Thursday, October 28, 2010



  • Thursday, October 28, 2010
  • Samuel Albert
  • Share
  • By G. T Nurse , Clanship in Central Malawi pp 50-62 (1978)

    As might be expected, the clan structure of the central Ngoni of Malawi is one which has developed from the system current in the Nguni lands of Natal and Swaziland towards the end of Mfecane, the period of disturbances surrounding the rise of the Zulu power. It was as a consequence of the Mfecane that the two Ngoni migration which terminated in central and east Africa set out from their original homeland. Some of the modifications which have taken place in the clan structure of the Maseko Ngoni have been due to the exigencies of the migration, while others are recognizably the consequences of contact with the Maravi.

    The description of Bryant (1929) gives the evolution of the clans of the Zulu presents a picture very different from that to be seen among the modern Ngoni of Central Malawi. In the first place, it is at the opposite pole from the end-result of Maravi clan evolution. The clans of the Nguni in the south were territorially based , hierarchical inasmuch as each clan had and acknowledged head, and arose from fortuitous geographical isolation rather than from regular segmentation and reciprocal recognition of functions. Bryant as readily calls the groups he describes as "clans" as "tribes", and indeed some of them, such as the Zulu themselves, become identified with large entities which functioned politically rather than socially. Each clan was able to put a number of fighting men into the field under the clan leader, and the purposes of associations among clans were fundamentally military. Apparentation/affiliation linkages were acknowledged, but the alliances which depended on them were tenous and liable to severed when political circumstances were unfavourable to them.

    Neither of the Ngoni migrations to Malawi was a clan migration. The absence of any clan in Zululand or Swaziland known as Jele today might suggest that at least the whole of the Jele clan was involved in that march, but the Jele did not travel alone and the clans which accompanied them have left numerous representatives in the South. In addition, the followers were recruited on the march, from among the Tsonga, the Shona peoples and even, possibly, the Lozi. Those segments of the clans represented  which remained in the South were larger than the segments which left , or at least they all, with perhaps the sole exception of Jele, now have a more representatives in Swaziland and South Africa than they have in Malawi. Moving from clan territory deprived the migrants of the territoriality of the clans they were exporting; leaving their clan heads behind shattered another facet of their clan structure. New heads of clans were appointed by both Ngwane Maseko and Zwangendaba Jele, and clan territories were allotted once settlement in the new homeland had been consolidated. But the first setllers were well aware that the clan heads were not the real ones, and their descendants were certainly no more inclined than they were to believe otherwise. The clan lands were foreign lands, soon populated by foreigners whose fore-bearers had been temporarily displaced. Any attempt to reconstruct in Malawi the patterns of Zululand inevitably failed.

    It must be admitted, however, that the failure is hardly recognised by many of the Ngoni. Old men cling to a fiction which assigns to them the lands occupied and worked by Ntumba who have no present awareness of their claims. Regiments, partly regional, partly made up of particular clans, still exist as an educational framework for young men growing up in their age-sets; regiments which neither can, nor would, ever fight again as regiments, and whose very existence is only made apparent during gatherings the royal headquarters at Lizulu. Read (1956) records that during the 1930's she was assured that the Ntumba were steadily adopting Ngoni customs. In the 1960's I still heard this maintained, even while all around

    TABLE I:


    A. Clans listed by Read (1956) as Swazi, occurring in the lists of Kuper(1947) and Bryant (1929); found in the central Ngoni area and still encountered in 1964/68:
    Maseko Magagula Mlangeni
    ManyoniMagaseni(given by
    Read as Maganeni)
    Mashabana (now Maswani)
    Maziya Nqumayo(now Khumae and probably originally Nqumayi)Dube
    B. Clans listed as in A but not encountered in 1964/68:
    Tabete Magininda
    C.Clans listed as in A, but as found in the northern Ngoni area; encountered in the central Ngoni area in 1964/68:
    Thole Gama Sibande
    Mngomezulu Nkambule Ndlovu
    D. Swazi clans listed by Read as found in the central Ngoni area, not occurring in the list of Kuper and Bryant; still encountered in 1964/68:
    Magwagwa (Ngwangwa) Ngozo Ziphondo
    Gwati Nzunga Phungwako
    Ndumba Mauya Likhuleni
    Maile Mhlambi (now Msambi)
    E. Clans listed as in D, but not encountered in 1964/68:
    Nhlahla Ngqongwana Manokera
    (N.B.Ngwana is still used among the praise-names of the king, the Inkosi ya Makosi).
    F. Clans listed as in D, but as found in the northern Ngoni area; encountered in the central Ngoni area in 1964/68:
    Nhlane (Nsani) Gausi Jele
    G. Swazi clans not listed by Read, but occurring in the lists of Bryant and encountered in the central Ngoni area in 1964/68:
    Ndlela (Ndila) Buyeni (Bieni) Mgabi
    Ngwenya Malunga Malinga
    Ndau (Ndawu) Nyati Zulu
    Puti Msimang (Simango) Sitole
    H. Clans claiming to be Swazi, not listed by Read, and now known by names found in the lists of Bryant not as izibongo but as izitakazelo:
    Mboma (formerly Sikakane)
    Bengo or Bengu (formerly Ngcolosi)
    I. Additional clans, listed by Hodgson (1933) as "true Ngoni" or "Swazi" and as found in Dowa district; occurring in other contexts in the lists mentioned above:
    Mafuleka (Maluleka) and Nkosi (mentioned by Read as northern and occurring in the lists of Kuper and Bryant)
    Gwane (Ngwane listed by Bryant)
    J. Additional clans listed by Hodgson as "Swazi" but not found in other lists; or classed otherwise by Hodgson but present in other lists:
    Manyatera (just possibly identical with Manokera)
    Zungu (called "Thonga" by Hodgson, but mentioned by Bryant)
    K. Clans of non-Ngoni and non-Matengo origin, now accounted Ngoni:
     (a) In the central Ngoni area:
    Soko (Karanga) Moyo (Karanga) Shumba(Zezuru)
    Khwinda (Venda) Nsangu (Safwa) Nyamazani(Tsonga)
    Mlenje (Sukuma) Hara (Karanga) Thengo (Lozi)
    Cisi (Sukuma) Sawa or Shaba
    Nyai (Rozvi, Lozi or Sotho)
    (b) In Dowa district, additional to the foregoing; listed by Hodgson:
    Sambo (Thonga) Makamu (Thonga) Nthara (Sukuma)
    Sungwani (Thonga) Liwinde (Sukuma) Newa (Karanga)
    Chipeta (Sukuma) Honde (Karanga) Mlipo (Sotho)
    Mapala (Karanga) Mphepo (Sukuma) Chizwa (Sukuma)
    Cika (Sukuma) Mpumulo(Thonga) Mashatira (Safwa)

    the accommodations made by the Ngoni clan system to the far more sturdy and coherent system of the Maravi were plain to see.

    In the 1930's Read (1956) assembled a list of the clan names of the central Ngoni. Her information came mainly from people living in the Ngoni "heartland" that stretches along the Kirk range between Mvai and Domwe mountains in the Nceu and Dedza districts of Malawi and the adjacent parts of Mocambique. It seems to me incomplete, in the light of information still available in the area. In Table I I have gathered together and tentatively analysed all the information at my disposal, including the material published by Read and Hodgson (1933), using for comparisons the lists made by Kuper (1947) and Bryant (1929).

    These lists call for quite a lot of comment. The Ngoni themselves say that they came from Ndwandwe, an area in northern Zululand which immediately prior to the migration was not subject to the Zulu king Shaka but was about to be conquered by him. It is not easy to ascertain whether all the clans which claim to have originated in Ndwandwe did in fact do so. An attempt has been made recently to collate them with clans still extant or remembered there (Nurse 1973), and it has been established that such an origin is feasible for many of them. There can be virtually complete certainty about those placed in categories A, B, C and I in Table I. In category D occur some of the names said to be "Jena", such as Nzunga, Ngwagwa and Phungwako. The Jena are, within the framework of the present Ngoni political system, more of a class, an aristocracy, than a tribe. Despite these pretensions to grandeur, the absence of their clan names from lists assembled in Zululand and Swaziland gives some substance to the hypothesis that the Jena in actual fact descend from a section of the Karanga who bear that name (Nurse 1969), and that they represent the massive recruitment that was probably necessary following the defeat of Nxaba Msane by Soshangane in Gazaland shortly after the commencement of the march (Omer-Cooper 1966). On the other hand, some of the names in category A, such as Magagula, Magaseni and Masawani, are also regarded as Jena; and so are one, Moyo, which is undoubtedly of Karanga origin, and another, Ndau, which in spite of its occurrence in Bryant's list could very easily be a tribal (Shona) name assumed as a clan name, since the migration did pass through territory inhabited by the Ndau. But Mhlambi, Ngqongwana, Nhlahla and Nhlane are on phonological grounds unlikely to have originated anywhere but among speakers of Nguni languages, and it is likely that Swazi or Zulu origins can be attributed to all the names in categories D, E. and F.

    Phonological analysis can often prove quite helpful in these investigations. For instance, the absence of Gausi from the southern lists becomes explicable when one considers the changes likely to have happened to Gcabashe, listed by Bryant, in the mouths of the Maravi. The initial voiced click sound could become either an unvoiced dental affricate or a voiced velar plosive, depending on whether the voicing or the click was more distinctly heard; the bilabial implosive would become a semi-vowel and, as the voiceless alveolar fricative developed from the palato-alveolar, merge with the subsequent vowel. The product of all this would be Gausi; the terminal vowel is frequently anteriorized in Maravi versions of Nguni names.

    Contrary arguments can be used to establish the distinction between Njobvu and Ndlovu. Both mean elephant; both are clan names, one found among the Maravi, the other among the Zulu and Swazi. As a matter of ultimate cognation, the phoneme which in the Nguni languges has become dl has in Cewa become j. This, however, represents the end-result of many centuries of separate linguistic evolution. When the dl sound is articulated in Cewa surroundings today, it is not immediately heard as j but as d; the Ngoni clan name Ndlela, a path, is not pronounced by the Ntumba as Njila, a path, but as Ndila. Any genealogical connection between the Njobvu and Ndlovu clans can consequently be dismissed. The chance that they share the same totem and avoidaled them into cibale association, but it is an association by implantation, almost certainly of the paronomasic type, and nothing more. Ndlovu would, in fact, be much more likely to nave become Ndovi, a name which actually does occur in northern Malawi in the Ngoni areas, but which is not of Ngoni origin but appears to derive from the Mwandovi clan of the Lakeshore Tonga.

    The mutation of the Ngoni hl phoneme to s in Malawi is found not only in the Nhlane/Nsani transformation mentioned earlier but also in Pahla, which is now Pasa, and Mhlambi, which has become Msambi. My informants are unanimous about the identity of Nhlane with Nsani, though there are more indications than the one discussed in an earlier chapter that the matter is not simply one of direct phonological equivalence. Bryant (1929) lists a clan Msane as well as Nhlane, and it was the clan of Nxaba, one of the original leaders of the Maseko migration; the dialect of Chewa spoken in Dedza and Ntcheu districts does contain an obscure nasal variously articulated as m or a. Moreover, Faria e Souza (1674) mentions a certain "Nessani", possibly NaNsani, as the eighth wife of the then "Monomatapa", at a period when the Malawi Empire was at its height and must inevitably have had some form of contact with the court of the Mwene Mutapa. Members of the Nsani clan more often claim to be Ntumba or Cewa than Ngoni; as I have suggested above, this appears to be an instance of implantation which through paronomasia or even homophony has amounted to virtual fusion. I have also touched on the tendency for the terminal vowels of Ngoni clan names to become sharpened or frontalized in their Maravi versions, so that the equivalence of Masawani with Mashabana is not as unlikely as it might at first sight seem; in any case, it is attested by my informants and vouched for by members of the clan concerned. There is a tendency for lateral and palatal clicks, irrespective of voicing, to become velar sounds in the mouths of central Africans (Nurse 1968), and thus one finds the well- acknowledged change of Nqumayo to Khumae, and is tempted to regard Gqoli with its dark l as a possible source for Khooni, a clan which is otherwise difficult to fit into the Ngoni complex.

    There has been much discussion of the origins of the Jere clan, the original obscurity of which has been circumvented by numerous loyal inventions. The Qeko clan from which Zwangendaba, the founder of the northern Ngoni kingdom, is said to have sprung (Cullen Young 1932), is not mentioned by Bryant, who suggests that the clan of the northern Ngoni kings derives from Gumbi, which has Jele as its isitakazelo. My own field-work in Ndwandwe extracted information in support of this from at least one informant; others admitted that it could have been the case, but were not certain. The persistent claims of writers from northern Malawi (e.g. Ciwambo 1965) and those who have taken their information from northern sources (Lancaster 1937, Wiese 1900) that the Maseko were originally subordinate to the Jele seem highly unlikely (Nurse 1967, 1973). In Ndwandwe the Jele are known of only dimly, and the Gumbi are of little account; whereas in Swaziland there are still Maseko chieftains whose funeral rites are identical with those Rattray (1907) describes for Cikuse, the father of Gomani I. The highly coloured account given by Bryant (1929) of the splitting up of the Ngoni south of the Zambesi is far too circumstantial to be other than fiction. There is nonetheless a curious echo of authenticity in his mention of the name Nyamazani, that of a clan still existing among the central Ngoni and said to be of Tsonga origin. The traditions of the central Ngoni contradict the Jere claims decisively and without any resentment of them (Read 1956, Philip 1965, Nurse 1967). Archdeacon Johnson's (1922) ostensible support for these claims is very difficult to take seriously without further knowledge of his sources; even if we knew who his informants were we might not be as prepared to credit them as he was. A rather more august source for Jere has been postulated by Rangeley (1966), who believed the clan to have arisen from Kumalo. One of my informants certainly stated that Zwangendaba was originally Nxumalo, but this was probably a confusion with the foregoing. All in all, the suggestion that the clan is really Gumbi seems by far the most likely.

    The linkages of the Ngoni clans with one another are quite different in type from those found among the Maravi. The obvious explanation for this is, that the Maravi clans have evolved in the areas which are still occupied by Maravi, and most of the data required for the analysis of their evolution are still to all intents and purposes accessible in situ. The Ngoni invaders, on the other hand, departed from the country where their clan evolution had already advanced to a fairly well consolidated stage. Information about the original linkages among the Ngoni clans is not easily procurable in Malawi, and when it is it tends to be internally inconsistent and often very confusing indeed. The migrants did not themselves necessarily consist of, or even include, members of clans which were in any way linked with one another, and on the march they assimilated a number of non-Ngoni of rather diverse provenances. There was even a fundamental departure from true Nguni custom before settlement in Malawi had begun. The Nguni of the south abhor the idea of marriage between known cognate kin, while the Ngoni of Malawi practise preferential cross-cousin marriage and deny that it is something they have learnt from the Maravi. This kind of marriage is common among the Venda and the Tsonga, and it appears that when recruits were taken from those peoples there was a simultaneous assimilation of some modicum of custom.

    Nor is it particularly likely that any of the migrants who left Ndwandwe and Swaziland were especially tenacious of or well informed about the minutiae of traditional practices. For any account nowadays of associations among the original Ngoni clans it is necessary to refer back to the areas from which they came; and these have themselves changed appreciably as a result of the Mfecane and its aftermath. The closest account of what they were like before that is to be found in the ponderous pages of Bryant (1929), concealed beneath plea-santries so elephantine and a web of puns so puerile that it is often hard to realize that the information thus concealed and revealed is unique and probably accurate. The only Ngoni clans in Malawi which have apparentation/affiliation relationships, which in fact they no longer acknowledge, are Nkosi and Ngwane, which sprang by splitting from Mlangeni before the migration. Of the rest, Gama, Magagula, Maseko, Msimango and Mngomezulu are said by the central Ngoni, and by Bryant, to have come from much the same geographical area: all are first heard of in the Usutu valley in Swaziland, and may very well all be ultimately of Sotho origin. Maseko is certainly still extant among the southern Sotho, and Msimang or Msimango is distinctly Sotho in form, though Mngomezulu is not and the other two are equivocal. Puti and Nyai, besides being clan names, are sometimes used by the central Ngoni as generic terms for the Sotho peoples. The latter name is not mentioned by Bryant, but the former is said by him also to come from the Usutu valley. Ndlela and Ndlovu claim a common parent clan in Cunu, which is not found in Malawi, neither is Xulu, the alleged parent clan of Mboma, which is actually an isitakazelo name for Sikakane, and Buyeni, which is known in Malawi as Bieni. Malinga is definitely Swazi, and the rest of the clans with undoubted origins in the south come from various parts of northern and eastern Natal. Malunga will be discussed below.

    The linkages recognized among the clans today are by no means consistent with their origins. There is no acknowledgement of brotherhood between Nkosi and Ngwane or between Mboma and Bieni, while some marriage prohibitions between clans have obviously arisen quite recently. For instance, I have been told that cibale exists between Maseko and Nyamazani, but also that the mother of Cikuse was NaNyamazani. These two pieces of data are not compatible unless one recognizes that a change in the pattern of relationship between the clans has come about within the past hundred or so years. Nyamazani is said to be a Tsonga clan, and presumably NaNyamazani was a wife acquired by Mputa Maseko on the mbefore full assimilation of the Tsonga recruits into the migrating party had taken place. The later relationship suggests not so much an association on the Maravi pattern of implantation by accommodation, paronomasia or adoption as an extension on clan level of the original Nguni prohibition of marriage between cognates. It is consequently not even a variant type of implantation, but rather a projection from a particular to a general case. It would have been interesting to determine whether any similar relationships had arisen through cognation between non-royal Ngoni clans.

    The brotherhood between Kambule and Mlangeni which exists in Malawi is not found in Swaziland or Ndwandwe; nor is it possible that that between Msimango and Soko could have sprung up there, since Soko is a clan of Karanga origin. This latter association may indeed represent a form of paronomasic implantation at a different level from that described earlier. Both clans take as their totem the monkey, and this certainly happened before they were ever in contact, since the Soko of Rhodesia and the Msimang of Swaziland still have the same totem. Apart from this isolated and perhaps fortuitous instance, and the possible role played by cognation, it appears that all internal linkages among Ngoni clans, and of Ngoni clans with those which joined the migration, are of only the one type: that of implantation by status or service adoption. An apparent exception, and a possible parallel with the relationship between Soko and Simango, is the rather tenuously persistent connection between Ndila and Ndlovu. Though it is claimed by the central Ngoni that this is genuinely genealogical, and on the foregoing hypothesis it ought to be due to adoption, it could nave arisen on account of the association of Ndila with Maseko and the consequent indirect imposition on the former of the royal taboo on the eating of elephant meat. The linkage with Ndlovu could hence have arisen paronomasically.

    It is consequently not surprising that similar implantation linkages should have grown up between Ngoni and Maravi clans. For military reasons, and mainly on account of their own relatively small numbers, the Ngoni have shown a constant eagerness to assimilate the peoples among whom they have travelled and settled. A sense of clan identity is emotionally stronger and functionally more efficacious than any consciousness of membership of a tribe. The natural desire of a settled people to accommodate the new into the familiar would hasten and facilitate the acceptance by the Maravi of the newcomers into a pattern of linkages which by that time would no longer have appeared unduly foreign to the Ngoni. Today the demarcation between Ngoni and Ntumba is not abrupt. There are large numbers of intermediate people who will claim at one time to belong to one tribe, at others to the other, and who can produce clan names to substantiate whichever claim happens at the time to suit them best. But the legitimization of these manoeuvres is to be found not in the Ngoni but in the Maravi approach to clanship. Victi victorem capiunt; the military conquest by the Ngoni is in the process of being turned by the Maraavi Ntumba into aa social triumph for themselves.

    Note from the moderator: izibongo means family name while izithakazelo is kinship group praises


    Post a Comment