local level. The Nkoya case shows the potential, the Kalanga case the limitations of this convergence. What happens to a language when it is committed to writing, pummelled into the desired orthographic shape, scrutinized for its potential to lexically and syntactically convey the alien images of an imported world religion with its alien theological classifications and nuances? Above we have seen how this Christian/ethnic convergence in the language process can be illuminated by the idea of a national political space, in which the language is to claim and fill a specific niche given prevailing political and language-policy conditions. However, the same phenomenon could also (cf. Fardon & Furniss 1991) be described as reification or objectification — at the hands of academic linguists, missionary linguists, administrators and educationalists but also and primarily at the hands of native speakers themselves (cf. Fabian 1986).

In the process, a language is named, standardized, variants become perceived as dialects and subsumed under the general chosen name (a form of hegemony in itself, but one we cannot go into here), and the lexicon has to be deprived of some of its capacity to endlessly incorporate new matter from adjacent language communities: for language defines itself by opposition to other languages at the local scene, so some linguistic forms will have to become marked as one’s own and other as alien. More than any other part of institutionalized culture, language is encoded in formal rules whose infringement almost immediately causes puzzlement, ridicule, rejection or a breakdown of communication among listeners and readers. This capability of encoding and displaying identity or alienness in social interaction, incidentally, must be one of the reasons why, among all possible culturally produced materials, it is primarily language on which ethnicity feeds and thrives. Lexical and syntactic purism is one of the hallmarks of ethnicity. For example, when I edited the Likota lya Bankoya manuscript in consultation with a few readers’ committees of native speakers of the Nkoya language, the latter insisted that Rev. Shimunika’s Nkoya text be cleansed of all anglicisms and Lozi-isms, even if these were totally accepted in the contemporary spoken Nkoya, and the ‘purer’ alternatives were felt to be awkward, obsolete or not generally understood.

It is impossible to assign a definite starting date to the process of linguistic objectification, even with reference to a specific language and access to historical linguistic data. It is difficult to see how any language could maintain such minimum stability and persistence over time without which it would not deserve the term language, if objectification did not already exist in some inchoate form. One thing can be said in this connexion however: the model of the nation-state, in which a unique language coincides with state power, is alien to most pre-colonial African contexts. Eighteenth- and nineteenth-century states in central western Zambia outside the Zambezi flood plain were multi-lingual and multi-ethnic; political power was not linguistically or ethnically marked to the extent it was to be in the Lozi state; and there are indications that, as a result, language boundaries were more fluid — as if the objectification process was still in an early stage. Certainly the objectification process is very much intensified in a context of political and economic incorporation — the very cradle also of ethnicity.

When selected elements of collective symbolic production (e.g. a language; a cult; a vision of the past) are drawn into the orbit of a group’s identity formation in inter-ethnic relations, these elements tend to be objectified to the extent of fossilization. For the social analyst like myself, who is not a linguist, there is even the risk of overlooking the dynamics of language and treating it as if it were an independent variable in the ethnicisation process, rather than being shaped itself in that very process. Both Nkoya and Kalanga are relatively ancient linguistic presences in the region where they are found today. A discussion of the historical linguistics of these languages is beyond my competence, but it is important to realize that we have identified at least two contexts in which objectification is particularly manifest: when which the language becomes a focus of an emerging ethnic identity, and when language is committed to writing for the first time, often in a context of the dissemination of a world religion.

In the objectification process the named, standardized and purified language becomes imbued with an emotive relevance, a socially constructed sense of identity and of opposition vis-a-vis other, rival expressions at the local and regional scene. It is for this reason that I cannot entirely go along with David Parkin’s (1991) otherwise illuminating idea of approaching language objectification in Africa (specifically on the Swahili coast) as a form of commoditification: as if the language becomes a commodity which is cut and dried, strictly demarcated and regulated like an industrial product, freely exchangeable — and disposable — in a market of money, power and prestige.

The ethnicisation process, viewed here as the ideological and organizational response to incorporation in a national political space, inevitably implies alienation: identity only has to be constructed when it has become problematic, in the face of the intrusion of an otherness so massive and powerful that it can no longer be incapsulated by means of the usual mechanisms through which the local society accommodates newly-born members, in-marrying spouses, strayed travellers and other isolated individuals. In the modern world, such alienation often springs from material disowning, in the process of the imposition of alien political power (e.g. the colonial state) and the world-wide penetration of the capitalist mode of production. In the process people lose (give up, often) much of what they then subsequently realize was once their own: a supportive kinship system, expressive art forms, a symbolically powerful and meaningful cosmological order. This is the familiar context for a discussion of commoditification (cf. van Binsbergen 1992b).

Perhaps in the final analysis, the Botswana Kalanga’s underplaying of language, the submission to Tswana linguistic hegemony and the lack of success of Kalanga ethnic mobilisation along language lines, could be explained in terms of the Kalanga language (in the highly proletarianised situation of northeastern Botswana) having become a commodity, easily exchanged for another (Tswana) whose higher market value is undeniable. In such a context one makes a fool of oneself if one were to publicly cherish a despised minority language as a vehicle for literary expression, or evoke the splendour of a medieval state which has long since disappeared and which cannot be linked in other than negative terms to the present dominant group (the Tswana) in the local nation-state.

Perhaps it is also as a result of proletarianisation that Kalanga ethnicity primarily expresses itself in the political domain rather than in the cultural, literary, religious or ethno-historical domain. One of the most disconcerting aspects of symbolic life in Botswana today is that local historic (i.e. African) culture is largely absent from public life; it is only allowed to enter public discourse in the very form (fossilized, commoditified) which Parkin stresses for language in the contemporary context. In order to be acceptable for public consumption (I use this word purposely) historical elements of African rural culture (e.g. the traditional judicial process in the kgotla or village court, which features in official discussions of the very different modern customary courts)[26] have to be selected, taken out of context, deprived of their meaning, reduced to textbook truisms, and then added as harmless ornaments (as duly processed packages of ‘identity’) to a consumerist life-style whose principle reference group is the urban middle class of nearby South Africa, known in detail through the media and personal exposure. The prominence of this pattern among the Botswana population especially in the urban areas is striking, even if we realize that the difference with other African societies is only gradual. The pattern’s contradictions are particularly manifest in the racially-conscious environment of Francistown (cf. van Binsbergen, in press (a)). The problem is not that historical African cultural forms — referring to a past rural order and its contemporary partial survival, real or imaginary — are not there and do not decisively inform people’s ideas and actions, but that they are largely censored out of public reference: they have gone underground, where they are safe from harassment by dominant White culture and its local, predominantly African representatives. The cult of the High God Mwali, the place of ancestors in everyday and ritual life, healing cults, sorcery, ritual violence, divination, the symbolic basis of family life and of production and consumption — only at the cost of rather greater personal commitment and patience than ordinarily required in other parts of the continent is the expatriate researcher allowed a glimpse of the extent to which the ‘Kuwayt of Africa’ shares in a general African cultural orientation. Under such circumstances there is no premium on a traditionalising and historicising symbolic production as a venue for ethnic expression [27] (the path the Nkoya have taken): instead, one confronts the formally organized texture of social power, in the thoroughly respectable, political domain, which is agreed to be organized along modern principles.

All this sounds like praise rather than criticism of Parkin’s idea. Yet, especially in a context of alienation and proletarianisation, powerlessness in the face of an authoritarian state, and the shattering — in modern migrants’ consciousness — of a village-orientated symbolic microcosm, language for the native speaker tends to be the last refuge of owning and belonging, of competence and identity. Cultural reconstruction and revitalisation, such as is at the root of many ethnic movements in the form of literary and ethno-historical production, seeks to rebuild an imaginary world of belonging in order to combat the disowning that characterizes the ethnic groups collective experience in the outside world. But while this is eminently true for the Nkoya (whose language has been objectified, but certainly not commoditified), it is far less true for the Botswana Kalanga. Straddling both urban and rural commitments, survival strategies and cultural expressions, the Nkoya were never effectively proletarianized, and the viability of their rural culture and their language testifies to this. But why have the Kalanga responded so very differently? Was it because they have been much more effectively proletarianised, with no longer any viable rural society to return to, since land alienation, over-grazing and drought have led to the collapse of the local ecosystem? Or because, while they became proletarianised and because of this condition, their success in the modern world away from the lost village has been so incomparably greater than that of the Nkoya?

Our argument has perhaps illuminated the role of language in the trajectories of two African ethnic identities, but does it also contain a lesson for language policy?

One striking point which emerges is that under similar conditions of withheld state recognition, the two languages, Nkoya and Kalanga, and the ethnicities associated with them, could traverse such different paths while the formal language policy in the respective countries formally remained unchanged. The existence of a restrictive Zambian language policy did not prevent the Nkoya from engaging in cultural and ethno-historical self-reconstruction, and on the Kalanga side we have seen plenty of reasons why even if the Botswana language policy would have been less restrictive, it would have been unlikely for the Kalanga to have produced a less tepid ethnic and linguistic response. The specific nature of the existing language policy hardly explains what happened — instead, political and economic factors cast much more light on the correspondences and differences of the two cases.

In principle this means that the data presented in this paper do not in themselves suggest a particular ideal form for a national language policy in African states. The following remarks therefore, although inspired by my research, are basically personal.

Much as I love the two languages discussed here, and would regret to see them disappear from the treasure of universal human culture, I do not think that the only, or even best, way to safeguard their future existence and to utilize their present-day potential for self-expression, communication and citizen participation, is to include them in a national-level formal language policy. They must be acknowledged and accommodated in policy, but only at the regional and local level.

(a) a peasant farmer being forced to use any other language than his own in first-line administrative, medical, judicial and developmental contact with the state (which, however common in both Zambia and Botswana today, constitutes an infringement of his human dignity and human rights, and effectively prevents him from citizen participation), and

(b) national-level institutions (such as parliament, the university, the High Court) being burdened with a multiplicity of languages, which may boost ethnic pride but at enormous financial sacrifice and at the risk of international isolation.

The latter even suggests that the national political space within which ethnic and linguistic processes evolve (since this is the constitutionally defined space for the legitimate exercise of state power), might very well be too narrow for a meaningful language policy to be defined. Botswana, with only one million inhabitants, has risked to isolate itself from international production and circulation in the intellectual, artistic, and technological fields, by allowing Tswana to be used at the national level beside English; Zambia’s national-level policy, of not allowing any African language to be used at the national level, seemed the better choice, but we have seen how the decline of the state is forcing it to compromise in this respect.

When the costs of thwarted citizen participation and frustrated ethnic pride at the local and regional level are weighed against the costs of consistent plurality of official languages in politics and public administration, formal education, industry etc., we have to look for a formula which balances efficiency with equity: a graded model which insists on the use of one official language at the national level, while for the regional and local level making generous provision for the use, preservation and propagation of such plurality of languages as actually exists within the national boundaries. In this way basic human rights are safeguarded; the obvious role of the mother tongue in alfabetisation is recognized (even reasons of efficiency — no modern state can afford to waste the intellectual and technological potential of the youths who happen to have a minority language as their mother tongue); the requirements for effective local and regional communication in social, cultural, religious and political matters — surely essential in a democratic state — are met; and language-centred ethnic frustration is far less likely to threaten the stability and integrity of the state.

In post-colonial Africa, and in the world at large today, there are numerous examples to show that such threat a can be very real. A Kalanga activist like Kwele liked to see the Kalanga case in the same light, and his standard ethnic rhetoric included the phrase ‘or else the guns will speak.’ In the light of my analysis this prediction is, for the moment, somewhat unrealistic, but the experience of humiliation on which such utterances are based, is both undeniable and unnecessary, and deserves our concern and our intellectual efforts.

The argument also has some bearing on the issue of development. In the first place symbolic reconstruction and social revival at the local and regional level — such as language-based ethnicisation often entails — could be recognized, more than is the case at present, as an essential element of development. The practical design and implementation of a graded unitary/pluralist language policy as advocated above would qualify for an exciting form of development cooperation.

More often however, development is conceived in terms of a population’s increased economic opportunities. In this respect the lessons of my analysis are far from straightforward. Among the Nkoya, the post-colonial state’s positive if partial and opportunist response to ethnic aspirations (a package including language, traditional leaders, modern representational bodies at the regional and national level) brought the Nkoya much closer, not only to the state, but also to a modernizing economy which is clearly developmental — even if most Nkoya villagers did not exactly benefit from this type of development. Among the Botswana Kalanga, their relative access to a modernizing economy and to the state (albeit not specifically as Kalanga) prevented their ethnic and language aspirations to take wing. In both cases one has the feeling that the state’s formal language policy has remained rather irrelevant since it failed to make provision for the local level of peasants and poor urbanites; it is primarily at this local level (where plurality of languages is a reality, especially in Africa) that development is realized, or fails to be realized.

Alexandre, P., 1972, An introduction to languages and language in Africa, Heinemann: London, Ibadan, Nairobi.

Amselle, J.-L., 1985, ‘Ethnies et espaces: Pour une anthropologie topologique’, in: Amselle, J.-L., & E. Mbokolo, ed., 1985, Au coeur de l’ethnie, Paris: La Decouverte, pp. 11-48.

Baylies, C., & M. Szeftel, 1992, ‘The fall and rise of multi-party politics in Zambia’, Review of African Political Economy, 54, July 1992, pp. 75-91.

Beach, D.N., 1980, The Shona and Zimbabwe, 900-1850: An outline of Shona history, Gwelo: Mambo Press.

Bernard, St., ‘Botswana: un multipartisme fragile et menace?’, Politique Africaine, 36, Dec. 1989: 125-128.

Bhebe, N.M.B., 1973, Missionary activity among the Ndebele and Kalanga: A survey, in: A.J. Dachs, ed., Christianity south of the Zambezi, I, Gwelo [Gweru]: Mambo Press, pp. 41-52.

Blake-Thompson, J., & Summers, R., 1956, ‘Mlimo and Mwari: Notes on a native religion in Southern Rhodesia’, Native Affairs Department Annual (NADA), 33: 53-58.

Brown, E.D., 1984, ‘Drums of life: Royal music and social life in Western Zambia’, Ph.D. thesis, University of Washington, School of Music; University Microfilms, Ann Arbor.

Chirenje, J. M., 1977, A history of Northern Botswana 1850-1910, London: Rutherford.

Clay, G.C.R., 1945, History of the Mankoya district, Livingstone: Rhodes-Livingstone Institute, Communication no. 4, ‘re-roneod’ [sic] 1955.

Crowder, M., 1988, ‘Botswana and the survival of liberal democracy in Africa’, in: P. Gifford & W.R. Louis (eds), Decolonization and African independence, 1960-1980, New Haven (Conn.): Yale University Press, pp. 461-476.

Daneel, M.L., 1970, The God of the Matopo Hills — An essay on the Mwari cult in Rhodesia, The Hague/Paris: Mouton for African Studies Centre.

Doke, C.M., 1931a, A comparative study in Shona phonetics, Part IV, Johannesburg: University of Witwatersrand Press.

Doke, C.M., 1931b, Report on the unification of the Shona dialects, Salisbury: Government of Southern Rhodesia, printed by Stephen Austin & Sons Ltd.

Doke, C.M., 1954, The Southern Bantu languages, Oxford: Oxford University Press.

Epstein, A.L., 1978, Ethos and identity: Three studies in ethnicity, London: Tavistock.

Fabian, J., 1986, Language and colonial power: The appropriation of Swahili in the former Belgian Congo, Cambridge University Press.

Fardon, R., & G. Furniss, 1991, ‘Language and languages: Frontiers and boundaries’, paper read at the conference on ‘African languages, development and the state’, Centre of African Studies (University of London), and EIDOS, London, April, 1991.

Fortune, G., 1949, Ndevo yenombe luvizho and other Lilima texts, Communications from the School of African Studies, University of Capetown, New Series, No. 21.

Fortune, G., 1956, ‘A Rozvi text with translation and notes’, NADA [Native Affairs Department Annual], 33: 67-91.

Fortune, G., 1959, A preliminary survey of the Bantu languages of the Federation, Lusaka: Rhodes-Livingstone Institute, Rhodes-Livingstone Communication No. 14.

Fortune, G., 1963, ‘A note on the languages of Barotseland’, in: Proceedings of Conference on the History of Central African Peoples, Lusaka: Rhodes-Livingstone Institute.

Fortune, G., 1969, ‘ 75 years of writing in Shona’, Zambezia, 1, 1: 55-67.

Gadibolae, M., 1985, ‘Serfdom (Bolata) in the Nata area’, Botswana Notes and Records, 17: 25-32.

Geertz, C., ed., 1963, Old societies and new states: The quest for modernity in Africa and Asia, New York: Free Press.

Gillett, S., 1973, ‘The survival of chieftainship in Botswana’, African Affairs, 72, no. 287: 179-185.

Givon, T., 1971, The Si-Luyana language, Lusaka: Institute for African Studies, Communication no. 6.

Gluckman, M., 1943, Organisation of the Barotse native authorities: With a plan for reforming them, Lusaka: Rhodes-Livingstone Institute, Communication no. 1.

Gluckman, M., 1951, ‘The Lozi of Barotseland, N.W. Rhodesia’, in: E. Colson & M. Gluckman (eds.), Seven tribes of British Central Africa, Oxford University Press, pp. 1-93.

Gluckman, M., 1968, Economy of the Central Barotse Plain, Manchester University Press, Rhodes-Livingstone Paper No. 7, reprint of the 1941 edition.

Good, K., 1992, ‘Interpreting the exceptionality of Botswana’, The Journal of Modern African Studies, 30, 1: 69-96.

Greenberg, J., 1963, Languages of Africa, Bloomington: Indiana University Press.

Guthrie, M., 1948, The classification of the Bantu languages, Oxford University Press for the International African Institute.

Holm, J., & P. Molutsi, eds., 1989, Democracy in Botswana, Gaborone: Macmillan.

Holub, E., 1879, Eine Culturskizze des Marutse-Mambunda Reiches in Sud-Central-Afrika, Vienna: Konigliche und Kaiserliche Geographische Gesellschaft.

Kashoki, M.E., 1978, ‘The language situation in Zambia’, in: S. Ohannessian & M.E. Kashoki (eds.), Language in Zambia, London: International African Institute, pp. 9-46.

McCulloch, M., 1951, The southern Lunda and related peoples, London: Oxford University Press, Ethnographic Atlas of Africa.

Malikongwa, D.M., & C.C. Ford, 1979, ‘The history of the Bakhurutshe (the Phofu group) and the Bakalanga of Botswana’, unpublished MS, Institute of Adult Education, University of Botswana.

Maripe, K.T.T., 1987, ‘Open letter to His Excellency Dr Q.K.J. Masire, President of the Republic of Botswana’, n.p.: Botswana Peoples Party.

Masale, G., 1985, ‘Ethnicity and regionalism in Botswana and their impact on elections: A case study of the North-East and Francistown’, paper presented at the workshop on preliminary results from 1984 election study project at the University of Botswana, Gaborone, 16th to 19th May 1985.

Mautle, G., 1986, ‘Bakgalagadi-Bakwena relationship: A case of slavery’, c. 1840-c. 1930, Botswana Notes and Records, 18: 19-32.

Meebelo, H.S., 1986, African proletarians and colonial capitalism, Lusaka: Kenneth Kaunda Foundation.

Mitchell, J.C., 1974, ‘Perceptions of ethnicity and ethnic behaviour: An empirical exploration’, in: Cohen, A., ed., 1974, Urban ethnicity, London: Tavistock, pp. 1-35.

Mkanganwi, K.G., 1991, ‘Language planning in Southern Africa’, paper read at the conference on ‘African languages, development and the state’, Centre of African Studies (University of London), and EIDOS, London, April, 1991.

Mulford, D.C., 1967, Zambia: The politics of Independence, 1957-1964, London: Oxford University Press.

Murray, A., H. Nengwekhulu, & J. Ramsay, 1987, ‘The formation of political parties’, in: Morton & Ramsay 1987: 172-186. Morton, F., & J. Ramsay, (eds), The birth of Botswana: A history of the Bechuanaland Protectorate from 1910 to 1966, Gaborone: Longman.

National Executive Council, Botswana Peoples Party, 1984, This way Botswana: Botswana Peoples Party 1984 Election Manifesto, Francistown: Botswana Peoples Party.

National Executive Council, Botswana Peoples Party Headquarters, n.d. (1988), To the President of the Republic of Botswana, Francistown: National Executive Council, Bortswana Peoples Party Headquarters.

Ndebo etc., 1985, Ndebo Mbuya Yobuhe gwe Ndzimu: Four Gospels and Acts in KALANA [sic], Roggebaai, Cape Town: Bible Society of South Africa, third South African impression 1985, reprint of the 1st edition by the British and Foreign Bible Society, 1929; first South African edition 1979.

Nengwekhulu, H.R., 1979, ‘Some findings on the origins of political parties in Botswana’, Pula, 1, 2: 47-76.

Njimbo etc., 1988, Njimbo dzi no kudza Ndzimu: Kalanga hymn book, n.p.[printed and bound at Goodwood, Cape] : UCCSA, second edition, reprint of the 1967 edition, 32 pp.

Ohannessian, S., & M.E. Kashoki (eds.),1978, Language in Zambia, London: International African Institute

Papstein, R.J., 1989, ‘From ethnic identity to tribalism: The Upper Zambezi region of Zambia, 1830-1981’, in: Vail 1989: 372-394.

Parkin, D., 1991, ‘Language, government and the play on purity and impurity: Arabic, Swahili and the vernaculars in Kenya’, paper read at the conference on ‘African languages, development and the state’, Centre of African Studies (University of London) and EIDOS, London, April, 1991.

Picard, L.A., 1987, The politics of development in Botswana: A model for success?, Boulder & London: Lynne Rienner Publishers.

Prins, G., 1980, The hidden hippopotamus, Cambridge: Cambridge University Press.

Ramsay, J., 1987, ‘Resistance from subordinate groups: Babirwa, BaKgatla Mmanaana and BaKalanga Nswazwi’, in: Morton & Ramsay 1987: 64-81. Morton, F., & J. Ramsay, (eds), The birth of Botswana: A history of the Bechuanaland Protectorate from 1910 to 1966, Gaborone: Longman.

Republic of Botswana, 1983 etc., Laws of Botswana: The Constitution of Botswana, n.p. [Gaborone]: s.n. [Government Printer].

Roberts, S.A., 1972, ‘The survival of the traditional Tswana courts in the national legal system of Botswana’, Journal of African Law, 16: 103-129.

Schapera, I., 1943, ‘The native land problem in the Tati district’, unpublished report to the Bechuanaland Protectorate Government (cf. Schapera 1971).

Schapera, I., 1952, The ethnic composition of Tswana tribes, London: London School of Economics and Political Science, Monographs on Social Anthropology no. 11.

Schapera, I., 1971, ‘Native land problems in the Tati district’, Botswana Notes and Records, 3, 219-268.

Schapera, I., 1984, The Tswana, London: Kegan Paul International in association with the International African Institute, enlarged reprint of the 1963 edition.

Schapera, I., 1988, Praise-poems of Tswana chiefs, Cape Town/Oxford: Oxford University Press, 3rd impression (reprint of the 1965 edition).

Silitshena, R.M.K., 1979, ‘Chiefly authority and the organisation of space in Botswana’, Botswana Notes and Records, 11: 55-67.

Stokes, E., 1966, ‘Barotseland: Survival of an African state’, in: Stokes, E., & R. Brown (eds.), The Zambesian Past, Manchester: Manchester University Press, pp. 261-301.

Tapela, H.M., 1976, ‘The Tati district of Botswana, 1866-1969’, Ph.D. thesis, University of Sussex.

Tapela, H.M., 1982, ‘Movement and settlement in the Tati region: A historical survey’, in: Hitchcock, R.R., and M.R. Smith, (eds), Settlement in Botswana, Gaborone: Botswana Society, pp. 174-188.

Testamenta, 1952, Testamenta ya yipya/Nyimbo [Nkoya New Testament and Psalms], London: British and Foreign Bible Society.

Vail, L., ed., 1989, The creation of tribalism in Southern Africa, London/Berkeley & Los Angeles: Currey/University of California Press.

van Binsbergen, W.M.J., 1977, ‘Law in the context of Nkoya society’, in: S.A. Roberts, ed., Law and the family in Africa, The Hague/Paris: Mouton, pp. 39-68.

van Binsbergen, W.M.J., 1985, ‘From tribe to ethnicity in western Zambia: The unit of study as an ideological problem’, in: W.M.J. van Binsbergen & P. Geschiere, eds., Old modes of production and capitalist encroachment: Anthropological explorations in Africa, London: Kegan Paul International, pp. 181-234.

van Binsbergen, W.M.J., 1986, ‘The postcolonial state, “state penetration” and the Nkoya experience in central western Zambia’, in: W.M.J. van Binsbergen, F. Reijntjens & G. Hesseling (eds.), State and local community in Africa, Brussels: Centre d’Etudes et de Documentation de l’Afrique (CEDAF), pp. 31-63.

van Binsbergen, W.M.J., 1987, ‘Chiefs and the state in independent Zambia’, Journal of Legal Pluralism and Unofficial Law, no. 25-26, 1987: 139-201.

van Binsbergen, W.M.J., 1988, (ed.), J. Shimunika’s Likota lya Bankoya: Nkoya version, Research report No. 31B, Leiden: African Studies Centre.

van Binsbergen, W.M.J., 1992a, Tears of Rain: Ethnicity and history in central western Zambia, London/Boston: Kegan Paul International.

van Binsbergen, W.M.J., 1992b, Kazanga: Etniciteit in Afrika tussen staat en traditie, inaugural lecture, Amsterdam: Free University.

van Binsbergen, W.M.J., in the press (a), ‘Making sense of urban space in Francistown, Botswana’, paper read at the Symposium ‘Urban Images: Cities and Symbols. Symbols and Cities’, Centre for Non-western Studies, Leiden University, 6-9 January 1992; in the press, in: P. Nas, ed., Urban images, Leiden: Brill.

van Binsbergen, W.M.J., ter perse (b), ‘African Independent churches and the state in Botswana’, paper read at the conference on ‘Power and Prayer’, Institute for the Study of Politics and religion, Free University, Amsterdam, 10-14 December, 1990; in the press, in: A. Koster, ed., Power & Prayer, based on the same conference.

van Waarden, C., 1988, The oral history of the Balakanga of Botswana, Occasional Paper no. 2., Botswana Society: Gaborone: Botswana Society.

von Sicard, H., 1954, ‘Rhodesian sidelights on Bechuanaland history’, Native Affairs Department Annual (NADA), 31: 67-94.

Weale, M.E., 1893, Matabele and Makalaka vocabulary: intended for the use of prospectors and farmers in Mashonaland, Cape Town: Murray & St Leger Printers.

Wentzel, P.J., 1983, Nau Dzaba Kalanga: A history of the Kalanga, vol i. Text and translations; vol ii. Annotations; vol. iii. The relationship between Venda and Western Shona, Pretoria: University of South Africa.

Werbner, R.P., 1970, ‘Land and chiefship in the Tati concession’, Botswana Notes and Records, 2: 6-13.

Werbner, R.P., 1971, ‘Local adaptation and the transformation of an imperial concession in northeastern Botswana’, Africa, 41, 1: 32-41.

Werbner, R.P., 1975, ‘Land, movement and status among Kalanga of Botswana’, in Fortes & Patterson 1975: 95-120. Fortes, M., & S. Patterson, (eds), 1975, Essays in African social anthropology, London: Academic Press.

Werbner, R.P., 1989, Ritual passage sacred journey: The process and organization of religious movement, Washington/Manchester: Smithsonian Institution Press/Manchester University Press.

Werbner, R.P., 1990, ‘On cultural bias and the cosmos: home universes in Southern Africa’, in: M. Jackson & I. Karp, eds., Personhood and agency: The experience of self and other in African cultures, Uppsala: Acta Universitatis Upsaliensis, pp. 135-148.

Wilmsen, E.N., 1988, ‘The political economy of minorities and its bearing on current policy’, paper read at the conference on Botswana, Edinburgh: Centre for African Studies.

Wilmsen, E.N., 1989, Land filled with flies: A political economy of the Kalahari, Chicago & London: University of Chicago Press.

Wylie, D., 1991, A little God: The twilight of patriarchy in a Southern African chiefdom, Johannesburg: University of Witwatersrand Press, reprint of the original 1990 edition.

Yasutoshi Yukawa, 1987, A classified vocabulary of the Nkoya language, Tokyo: Institute for the Study of Languages and Cultures of Asia and Africa, Tokyo University of Foreign Studies.

[1] Field-work among the Zambian Nkoya was undertaken in 1972-74, and during shorter visits in 1977, 1978, 1981, 1988, 1989. and 1992 (twice). Field-work among the Kalanga of Botswana was undertaken in 1988-89 and during shorter visits in 1990, 1991 and 1992. I am indebted to the African Studies Centre, Leiden, for the most generous encouragement and financial support; and to research participants, to assistants and government officials in both Zambia and Botswana and to members of my family, for invaluable contributions to the research. An earlier version of this argument was presented at the conference on ‘African languages, development and the state’, Centre of African Studies (University of London) and EIDOS, London, April, 1991; in this context I wish to thank the conference organizers, Richard Fardon and Graham Furniss, and the participants, for stimulating discussions; and the African Studies Centre, Leiden, for financing my participation. Rob Buijtenhuijs made useful comments on an earlier draft.

[2] Cf. Mitchell 1974, Epstein 1978, and references cited there; major recent additions to this literature in Vail 1989, including Robert Papstein’s (1989) analysis of Luvale etnicity which has considerable parallels with the Nkoya case. I have given an overview of rural ethnicity studies on Zambia in van Binsbergen 1985; 1992a deals with the interplay between twentieth-century ethnicity and the production of images of the precolonial past.

[4] On the Nkoya, cf. Brown 1984; McCulloch 1951; van Binsbergen 1977, 1985, 1986, 1987, 1992a, 1992b; Clay 1945. On the Lozi and Barotseland (Western Province, Western Zambia) in general, cf. Gluckman 1943, 1951, 1968; Prins 1980; Mutumba Mainga 1973; Stokes 1966; Caplan 1970. The only linguistic publication on the Nkoya language is Yasutoshi Yukawa 1987.

[5] This is not only the Nkoya’s self-image (cf. van Binsbergen 1992a), but also the opinion of, among others, the ethnologist McCulloch (1951: 93) and the linguist Fortune (1959: 26).

[8] Throughout the colonial period, those Nkoya royal chiefs whose chieftaincies had survived functioned as members of the Lozi aristocracy and in this capacity boasted their own courts; proceedings were preferably conducted in Lozi, but use of Nkoya was not ruled out. Shortly after Independence (1964), the central state instituted Local Courts, with state-trained judges and assessors who de jure were independent from the chief (not de facto, since they were members of the local aristocracy and appointed in consultation with the chief). At an unofficial level below the Local Court each valley would continue to have its court presided by a senior member of the chief’s council; proceedings there would mainly be in Nkoya. An interesting development in Nkoyaland in the late 1980s was the spontaneous, but state-tolerated institution of mabombola (‘palaver’) courts at chief’s palaces, administering a local customary law in the Nkoya language, but without any formal powers beyond reconciliation; cf. van Binsbergen 1977.

[9] Testamenta 1952. A list of published texts in the Nkoya language is given in van Binsbergen 1992a: 441f.

[10] Mr. J. Kalaluka, son of a Lozi father and of a Nkoya mother, sister of Mwene Timuna Kahare. On his career, cf. van Binsbergen 1992a.

[11] Not to be confused with the South African political organization of the same name, by which it was however inspired.

[12] The Kalanga have received considerable ethnographic attention (especially in the work of R.P. Werbner (e.g. 1970, 1971, 1975, 1989, 1990), but much work remains to be done on their history (cf. Malikongwa & Ford 1979; Tapela 1976, 1982; von Sicard 1954). In recent years, the Botswana Kalanga have been the subject of extensive oral-historical and especially archaeological research by van Waarden (1988), and linguistic and ethno-historical research by Wentzel (1983). Fortune was the first to describe the Kalanga language (Fortune 1949, 1956, 1969). Elements of Kalanga can also be found in the classic works of Doke (1931a, 1931b, 1954); a very early wordlist is Weale 1893. A first impression of the historical data on Kalanga as a language can be gleaned from Beach 1980: xi, 189, 243, 258-9, 265, 279 and passim. On ethnic relations in the area, cf. Masala 1985; on rural land alienation, cf. Schapera 1943, 1971.

[13] In the Botswana constitution, the section on the House of Chiefs is the only part listing ethnic groups, and (as is clear from the context) exclusively with a view of defining the composition of that House; these groups are: Ngwato, Ngwaketse, Kwena, Tawana, Kgatla, Lete, Rolong and Tlokwa (Republic of Botswana 1983).

[14] For the place of these groups among the Tswana see Schapera 1952, 1984. Khurutshe are a sub-group of the Hurutshe Tswana of neighbouring South Africa; the Rolong Tswana also originate in South Africa but now are found on both sides of the border in the south of Botswana. One particular, massive category of immigrants had no impact of its own on the language situation beyond swelling the numbers of the existing groups: labour migrants from Zambia and Malawi for whom Francistown was a major stop-over between South Africa and their homes, and many of whom settled permanently in and around Francistown. While some have retained or assumed a distinct ethnic identity (notably as ‘Rotse’, which refers to Barotseland), virtually all have adopted Kalanga or Tswana as their language.

[15] For an official reading of the episode, where its scope and violence are relegated to minimum proportions, see Tlou & Campbell 1987. A recent reading is Wylie 1991: 162-172; also cf. Ramsay 1987. The topic is also currently covered in work in progress by such prominent historians as T.O. Ranger and N.Q. Parsons.

[16] On the Mwali cult, cf. Blake-Thompson & Summers 1956; Daneel 1970; Fortune 1973; Werbner 1989.

[17] Which meanwhile has found its way to authoritative textbooks widely available in Botswana, e.g. Tlou & Campbell 1984.

[18] At present I have insufficient data on these developments in Zimbabwe. Mkangawi (1991: 1) suggests that they are substantial but does not give details:

‘In addition, it is significant to note that although governments are active in LP [language planning], individuals, groups of individuals, or non-governmental organizations can also be involved in it, by, for example, trying to unite people in the language question, particularly where a minority language is involved (as Nsala-Malaba is doing for Kalanga in Zimbabwe).’

The developments hinted at are far from recent. Already in the late 1950s Fortune (1959: 8) wrote with reference to Zimbabwe:

‘There are a number of Kalanga and Lilima-speakers who are anxious to preserve and develop their language and to have it as a medium in their junior schools.’.

[19] Registrar of Societies, Gaborone, file no. H 28/90/258; this society was registered on 7th August, 1984; also cf. van Binsbergen, in press (b).

[20] Cf. Guthrie 1948; Fortune 1959; the continuity on the ground was repeated brought out in a practical sense in the course of my anthropological and historiographical field-work, when I found that my language skills, however limited, enabled me to communicate somehow with local non-speakers of Nkoya or Kalanga.

[21] There is an interesting concrete link between the Kalanga and the Nkoya situations: the Wankie area, the northwestern extension of the Kalanga language region, in the second half of the nineteenth century was tributary to the Kololo and subsequently the Luyana, so that Holub (1879) could list Kalanga as another Barotse subject tribe.

[22] For a related discussion of ethnicity in an idiom of space, cf. Amselle 1985.

[23] This applies to the most senior Nkoya royal chiefs, Mwene Mutondo and Mwene Kahare, of Kaoma district. Outside the district, and outside Barotseland, the history of Mwene Kabulwebulwe and Mwene Momba has been rather different, but then they never suffered Lozi incorporation to the same extent as their Kaoma counterparts.

[24] A point made by Richard Fardon at conference where this paper was first presented; also cf. van Binsbergen 1987.

[25] The usual term, in Botswana English, to denote the personal nominative of Tswana-hood.

[27] On this point the Botswana situation as described has strong parallels with present-day South Africa, and reflects such dilemma’s of cultural and symbolic reconstruction as have only very late in the liberation struggle gained explicit recognition and respectability.

[28] Not being a linguist, I include this entry only in order to allow those who are, to identify the languages professionally. I am aware that the classification of African languages has evolved greatly since Gurthrie’s pioneering work (Greenberg 1963; Alexandre 1972).

[29] Fortune contests Guthrie’s classification even if it is close to Doke’s (1954):

‘There is need for a closer examination of the Nkoya-Mbwela languages, scattered as they are, and influenced by others, it is at the moment impossible to indicate the true position.’ (Fortune 1959: 27).

Ohannessian & Kashoki’s authoritative Language in Zambia regrettably does not reconsider this question and comfortably classify Nkoya in a group of its own (H, comprising such minor dialectal variants as local speakers themselves would not hesitate to include in that cluster: Nkoya, Lukolwe/Mbwela, Lushangi and Mashasha) (Ohannessian & Kashoki 1978:20 and passim).

[34] Guthrie (1948) classification K21, which however does not seem to do justice to the fact, well recognized by Fortune, that Lozi is

‘a mixture of Southern Sotho and Luyana, now the lingua franca of Barotseland Protectorate and used in all African courts. It is mainly Sotho in morphology but has a great number of Luyana words which seem to be increasing. (...) The language Lozi, a combination of Sotho and Luyana, grew up between 1869, the date of the expulsion of the Kololo invaders, and 1919, at the Barotse court.’ (Fortune 1959: 41-42; cf. 1963).

On Luyana, the original court language of the Luyana kingdom with striking parallels with contemporary Nkoya, cf. Givon 1971.

[35] This was the year of official registration; considering the amount of time needed to prepare for official registration, the actual founding initiative dated from 1981-82.

variable	Nkoya	Kalanga
1. basic language data
Guthrie (1948) classification	L62	T16
estimated number of speakers in national territory	30,000	120,000
percentage of national population	0.80%	13%
formalized dialectical variation	hardly	yes
orthography	not an issue, rather standardized	bone of contention between Botswana and Zimbabwean native speakers
close affinity with languages outside country where this language is spoken	Lunda, Luvale (Zambia); (Mbwela, Ganguela in Angola)	other Kalanga dialects in western and southern Zimbabwe (the Botswana Kalanga is identified as the Lilima dialect; the Kalanga cluster is considered to be western Shona)
2. inter-language relations
national official language	English	English/(Tswana)
state-recognized language	Bemba, Nyanja, Lozi, LuvaleTonga, Kaonde, Lunda	Tswana
regionally dominant language	Lozi	Tswana
language most speakers in local rural area speak	Nkoya	Kalanga
language of official communication in local rural area	Lozi	Tswana
other languages in local rural areas	Lozi, Luvale, Luchazi, Kaonde, Tonga, Ila, Lunda, Mbundu	Tswana, Ndebele, English, Afrikaans, Shona, San
lingua franca of national urban centres frequented by native speakers	Lozi (Livingstone), Nyanja (Lusaka today), Bemba (Copperbelt, pre-WW II Lusaka)	Tswana/(Kalanga) (Francistown), Tswana (other Botswana towns)
3 ethnic self-perception
language name in use as ethnic name since	mid-19th century	17th century
is ethnic identity underpinned by historical consciousness?	yes, spanning the period 16th-20th century	yes, but mainly with reference to colonial period
contemporary self-perception of ethnic group in terms of local precedence	claims historical precedence as the first group to arrive in central western Zambia from Zaire	dim historical awareness, many members accept the erroneous Tswana claims as to the Bo tswana Kalanga being recent immigrants from Zimbabwe
4. inter-ethnic relations
nationally dominant ethnic group	Bemba, Tonga, ‘Nyanja’, Lozi	Tswana
regionally dominant ethnic group	Lozi	Tswana
the language’s status in national society	very low	low
main perceived ethnic enemy	Lozi	Tswana
when native speakers ‘pass’, they pass as	Lozi, Bemba	Tswana
economic position of average native speaker/member of this ethnic group	below national average	at or above national average
5. language and state institutions except formal education
is it allowed to use this language in the lowest courts?	allowed but discouraged in testimony; judges prefer Lozi, clerks record in English	formally yes but discouraged
can state officials use this language in dealing with the local population?	not until late 1980s	no
is this language admissible for use in Parliament?	no	no
state attitude towards this language	reticent but increasingly positive	strongly negative and hostile
is this language used in broadcasting?	no, except in popular and traditional songs	no
6. language and formal education
language used in schools	Lozi (Nkoya only occasionally and informally	Tswana (Kalanga only occasionally and informally)
school primers locally available in print	not any more	no
relative educational level of native speakers of this language	far below national average	well above average
7. politics and language
was this language ever the state language of a pre-colonial state?	yes, since late 18th century	yes, from 17th to early 19th century
is this language used in royal/courtly ceremonies today?	only by musicians but throughout western Zambia (since early 19th c.)	no, except at village level
is this language used in local-level traditional politics today?	yes, by the district’s major traditional ‘rulers’	yes, but not the district’s major one
do representatives of this ethnic group have access to national-level traditional politics today?	yes	no
do representatives of this ethnic group have access to modern regional politics today?	limited but increasing	extensive
do representatives of this ethnic group have access to modern national-level politics today?	limited	extensive
political party favouring this language	UNIP, MMD (ruling parties)	BPP, BPU (opposition parties)
is language an issue in oppositional political rallying?	it was in the 1970s, not any more	yes!
8. language and expressive culture
literary works available in this language	none exist	some exist but none are available locally
pious literature available in this language	rather abundantly	no evidence except hymn book and part of NewTestament
ethno-historical literature available in this language	yes	yes but not available locally
is this language served by an ethnic/linguistic society? (since)	yes, Kazanga (1983)	yes, SPIL (1984 )
profile of that society’s membership	primary school teachers, local-level politicians, paramedics	university students, secondary school teachers, highly-educated professionals
is there a periodical in this language?	no	yes but moribund (Tjedza)
does linguistic self-presentation take place in conjunction with other cultural forms (music, dance)?	yes, emphatically so	yes, but problematically so
9. religion and language
was this language first committed to writing by a Christian mission?	yes	yes
Bible translation (year)	New Testament and Psalms (1953); Old Testament recently completed in MS	Gospels and Acts (1929); new translation currently in preparation
language used in local churches	Nkoya, Lozi	Kalanga, Ndebele, Tswana
(main) hymn book in	Mbunda/Nkoya/Lozi	Tswana, Sotho; Kalanga
is there any church exclusively using this language?	no	no
is this language used in non-Christian cults?	yes, but together with Luvale and Lenje	yes, but together with Ndebele, Shona, Venda,

page last modified: 20-04-13 12:24:28

© 1994-2002 Wim van Binsbergen