The Malila: Preliminary Notes On Language, History And Ethnography

THE MALILA

Preliminary notes on language, history and ethnography Martin Walsh #1, April 1998

Introduction The Malila are a relatively unknown ethnic group in south-west Tanzania. Very little has been written about the language, history and ethnography of the Malila and they appear never to have been the subject of research except in the context of studies of neighbouring and related peoples. The following notes comprise a first attempt at compiling information on the Malila and their language from existing sources, both published and unpublished. This first draft is necessarily incomplete, being based upon a review of sources and notes in the author’s possession. Location The traditional territory of the Malila is in the highland area of the same name (Malila or Umalila), located to the south-west of Mbeya, roughly half-way between the regional capital and the Malawian border. The highlands rise to a peak of 2,456 metres above sea level at Mbogo, and, according to Cribb and Leedal, they are primarily drained by the northern Songwe River into Lake Rukwa (1982: 3-4). The traditional boundaries of Umalila are not shown clearly on available maps. Historically the Malila appear to have shared borders with the Songwe (Safwa) and possibly Poroto to their north and north-east, the Mbozi Nyiha to their west, the northern Lambya and Ndali to their south and south-east, and Nyakyusa speakers to their east (including the Nyika and Penja of Rungwe district?). According to notes in Mbeya District Book, the territories of the following Nyiha chiefs once bordered Umalila: Mwezimpya (in the north), Mwamlima, Mwamengo, Mugaya, and Mwembe (in the south) (‘Tribal history and legends: WaNyiha tribe’). The Songwe (lowland Safwa) chief whose land bordered Umalila appears to have been Mirambo (‘Sketch Map of Usongwe’, also in Mbeya District Book). It seems that most Malila now live in the south of Isangati division in Mbeya district, with some in the northern part of Ulambya division in Ileje (formerly part of Rungwe) district. Mariam Slater’s map shows two principal areas of Malila settlement, one focusing on the village of Isangati itself and bordering Safwa territory, and the other around Ilembo, close to Lambya country (1976: ix). Both of these places are in Isangati division. Van Hekken and Thoden van Velzen, however, worked in the Malila village of Ibala, which lies on the southern fringe of Umalila in Ulambya (Bulambia) division (1972: 21).

1

Population The German administration estimated that there were about 1,700 Malila, 567 of whom were ‘men fit for bearing arms’ (Admiralty 1920: 66, describing conditions c.1910-11). Judging by later British census returns, this was probably a gross underestimate. The Native Census of 1921 reported a population of 7,996 Malila in Rungwe district (this was before the creation of Mbeya district). The 1931 census, however, indicated a total of only 5,878 Malila throughout the Southern Highlands province. This apparent drop in numbers is perhaps no more than a reflection of the comparative inaccuracy of the earlier census, though it is also possible that shifting perceptions of ethnic identity (and in particular the desire to be identified with more salient neighbouring groups) influenced the census returns. In the 1957 population census of Tanganyika, 20,745 people identified themselves as Malila (Polomé 1980: 4). This is the most up-to-date figure available for the Malila: subsequent Tanzanian censuses have not incorporated data on ethnic affiliation. Ethnographic classification In her survey of The Peoples of the Nyasa-Tanganyika Corridor (1958), Monica Wilson classed the Malila and closely related peoples under the heading ‘Nyiha’: “The Nyiha consist of a number of scattered groups living mostly on the drier parts of the table-land between the Lakes. They include (i) the Nyiha around Mbozi in Mbeya district; (ii) the Lambya who adjoin them in Rungwe district; (iii) the Lambya to the south of the Songwe in Nyasaland [now Malawi]; (iv) the Wandya who adjoin them (to be distinguished from the Wanda at the south of Lake Rukwa); (v) the Lambya of Northern Rhodesia [now Zambia]; (vi) the Nyiha of Nothern Rhodesia [Zambia]; (vii) the people of the wet Malila plateau adjoining the Lambya of Rungwe district; (viii) scattered groups on the Fipa plateau; and (ix) the Nyiha of Rungwe district.” (1958: 28)

Wilson, famous for her earlier work (together with Godfrey Wilson) among the Nyakyusa, walked through Malila in 1954, but has very little to say about the place and its people: “The Malila plateau adjoins both Nyakyusa and Lambya country, and people living there said that in language and custom they resembled the Lambya, as they did in their style of dress and building. [The explorer Joseph] Thomson refers to Nyiha villages on the Malila plateau in 1879 so the cultural similarity is nothing new.” (1958: 28) “Lambya and Malila are names of localities which have come to be applied to the people who occupy them.” (1958: 29) “Traditionally the Nyiha wore skin clothing and elaborate bead ornaments, and this dress was still visible in 1955 [sic] in remote villages on the Malila plateau.” (1958: 32)

In concluding her survey, Wilson listed ‘A study of the Nyiha ‘people’ as one of a number of ‘obvious topics’ for ethnographic research: “They are the only large group in the area on whom there is no professional study either published or forthcoming. They appear to have been very early inhabitants of the Corridor, and they have been so conservative that detailed field work might yet provide a great deal on

2

traditional history and culture. Are the various ‘dialect groups’ here classed as Nyiha indeed similar in language and custom?” (1958: 61-62)

When Wilson wrote these words she appeared to be unaware of the collection of 14 northern Lambya texts published by Joseph Busse (1939/40) and the same author’s forthcoming monograph on the language of the Nyiha of Mbozi district (1960). Nonetheless, her survey did stimulate a significant body of subsequent ethnographic and historical research among the Nyiha of Mbozi (the published results of which include Brock 1963, 1966, 1968; Knight 1970, 1974; Slater 1976; and Gartrell 1979). This research, together with anthropological work on the Fipa, has also added (though not much) to our knowledge of the Nyika who live to the west of Lake Rukwa (Willis 1966). The precolonial history of the Lambya in Malawi has been studied more recently as part of a wider investigation of the history of the Ngonde kingdom (Kalinga 1974; 1978). The Malila and most of the other groups mentioned by Wilson, however, remain largely unstudied, and, although some of their linguistic relationships are now clearer, we are still in no position to provide a definitive answer to Wilson’s closing question.

Language The language of the Malila has never been subject to detailed investigation, though sufficient information (mainly lexical) has been collected to provide an indication of its position in a genetic classification of the better-known Bantu languages of southwest Tanzania. Linguistic data Unfortunately, most of the linguistic data on Malila which has been used to classify the language remain unpublished. The only published lexical material is a list of 22 plant names recorded by the botanists Cribb and Leedal (1982: passim.). Their list, together with botanical identifications, is reproduced in a later section of this paper. Given the manner of its collection, it is unlikely that these terms are recorded with any phonological accuracy. Using his own lexical material, Derek Nurse has recorded the reflexes of proto-Bantu consonants in Malila (‘Malela’) and three other ‘Nyika’ languages: Mbozi Nyiha, northern Lambya, and Tambo (1988: 104). In this table he shows Malila as a seven vowel language with no contrast between long and short vowels. However, it should be noted that not all of the entries for other languages in this table (which is incomplete) are confirmed by other sources, and it should therefore be treated as no more than a provisional guide to Malila phonology. Linguistic classification Malila was not mentioned in Harry Johnston’s A Comparative Study of the Bantu and Semi-Bantu Languages (1922), although he had earlier travelled across the NyasaTanganyika Corridor and collected data on some of its languages. Johnston classified Malila’s closest neighbours in his Group M, ‘The North-West Nyasa Languages’, a grouping which includes a selection of languages from all three Corridor sub-groups (as currently defined) as well as Tumbuka and related languages in what is now Malawi (1922: 59-63). Guthrie (cited in Polomé 1980: 17) also classified Malila

3

(M24) in his own Group M, the composition and sub-grouping of which has remained largely unchanged in subsequent classifications. Guthrie’s Group M corresponds to Heine’s ‘Fipa-Konde’ group in the latter’s genetic classification of the ‘Eastern Highland’ Bantu languages, based on 100-word lists (cited in Polomé 1980: 18). In Derek Nurse’s more recent classification of the languages of south-west Tanzania, this group is renamed ‘Corridor’, and Malila (or ‘Malela’) is placed in a sub-group of the Corridor languages labelled ‘Nyika’, together with Nyiha, Lambya and Safwa (1988: 20). Nurse’s classification is based primarily upon 1,000-word lists (collected in Dar es Salaam in the 1970s with Gerard Philippson), from which 400-word lists were drawn for lexicostatistical comparison (1988: 18). The lexicostatistical analysis presented by Nurse shows that Safwa is the most divergent member of the Nyika sub-group, with Lambya splitting off from Malila and Nyiha only slightly before they separated from one another (1988: 91). While the linguistic (and cultural) divergence of Safwa from other members of the sub-group is apparent to most observers (Mwakipesile’s claim that the Malila are a Safwa sub-group being an exception, n.d.: 35), the precise nature of the relationship between the other members is less clear. Consider the differing perspectives offered by Busse and Knight: “We can say the following about the relationship between Nyiha and neighbouring languages. Lambya stands particularly close and is spoken as a dialect of Nyiha. Whether this has come about as a result of long geographical proximity, or whether it has always been the case, cannot be established. Nyiha is also closely related to Safwa, Kinga, Ndali, Malila and Poroto. In the lexicon, phonology and grammar the connection is apparent.” (Busse 1960: 84) “The Malila are related to the Nyiha, their languages being virtually identical. Malila claim to be able to converse much more easily with Nyiha than Safwa or Songwe [the western or lowland Safwa].” (Knight 1974: 27)

Given the imprecision of lexicostatistical analysis at this level, especially when based upon a limited sample of dialects, it would be prudent to say that the available data are insufficient to clarify these relationships further at present. It may well be that Malila and its close relatives form a dialect continuum: whether or not they do, and what direction linguistic relationships take, will only be revealed through further research. Linguistic classification has, at least, clarified some of the wider connections of Malila. It is evident that the Safwa and Nyiha ‘peoples’ belong, both linguistically and culturally, to the same sub-group (Nurse’s ‘Nyika’), as Monica Wilson herself suspected (1958: 41). It is also apparent that the ‘Nyika’ languages form a higherlevel grouping together with their western neighbours, members of the ‘Mwika’ subgroup (which includes Nyamwanga, Mambwe, Lungu and Fipa). The genetic relation between the Nyika-Mwika languages and the Nyakyusa-Ndali languages (in Nurse’s terminology) is less certain, though there are reasonable grounds for classifying them together in a wider Corridor group, as Guthrie and others have also done. There is, however, a sharp dividing line between the Corridor languages and the Southern Highlands languages to the east, a linguistic boundary which cuts across Monica Wilson’s original conception of the Nyasa-Tanganyika Corridor as a much broader cultural area extending to the east and north-east of the northern end of Lake Nyasa.

4

Despite continuing uncertainties (for example over the immediate external relationships of Corridor and Southern Highlands), this genetic linguistic classification has obvious historical implications. It puts paid to Kalinga’s notion of a historically related group of peoples (the so-called ‘Ngulube peoples’) united by common ritual traditions, unless the distribution of these traditions is understood in terms of later cultural diffusion. Unfortunately, anthropologists and conventional historians have been slow to catch on. The most recent anthropological / historical overview of the cultures of south-west Tanzania almost completely ignores the implications of Nurse’s linguistic analysis, despite the fact that it was written to complement Nurse’s paper (Park 1988; Nurse and Park 1988).

History Nothing has been written specifically about the history of the Malila, though there are scattered references to them and their rulers in the published and unpublished literature. Geography and the available linguistic evidence suggest that the Malila may only have emerged as a distinct ethnic group in the relatively recent past. It is possible that they developed a separate identity (separate from that of their other Nyika neighbours) largely as a result of their comparative isolation in the highlands of Malila, and that this isolation fostered a degree of unity which persisted through into the colonial period. Apart from geographical location (recalling Monica Wilson’s observation that Malila was originally the name of a place rather than a people), the focus of Malila ethnicity in at least the recent past appears to have been a single chiefship, all of the known rulers sharing the same family name, Mwaluvanda (which is variously transcribed in the literature). It is impossible to deduce from existing sources, however, when and how this particular lineage came to prominence. It may originally have been only one among a number of localised polities, in which the ritual components of leadership were as important as (or more important than) the political (as Harwood emphasises in his study of the Safwa, 1970). What, if any, impact the institution of Indirect Rule by the British may have had in this context is also obscure. Origin traditions The oral traditions of at least two traditional Nyiha polities, the Mwamlima and Nzowa chiefships, claim a common origin with the Malila ruling line. This historical connection was first alluded to in writing in Mbeya District Book: “Mwalima’s ancestors originally came from Uwanje and may be that departure there from coincided with that of Nzowa’s forefathers. It is interesting to note that Mwalima is related to Mwaruanda, the present chief of Umalila.” (‘Tribal history and legends: WaNyiha tribe’, written c.1926)

A rather different and more detailed version of this tradition, which makes explicit reference to the common origin of the Mwamlima and Mwaluvanda (‘Mwaruanda’) lines, together with other chiefs in the region, was given to the anthropologist Beverley Brock in 1961:

5

“The elderly Mwamlima chief, Msawira, traces his ancestry back to Gogo country, with a suggestion of Somaliland origin (Ubarawa) before that. He says one of the party stayed in Gogo country, another dropped out to found the Merere line of chiefs among the Sangu, a third founded the Zumba line, chiefs of the Guruka (or Galuha) division of the Safwa, and a fourth became the Malila chief, Mwaruanda. When Mwamlima’s ancestor reached the hills around Iyula, he found people already in occupation. These people are often said to have been the ancestors of the present Mgala clan, although Mwamlima will not concede this.” (Brock 1963: 126)

A similar version was given to John Gay in 1981 by Juma Mwamlima, the son and heir to the title of Brock’s informant: “His distant ancestors came from Ugogo, and his grandfather told him that the most remote ancestors had come from Balawi in Somalia. They didn’t stop in Kenya but came straight on to Ugogo in Tanzania… The family stopped in Njombe, along with the ancestors of Zomba, Melele and Malwanda.” (from an interview dated 29 August 1981, unpublished field notes on the Nyiha)

It seems reasonable to assume that the latter two versions both derived from the same source: Shimamula Mwamlima, who ruled in the immediate pre-colonial period and was the father of Msavila (Brock’s ‘Msawira’) and grandfather of Juma. Both Brock (1963: 256) and Gay were given similar chiefly genealogies, of nine and eight generations depth respectively, from Juma Mwamlima (who acted as chief from 1948 (or 1949) until Independence) to an ancestor called Sisole. The genealogy given in Mbeya District Book only counts six generations of chiefs, from Mwamlima (probably Msavila, who was already ruling in 1914) back to an ancestor called Mwakombe, not Sisole. Brock discusses this tradition as a ‘mythical charter’. A few years before her fieldwork some members of the Mgala clan had laid claim to the chiefdom on the grounds that they were the original inhabitants of the land. At the resulting hearing before a district officer they had produced a copy of the second volume of KootzKretschmer’s Die Safwa, highlighting the following passage: “But Mwamlima left this land (Mwaruanda’s) and went through the forest Irindi to Unyiha and came to Vugara and took the land away from the Vagara (Wagala or Mgala) and settled there.” (1929a: 243; Brock’s translation)

Their claim was dismissed, but feelings still ran high in 1961 and it was evident that different versions of the origin tradition given to Brock reflected competing positions – each side claiming to have been the source of all the local accoutrements of civilisation, including fire and iron tools (Brock 1963: 126-127). As Brock points out, such traditions do not provide easy material for the determination of objective history, and this is borne out by the analysis of other aspects of them, in particular the cross references to shared origins with other chiefs. The recorded Mwamlima traditions do not make any explicit link with the traditions of other lines of Nyiha chiefs, though there are clear parallels with the origin myth of the Nzowa chiefship, as suggested by the anonymous British contributor to Mbeya District Book, quoted above. The version of the Nzowa’s history which he recorded has the founder of the line (called Nyambo) also coming from Uwanji:

6

“Nyambo originally came from Uwanji in the hills above Chosi. Upon leaving there he passed North through Usangu and proceeded first to Ukimbu, before finally arriving in Unyiha, where he was permitted to settle down in the neighbourhood of Shiwinga Hill by the people already living in those parts.” (‘Tribal history and legends: WaNyiha tribe’, written c.1926)

Beverley Brock was given contradictory versions of the ultimate origin of the first chief of the Nzowa line (called Shamuntu); one suggestion being that he came from Somaliland, another that he was a ‘Mang’ati’ (the common name of the Barabaig and other Southern Nilotic peoples in northern Tanzania). All were agreed, however, that he and his party travelled via Ugogo and Ukimbu, before settling among and becoming chief over the fire-less, hunter-gathering, indigenous inhabitants of the Mbozi plain, members of the Shupa clan (1963: 125). Brock discusses the parallels between this story and the versions of the Mwamlima tradition which she collected in some detail (1963: 131-132). Mariam Slater also heard different versions of the Nzowa tradition: Gilbert Nzowa, the last chief of the line, gave her an account which made the Mwamlima link explicit: “Although a former literate chief had written down the history of the Nzowas, there was no official arbiter in such matters. Recitations differed with every telling. Gilbert started with a founding ancestor from Kenya, perhaps a Gogo, he said… ‘He was a warrior and a hunter. He traveled to Tanganyika and married a Kamba… He fought the Nyika and also the Nyamwanga and Nyakyusa. After conquering, he became chief of the Igamba Nyika… The Mwamlima chief…came from the same place…’” (1976: 50-51)

Although the recorded Mwamlima traditions do not make explicit reference to Nzowa, they do claim a common origin with the chiefs of Umalila (Mwaluvanda), Uguruka (Zumba), and Usangu (Merere). The link with Mwaluvanda is also affirmed in the tradition recorded by Kootz-Kretschmer and cited above. Unfortunately we have no information on the parallel Malila traditions, assuming that they exist. In the case of Zumba, we merely have the statement, recorded in Mbeya District Book, that “His ancestors are reported to have come originally from Usangu.” This is not improbable, given that Uguruka borders Usangu, though it does add weight to a suggested link with the chiefs of Usangu. The detailed traditions of the Sangu themselves are, however, perhaps the most interesting in this context, not least because they provide an indication of the possible source of some of the claims in both the Mwamlima and Nzowa origin myths among the Nyiha. Sangu trace the descent of their royal family from a powerful stranger called Mbalawe. According to the detailed version of this tradition in Mbeya District Book, the stranger was a tall and fair-complexioned medicine-man who declared that he came from Barawa (Brava) on the Somali coast. He arrived in Usangu via Uwanji, impregnated the daughter of a local ruler, and left after showing her the medicine of chiefship and war. With the help of this medicine (and a subsequent gift of iron spears), their son, Njali, struggled to conquer all of the other local rulers in Usangu, a job eventually completed by his own son and successor, Mwahavanga (from ‘A History of Usangu. Related by the Wazee of Utengule’, dated 6 January 1930).

7

As in the Nyiha case(s), there are many different versions of this story, though the basic outline remains the same. Mwahavanga is the first Sangu chief about whom anything certain is known: he appears to have died at a ripe old age in the late 1850s. A bitter succession dispute followed, from which his maternal grandson, Tovelamahamba Merere, emerged victorious. Merere, the most notorious of Sangu chiefs, died in 1893, having established a dynasty which has held on to the Sangu royal stool through to the present. Pro-Merere versions of the Sangu origin tradition tend to gloss over the circumstances of Merere’s accession (the Mbeya District Book history is no exception), and members of his family continue to claim descent (often direct patrilineal descent) from Mbalawe (Walsh 1984: 37-40, 106-124). The Sangu traditions are silent on the claims of other chiefs to a common origin with their own. Indeed, there are good grounds for suspecting that the Mwamlima and Nzowa chiefs (and probably the Zumba as well) borrowed the relevant elements of their own traditions from the Sangu. The Sangu were the dominant political force in Usafwa and Unyiha in the latter part of the nineteenth century, especially after they had established a capital-in-exile in Safwa country (in Mwalyego’s chiefdom) in the early 1880s. They appear to have begun modifying their own royal myth to reflect the growing importance of trade with the East African coast much earlier, during Mwahavanga’s reign. The name Mbalawe was almost certainly reinterpreted to mean ‘person from Barawa’ under the influence of coastal traders and advisers at the Sangu court; while the frequently-made claim that he was an Arab no doubt also derives from the same period. It is most likely that the two Nyiha chiefs then tacked the same traditions onto their own. Nzowa in particular seems to have had a good reason to do so, having allied with Merere at a time when other Nyiha chiefs (including Mwamlima) were still resisting Sangu domination. Ultimately all of the local chiefs were forced to capitulate, including, we can assume, the ruler(s) of Umalila. The persistence of certain ‘praise-names’ may give some indication as to which elements have been recently borrowed and which not in these different traditions, though there is no guarantee that these praise-names have not been ‘invented’ themselves at some point in the past. Members of the Sangu royal family (both the Mereres and the usurped Mwahavangas) still use the praise-name ‘Mbwanji’, and the claim to a Wanji ancestor almost certainly predates the tradition of a coastal origin. According to Brock, the Mwamlima chiefs use the praise-names ‘Mugogo’ and ‘Mwanjesingogo’ (1963: 256). This suggests that the Gogo connection alluded to in both the Mwamlima and Nzowa traditions may have some historical validity: it almost certainly does not derive from the Sangu sources. The first part of the name ‘Mwanjesingogo’ looks like a Nyiha equivalent of the Sangu ‘Mbwanji’, lending some credence to the local claims to a Wanji origin. This element may equally have been borrowed from Sangu traditions. Given the present state of our knowledge, it is difficult to choose between these and other alternatives. It is also possible, for example, that the Mwamlima and Nzowa families have borrowed traditions from one another. In the absence of corresponding Malila traditions, this does not tell us much about their own chiefly origins, mythical or otherwise. It does, however, point to some of the contextual factors that may have to be taken into account when they are available for analysis. By the same token, it is quite likely that they will enrich our understanding of the origin traditions of their neighbours.

8

The immediate precolonial period Our main sources of information on the Malila in the immediate precolonial period are the Safwa narratives recorded by the missionary Elise Kootz-Kretschmer before the outbreak of the first world war (1929a: 164-337). Although there are no more than scattered references to the Malila and Umalila in these narratives, they do provide some indication of the place of the Malila in the history of the wider region and the impact of regional events upon them. The most traumatic events for all of the Nyika peoples in the nineteenth century were the raids of the Ngoni, which began around 1840, and the raids of the Sangu, which escalated in intensity after 1877 when the Sangu were exiled from their homeland by the Hehe, took up residence in Usafwa, and sought to make most of the Nyika their subjects. Before the coming of the Sangu, the Bemba and Wungu (in some cases acting on behalf of coastal slave-traders) also raided in the area, albeit less frequently and more locally. Conflict between local leaders added to this grim pattern, as a result of which the Nyika suffered from recurrent famines, considerable loss of life, and frequent displacements of population. The Malila were not spared these events. Movements of population, often of small family groups, evidently predated the coming of the Ngoni, and a number of Safwa and Nyiha lineages claim origins in Umalila (Kootz-Kretschmer 1929a: 255-256, 287, 331; Brock 1966: 1). The Mbwila, a small group of people settled between the Poroto and the Sangu, also claim to have come from Umalila according to one version of their origin recorded in Mbeya District Book: “Mahinya [the name of a Mbwila chief] people came from Malila under chief Mwanjewa their old country which they state was near Mwaruwanda’s.”

Another, more detailed version, however, places their origin further west, in Unyiha. In any event, the British administration eventually recognised that they were sufficiently distinct from the Poroto and other Safwa to warrant recognition (in 1931) as a separate ‘Tribal Unit’ (‘WaMbwila tribe’, Mbeya District Book). Other groups of people sought refuge from local conflicts in Umalila itself. One of Kootz-Kretschmer’s informants described how his great-grandfather, Ntengwi (or Untengwi), who was the ruler of Izumbwe, “fled to the land of chief Mwarwanda, in Marira” after losing a two-day battle with the Safwa chief Nsweve (whose own father had migrated to Usafwa from Ungonde). The Malila chief Mwaluvanda gave Ntengwi and his people land on which to settle and over which he could retain authority. When Ntengwi died he was succeeded as local ruler by his son Mwanyima. However, for reasons which are not made clear, Mwanyima migrated east to Unyiha, ejected the local people (the ‘Vagara’ of ‘Vugara’) from their land, and settled there. His brother Unsera, the informant’s paternal grandfather, remained behind in Umalila (Kootz-Kretschmer 1929a: 242-243). Brock identifies the ‘Mwanyima’ who settled in Unyiha with the Nyiha chief Mwamlima, and members of the Mgala clan who later disputed the Mwamlima chiefship appear to have done likewise (see above). The name Ntengwi, however, does not appear in recorded Mwamlima genealogies, including that given to Brock herself. This makes the two stories difficult to

9

reconcile, unless Kootz-Kretschmer’s Ntengwi is identified with the ‘Ntunje’ who appears in John Gay’s list of former Mwamlima chiefs. The fact that the ‘family name’ of Kootz-Kretschmer’s informant was ‘Mugogo’ – also a Mwamlima praisename according to Brock – may be significant, but further elucidation is clearly required. These events took place before the coming of the Ngoni. The general outlines of Ngoni history are well known (for one version see Ebner 1987), although the precise details of their movements in the Nyasa-Tanganyika corridor in the 1840s and 1850s remain obscure. The fission of the Ngoni into different groups under different leaders, and their long history of raiding in this area, make it difficult to reconstruct chronology and events with any precision, though there is considerable scope for further research using the multitude of existing sources. The Ngoni repeatedly raided the Nyika peoples with devastating effect (Kootz-Kretschmer 1929a: passim.; Brock 1966: 3). On one occasion Safwa are recorded as fleeing to Umalila (KootzKretschmer 1929a: 191). On at least one other occasion, the Ngoni attacked in Umalila itself. The narrative of Mpori Nkwitixa Sishivozya, a woman who was born in Igwirizya village in Ufyomi (a part of Umalila mentioned as the original home of a number of Kootz-Kretschmer’s informants), describes how the Ngoni came and destroyed their crops and stores of grain, the resultant famine, and the subsequent flight of her family to Nzowa’s in Unyiha (1929a: 331-332). Unfortunately, it is not possible to date this particular raid, nor gain a clearer impression of how often the Ngoni raided Malila settlements in particular. The Sangu began raiding the Nyika peoples, in particular the Safwa, some time before the first appearance of the Ngoni. The Sangu chief Mwahavanga is said to have killed the Safwa chief Nsweve (see above), forcing his son Nswira to flee to Ungonde. The latter then returned to chase the Sangu away in turn (Kootz-Kretschmer 1929a: 196). This was before the coming of the Ngoni, who later forced Mwahavanga himself to flee eastwards from Usangu (some years before his death in Usangu at the end of the 1850s). Sangu attacks intensified, however, during the reign of Mwahavanga’s successor, Tovelamahamba Merere, who was forced to turn his attentions increasingly to the west of Usangu as a result of constant pressure from the Hehe in the east. Ultimately Merere was forced to flee from Usangu and settle among the Safwa. Sangu warriors raided for cattle, women and slaves throughout the eastern half of the Nyasa-Tanganyia corridor. In the process Merere conquered most of the Safwa and Nyiha chiefs, who remained subject to the Sangu until the closing years of the nineteenth century, when the German administration restored the Sangu (under Merere’s son and successor, Mugandilwa Merere) to their homeland. Merere’s final exile from Usangu began in 1877. Fleeing from the Hehe through Uwanji and (on some accounts) Ukinga, he made his first base at Idweri in the Poroto Mountains, fighting against the Poroto themselves. From there he moved to Ukukwe, where he settled by force on the lands of chief Mwakalinga. The cold and damp climate of Ukukwe, however, proved intolerable to the Sangu, who then pressed further west: “Merere could not stay even in Kondeland: his people could not bear the dampness of the Konde climate, being accustomed instead to the high and dry steppe [of their homeland, Usangu]. That is why Merere sent some of his subjects to Usafwa; but they were killed by the

10

Safwa. Merere then broke out of Kondeland and moved to the north-west over the Marira mountains, through the land of chiefs Mwamurima and Jirima, and into the country of chief Mwaxyambwa, on the boundary between Usafwa and Unyixa. He came to Mbinza, the village of the chief’s son Serengeni, overran and drove him away, destroyed the village, abducted women and children, and then made himself a stockade in the ruined village.” (Kootz-Kretschmer 1929b: 157; my translation, retaining original spellings)

Although there are a number of references in the narratives recorded by KootzKretschmer to Merere’s passage through Umalila, they do not tell us what, if any, effect this had on the local population (1929a: 172, 209, 260). One account is, however, a little more specific about the route taken by the Sangu chief: from the Rungwe area (‘Nkonde’), over ‘Magandja’ and the mountains in ‘Marira’ and ‘Jirima’, and on into chief Mwamlima’s country (1929a: 172). By the early 1880s Merere had established a more permanent base in Usafwa, in Mwalyego’s chiefdom. This large village was subsequently surrounded by a stone fortification and became the Sangu capital-in-exile, called Utengule like earlier (and later) Sangu royal villages. Safe behind the walls of their new capital, the Sangu successfully resisted subsequent Hehe attacks, and, when not on the defensive, launched numerous raids from it against their weaker neighbours throughout the region. There are two references in the accounts collected by Kootz-Kretschmer to raids in the Malila area, one an attack on the ‘Vapigu’ (1929a: 249), a name now associated with the northern Lambya. The other attack is described in the narrative of Murotwa Ntamanta, a great-grandson of the Safwa chief Nsweve (who was killed by the Sangu chief Mwahavanga) and grandson of Nswira (who returned from exile to chase the Sangu away). The informant’s father, Ntamanta Nswira, fled to Isongore in chief Jirima’s country, to escape the Sangu under Merere; and this is where he was born: “However, while my parents were living there, the owners of the land, the Marira, complained and said: ‘The Safwa should leave our country. Since they came here wild pigs have eaten the potatoes in our fields, whereas in the past we had no wild pigs. The Safwa are responsible: they change into wild pigs at night and eat our potatoes!’ And this is what they said to my father and his brother: ‘Go back home to Usafwa!’ This is when they realised that the Marira hated them. However, their enemies, the Sangu, were ruling in their own land. So they sat down to discuss matters and decided: ‘We must go to Merere and humble ourselves before him.’ My father took a mat and a dog, and went with them to Utengule to see Merere, the ruler of the Sango. My father said to him: ‘Oh ruler, we have come to surrender ourselves to you; we willingly return to our country, Ivindji; we will be your subjects and work for you.’ Merere gave his consent, telling him: ‘You can build a home in your own country; I have not once taken it from you.’ Thereafter my father left Marira, returned home and built a house in the village of Igongo. After we had left Marira, the Sango went there, killed a number of people, and took others captive and brought them back to Utengule.” (1929a: 198-199; my translation)

Jirima is mentioned in a number of Kootz-Kretschmer’s texts. However, it is not clear whether he was an independent chief or was subordinate to Mwaluvanda.

11

First European contact A European presence at the north end of Lake Nyasa was first established in the late 1870s. In 1875 Scottish missionaries of the Livingstonia mission landed on the northern shore of the lake (in Ngonde), and in 1877 the British Consul in Zanzibar, J. Frederic Elton, became the first European to lead an expedition over the mountains to the north. The first European to travel through Umalila, however, was the young Scottish explorer, Joseph Thomson. In 1879 he travelled across the lakeshore plain of Unyakyusa, through Ukukwe, and on into Umalila, en route to Unyamwanga and Ufipa. Thomson described a ‘Nyika village’ in Umalila as follows: “Owing to the almost constant state of warfare in which they live, the Wanyika are compelled to live in stockaded villages. The huts are huddled as closely together as possible, leaving barely room to creep about among them. The area to be defended is thus lessened. At night their cattle are brought within the stockade, filling up all the odd spaces; and as the filth is never removed, the frightful condition of the interior or a Wanyika village may be conceived.” (1881: I, 285-286)

In the 1880s Livingstonia missionaries extended their work in the region, but did not establish a permanent presence north of the lake. Shortly thereafter they were displaced by German Lutheran and Moravian missionaries, who began work in this area in 1891. Although mission stations, both short- and long-lived, were founded among neighbouring peoples, none was established among the Malila themselves during this early period. Umalila appears largely to have been bypassed by both missionary and military travellers, and as a result we possess no detailed published reports about the Malila and their country (Wright 1971: passim.; Wilson 1977: 15). The German colonial period German administration at the north end of Lake Nyasa was established in January 1893, some eighteen months after the arrival of the first German missionaries. The new administration arrived in the form of a party led by Hermann von Wissman, Imperial Commissioner and Commander-in-Chief, with orders to take possession of the territory in the south of German East Africa secured under the terms of the AngloGerman Agreement of July 1890. Von Wissman’s party founded a station, Langenburg (later Alt Langenburg), at Rumbira Bay, and began to extend their authority over the region with a combination of diplomatic pressure and military force. Alt Langenburg remained the administrative centre of Langenburg district until the end of 1900, when the district seat was moved northwards to Tukuyu, which the Germans christened Neu Langenburg (Charsley 1969: 108-114) Umalila was administered as part of Langenburg district. A German military handbook describing conditions in 1910-11 provides the following scant information on the Malila:

12

The Malila in 1910-11 Name of tribe; seat of tribe; where thickly settled. Population (men fit for bearing arms). Muzzle-loaders; other arms. Past wars; capabilities; methods of warfare. Attitude towards German Government; its influence; tribal organization. Language; use of Swahili. Dwelling-places; how far defensible. Manner of life and subsistence; cattle; donkeys.

9. WAMALILA: inhabit highland of Malila. 1,700 (567) Muzzle-loaders, spears. ---Are timid and often flee at the approach of a European. Under many small Jumbes [headmen]. Kimalila. No Swahili. ---Agriculture and cattle. Much eleusine [finger millet] and maize; potatoes. (adapted from Admiralty 1920: 66-67)

The overriding impression derived from this report is that the Malila were comparatively isolated from and only weakly integrated in the German administration. Given that they were a minor ethnic group of no military importance, the Germans, for their part, had little need to pay much attention to them. It is interesting to note that the Germans did not recognise (or at least chose not to mention) the existence of a centralised chiefship in Umalila. This may reflect no more than a lack of information or interest: otherwise it suggests that the authority of the Mwaluvanda chiefship was weak and/or somewhat localised in the early colonial period. The British colonial period The most significant event during the early years of British administration was the institution of ‘Indirect Rule’ in 1926, ten years after the British had taken over from the Germans. Subsequent efforts to rationalise local administration by combining smaller into larger units can be traced in successive Annual Reports of the Provincial Commissioners on Native Administration. In 1925 the British recognised 30 separate ‘Tribal Units’ in the western part of Mbeya district. In 1926 this number was reduced to seven units (Nyamwanga, Nyiha, Wungu, Malila, Safwa, Wanda and Songwe), amalgamated for financial purposes into a single Unyamwanga (or Unamwanga) Native Treasury. In 1928 an attempt was made to amalgamate these units further into a single federation subordinate to the Nyamwanga chief. However, this led to considerable resentment, and the new federation had to be dissolved in 1929: only financial amalgamation (now of six units) was retained. Not to be thwarted, the British continued to press for further rationalisation. At a meeting of chiefs in September 1935 it was unanimously decided to form a new and larger federation and council of chiefs. The amalgamated Nyamwanga Native Treasury was combined with those of the Sangu, Kimbu and Kipembawe into one, and the new council of chiefs was given the mandate to decide on how to apportion revenues. The council was also given deliberative functions, although each chief

13

retained executive authority in his own area. At the same time all the chiefs’ courts were upgraded (to Grade ‘A’), with the council of chiefs, which met quarterly, functioning (together with the District Officer) as an appeal court. Throughout this process, then, the Malila unit remained with a degree of autonomy. Given the German’s relative lack of interest in Umalila, it would be interesting to know why the British chose it as the focus of one of their smaller ‘Tribal Units’, and to know why it endured, albeit with reduced powers like other units in Mbeya district. It would also be interesting to know to what degree the Malila unit was dominated by the Malila themselves, and what (if any) role other ethnic / linguistic groups played within it. Like Usafwa, and to a lesser extent Unyiha, Umalila certainly admitted considerable numbers of Nyakyusa immigrants from Rungwe district: this pattern was noted in 1937, and ascribed to the weakness of the Safwa and Malila chiefs. By 1940 it was recognised that the district federation of chiefs lacked “vitality and unity”, though it was “a valuable forum for discussion”. As the following comments suggest, the British became less and less happy with the native administration of Malila as the decade progressed: “The administration of the Umalila Chiefdom continues to be carried out satisfactorily by the Chief…” (Annual Reports of the Provincial Commissioners for the Year 1945, Dar es Salaam, 1946: 80) “In Mbeya, the Native Authorities of Usangu, Usafwa (proper), Umalila, Unyiha and Unyamwanga are adequate if not brilliant.” (Annual Reports of the Provincial Commissioners for the Year 1947, Dar es Salaam, 1948: 108) “The Native Authorities in the Mbeya District remain with two exceptions unimpressive. Many are reactionary, some lethargic, and some inactive through old age. Chiefs Lyoto of Usafwa and Mwaliego of Usongwe are the only two who seem to administer their areas with a proper attention to duty.” (Annual Reports of the Provincial Commissioners for the Year 1948, Dar es Salaam, 1949: 106)

In the 1950s attention shifted away from the chiefs and towards the development of a hierarchical council system. By 1959 Mbeya could boast of possessing district, divisional, sub-chiefdom, and parish councils, although it was admitted that most divisional councils were superfluous and that parish councils only existed in Usafwa and Umalila. The development of a local government structure which was not dependent upon the energy and/or interest of individual chiefs effectively paved the way for the institutional changes which were made shortly after Tangnayika was granted independence, and which included the abolition of chiefship as a government office.

Ethnography Very little has been written about the Malila; in part because Umalila has only been visited briefly by professional anthropologists working among neighbouring peoples. Monica Wilson, who reports that she walked through Umalila in 1954, did little more than remark on the survival of traditional skin clothing and bead ornaments in this period (see above). Mariam Slater took in the Malila on her first tour in search of a

14

field site, but has equally little to say about them (see below). Our knowledge of Malila ethnography is therefore extremely scant, and the following notes have been pieced together from very disparate sources. Kinship and the Family Mariam Slater’s sole ethnographic observation on the Malila is contained in the following short statement about their kinship system: “…I verified that although Malila, Lambya, and Nyika [Mbozi Nyiha] are mutually intelligible, their kinship systems differ sufficiently to classify them as separate groups.” (1976: 61)

Unfortunately Slater does not elaborate further. Given the historical relationship between and close proximity of different Nyika-speaking peoples, we should not be surprised to find frequent intermarriage between them, whatever the differences in their kinship systems. Indeed, the only detailed case we have falls into this category: One of Kootz-Kretschmer’s Safwa informants, Murotwa Ntamanta, who was born in Isongore in Umalila during the period of Sangu rule in Usafwa (see above), returned to Umalila to take a Malila wife (Havirimbi Ndwirije) in 1902. The bride’s father, Mwambanga, told him either to bring a cow or to build a house in Umalila if he was to be granted her hand. Unable to afford a cow, Murotwa went to live with his mother-in-law, Ndipupo Namesa, who was a well-known medicine-woman. She did not live with his father-in-law, but had borne three children by three different husbands, divorcing and remarrying after the birth of each child (presumably Murotwa’s bride was the daughter of an early union). Murotwa presented his fatherin-law with six goats, two hoes, and a piece of cloth worth two rupees. He also brought a girl, his niece, to live and work for Mwambanga, saying that he would redeem her once he had provided a cow. He was thus given Ndwirije as his wife and built a house there. In 1903 Murotwa’s new wife gave birth to a child, Nsatuje. He told his parents-inlaw that he must visit home with them, and they allowed him to go in July 1903 together with his wife and child. In June 1904 Murotwa went to work as a herder at Utengule, but his wife refused to join him, remaining with her in-laws in Nswira (Murotwa’s family home). When Murotwa subsequently returned to her, she agreed to accompany him back to Utengule, but his mother-in-law refused to let her go. Only when Murotwa had given her a piece of cloth (for making clothing) worth two rupees did she relent, and in January 1905 Murotwa went back to Utengule, this time with his wife (1929a: 203-205). Murotwa does not say whether he ever completed his bridewealth payments or redeemed his niece: the implication of his account is that the bride’s mother had a much stronger influence over her subsequent movements, and therefore had to be ‘bought off’ with a piece of cloth before she would allow her daughter to travel further afield. Traditional medicine Another interesting aspect of this case (which in isolation does not tell us much about Malila marriage practices) is the reference to the mother-in-law’s vocation as a

15

medicine-woman. The practice of Malila ‘bone-medicine’ is mentioned in particular by a number of Kootz-Kretschmer’s Safwa informants, though they do not describe what this involves or have anything more to say about its Malila origin (1929a: 241242, 252, 255. Kootz-Kretschmer refers to a description of the practice in the first volume of her Die Safwa, 1926: 251). Ethnobotany In their field guide to The Mountain Flowers of Southern Tanzania (1982), Cribb and Leedal list 22 Malila plant names. This provides the only published lexical data we have on Malila, let alone on Malila ethnobotany. Malila Plant Names Scientific name RANUNCULACEAE: Clematopsis villosa (DC.) Hutch, subsp. kirkii (Oliv.) J.Raynal & Brummitt ROSACEAE: Rubus spp. Including R. porotoensis R.Graham CRASSULACEAE: Kalanchoe densiflora Rolfe LEGUMINOSAE: Indigofera spp. Including I.atriceps Hook.f., I.mimosoides, I.smutsii, I.ramosissima, I.asterocalycina and I.spathulata Kotschya recurvifolia (Taub.) F.White

Lotus goetzei Harms Tephrosia interrupta Engl. CELASTRACEAE: Catha edulis (Vahl.) Endl. OXALIDACEAE: Oxalis spp. Including O. semiloba Sond. ACANTHACEAE: Thunbergia lancifolia T.Anders CAMPANULACEAE: Lobelia gibberoa RUBIACEAE: Rubia cordifolia L. COMPOSITAE: Artemisia afra Jacq.

Malila name

Comments

ikuwi

itononkwa mtonongwa itivwa

used medicinally Ndali itifya

ivizi

intenga

idava bazivanga nzuruti

Ndali ndenga Bena matenga Kinga matenga Safwa ibanga Ndali nsuluti Nyakyusa insuluti

insevelakwale sonya piriti ibambula

Nyiha ivambula

ilumbati

has medicinal and magical uses

16

Cineraria grandiflora Vatke Notonia abyssinica A.Rich

msinde itivwa

Stoebe kilimandscharica O.Hoffm. IRIDACEAE: Gladiolus dalenii van Geel LILIACEAE: Kniphofia thomsonii Bak. Gloriosa simplex L.

izaza

Canarina eminii Schweinf. ORCHIDACEAE: Disa robusta N.E.Br. Epipactis africana Rendle

fundofundo

leaves used to draw out blood

ishilungu susumba (not given)

an extract of the tuber is used to cure earache

vigogwa ndungulingu (adapted from Cribb and Leedal 1982: passim.)

Agriculture and Land Ibala, a Malila village in the north of Ulambya (Bulambia) division, was one of the fieldwork sites chosen by van Hekken and Thoden van Velzen in 1966-68 for their study of Land Scarcity and Rural Equality in (the then) Rungwe district. The published results of this study, however, provide few details of local practice (but rather more on events in their other two field sites): “The village of Ibala is situated in the northern part of Bulambia and on the southern fringe of the Umalila plateau (2000-2300 m.). The people of this area are Malila by tribe. They live scattered throughout the accidented [sic] terrain; some families live together in small hamlets, others prefer to have their huts at some distance from the others. Ibala consists of twenty ‘ten house groups’, seven of which were selected for intensive study. These seven TANU cells comprise a total of 79 farmers with their wives and children. Land is an open resource in this part of the district. Pyrethrum, however, the important and only cash crop of the area, can only be grown on fields which have a thick blanket of volcanic ash and a permeable sub-soil. These fields form a closed resource. Since we did not measure the acreage of all pyrethrum plots within our group [of] seven TANU cells, [we] were compelled to look for another indicator of economic differentiation. We decided to choose the quantity of pyrethrum sold to the local co-operative (‘the Bulambia Co-operative Society) by the 79 farmers whom we studied. The fact that a number of farmers sold their pyrethrum via kinsmen or friends in order to escape the registration fee of 20 shs. detracts from the validity of this indicator. However, after a year in this community we had the impression that only a few farmers actually followed this practise. A general climate of suspicion in economic dealings kept many people from selling their crops through their relatives and acquaintances. This also becomes plain by the fact that women in polygynous households, who reserved a few plots for themselves and may dispose of the produce as they see fit, opted in the majority of cases for separate registration. They preferred to pay an extra 20 shs. entrance fee rather than have their husbands sell for them.” (1972: 21-22)

Having analysed incomes from pyrethrum sales, the authors conclude that there was a significant pattern of economic differentiation in Ibale. The main reason for this, they argue, was the scarcity of plots suitable for cash crop production, while secondary causes included individual differences between farmers and the number of able-

17

bodied workers per household (1972: 23). However, the evidence which van Hekken and Thoden van Velzen produce in favour of this argument is far from convincing. It certainly tells us very little about agricultural practices or economic development among the Malila in general; though this was not, of course, the primary object of study (which addressed contemporary debates about the political economy of socialist Tanzania).

Conclusion It is evident from this brief survey that we know very little about the language, history and ethnography of the Malila – little, that is, that cannot be deduced from current knowledge about some of their better-known neighbours. In particular, we lack an understanding of what makes (or once made) the Malila different from neighbouring and related peoples. To what extent can the Malila be differentiated from their Nyika neighbours? Does Malila ethnic identity have strong historical roots? How was it changed, if at all, by the policies and practices of the British colonial administration? What does being Malila mean in the late 1990s, three and a half decades after the end of colonial rule? These and related questions cannot be answered at present, and will remain unanswered until more detailed research on the language, history and ethnography of the Malila is undertaken.

Acknowledgements This paper is based primarily upon a review of sources assembled by the author while engaged in anthropological research on the Sangu and other peoples of south-west Tanzania (1979-85, 1995ongoing). I would like to thank Beverley Gartrell, John Gay, Derek Nurse, George Park and Mariam Slater for their particular inputs to this research.

References Brock, Beverley (1963) A Preliminary Description of the Nyiha People of Southwestern Tanganyika, unpublished M.A. thesis, University of Leeds. Brock, Beverley 1-30.

(1966)

‘The Nyiha of Mbozi’, Tanzania Notes and Records, 65,

Brock, Beverley (1968) ‘The Nyiha (of Mbozi)’, in Andrew Roberts (ed.) Tanzania Before 1900. Nairobi: East African Publishing House. 59-81. Busse, Joseph (1939/40) 30, 250-272.

‘Lambya-Texte’, Zeitschrift für Eingeborenen-Sprachen,

Busse, Joseph (1960) Die Sprache der Nyiha in Ostafrika (Veröffentlichung nr.41, Institut für Orientforschung, Deutsche Akademie der Wissenschaften zu Berlin). Berlin: Akadamie-Verlag. Charsley, S. R. (1969) Publishing House.

The Princes of Nyakyusa. Nairobi: East African

18

Cribb, P. J. and G. P. Leedal (1982) The Mountain Flowers of Southern Tanzania: A Field Guide to the Common Flowers. Rotterdam: A.A.Balkema. Ebner, Elzear (1987) The History of the Wangoni and their Origin in the South African Bantu Tribes. Ndanda-Peramiho: Benedictine Publications. Gartrell, Beverley (1979) ‘Is Ethnography Possible? A Critique of African Odyssey’, Journal of Anthropological Research, 35 (4), 426-446. Gay, John

(1981)

Unpublished field notes on the Nyiha, Lambya and Safwa.

Harwood, Alan (1970) Witchcraft, Sorcery, and Social Categories among the Safwa. London: Oxford University Press. van Hekken, P. M. and H. U. E. Thoden van Velzen (1972) Land Scarcity and Rural Inequality in Tanzania: Some Cases from Rungwe District. The Hague / Paris: Mouton. Johnston, Harry H. (1922) A Comparative Study of the Bantu and Semi-Bantu Languages (Vol.1). Oxford: Clarendon Press. Kalinga, Owen J. M. (1974) The Ngonde Kingdom of Northern Malawi, c.16001895, unpublished Ph.D. thesis, University of London. Kalinga, Owen J. M. (1978) ‘The Establishment and Expansion of the Lambya Kingdom c1600-1750’, African Studies Review, 21 (2), 55-66. Knight, C. Gregory (1970) Unyiha: The Ecology of Agriculture and Change in an East African Community, unpublished Ph.D. thesis, University of Minnesota. Knight, C. Gregory (1974) Ecology and Change: Rural Modernization in an African Community. New York / London: Academic Press. Kootz-Kretschmer, Elise (1929a) Die Safwa, ein ostafrikanischer Volksstamm in seinem Leben und Denken (Vol.2). Berlin: Dietrich Reimer. Kootz-Kretschmer, Elise (1929b) ‘Abriss einer Landesgeschichte von Usafwa in Ostafrika’, Koloniale Rundschau und Mitteilungen aus den Deutschen Schutzgebieten, 124-131, 152-161, 184-191, 220-226. Mwakipesile, J. S. Dar es Salaam.

(n.d.)

Mila na Desturi za Wasangu, Wasafwa na Wasagara.

Nurse, Derek (1988) ‘The Diachronic Background to the Language Communities of Southwestern Tanzania’, Sprache und Geschichte in Afrika, 9, 15-115. Nurse, Derek and George Park (1988) ‘Regional Anthropology and Historical Linguistics’, Sprache und Geschichte in Afrika, 9, 7-14.

19

Park, George (1988) ‘Evolution of a Regional Culture in East Africa’, Sprache und Geschichte in Afrika, 9, 117-204. Polomé, Edgar C. (1980) ‘The Languages of Tanzania’, in E. C. Polomé and C. P. Hill (eds.) Language in Tanzania. Oxford: Oxford University Press / International African Institute. 3-25. Slater, Mariam K. (1976) African Odyssey: An Anthropological Adventure. New York: Anchor Press / Doubleday. Thomson, Joseph (1881) To the Central African Lakes and Back: the Narrative of the Royal Geographical Society’s East Central African Expedition, 1878-80 (2 vols.). London: Low, Marston, Searle & Rivington. Walsh, Martin T. (1984) The Misinterpretation of Chiefly Power in Usangu, South-West Tanzania, unpublished Ph.D. thesis, University of Cambridge. Willis, Roy G. (1966) The Fipa and Related Peoples of South-West Tanzania and North-East Zambia (Ethnographic Survey of Africa, East Central Africa, Part XV). London: International African Institute. Wilson, Monica (1958) The Peoples of the Nyasa-Tanganyika Corridor, Communications from the School of African Studies, New Series No.29, University of Cape Town. Wilson, Monica (1977) For Men and Elders: Change in the Relations of Generations and of Men and Women among the Nyakyusa-Ngonde People 18751971. London: International African Institute. Wright, Marcia (1971) German Missions in Tanganyika 1891-1941: Lutherans and Moravians in the Southern Highlands. Oxford: Clarendon Press. Archives and official documents A Handbook of German East Africa, Admiralty, 1920. Mbeya District Book, Tanzania National Archives, Dar es Salaam (microfilm copy in Rhodes House Library, Oxford). Tanganyika Territory. Annual Reports of the Provincial Commissioners on Native Administration for the Year [1929-37]. Dar es Salaam: Government Printer, 1930-38. Tanganyika Territory. Annual Reports of the Provincial Commissioners for the Year [1938-1959]. Dar es Salaam: Government Printer, 1939-60. Tanganyika Territory. Government Printer.

Report on the Native Census, 1921.

Tanganyika Territory. Native Census, 1928 [original not seen].

20

Dar es Salaam:

Tanganyika Territory. Census of the Native Population 1931. Government Printer, 1932.

Dar es Salaam:

African Population of Tanganyika Territory. Geographical and Tribal Studies (Source: East African Population Census, 1948). East African Statistical Department, 1950. Tanganyika. African Census Report 1957. Dar es Salaam: Government Printer, 1963.

21